{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.064693659796709,
  "eval_steps": 3000,
  "global_step": 54000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.691954421997071,
      "epoch": 0.0005615769079575448,
      "grad_norm": 13.75,
      "learning_rate": 2e-06,
      "loss": 10.8494,
      "mean_token_accuracy": 0.0,
      "num_tokens": 10897.0,
      "step": 5
    },
    {
      "entropy": 10.692028713226318,
      "epoch": 0.0011231538159150896,
      "grad_norm": 13.375,
      "learning_rate": 4.5e-06,
      "loss": 10.783,
      "mean_token_accuracy": 7.102272938936949e-05,
      "num_tokens": 23084.0,
      "step": 10
    },
    {
      "entropy": 10.69119052886963,
      "epoch": 0.0016847307238726343,
      "grad_norm": 10.5,
      "learning_rate": 7e-06,
      "loss": 10.5147,
      "mean_token_accuracy": 0.01880119484849274,
      "num_tokens": 35810.0,
      "step": 15
    },
    {
      "entropy": 10.683763885498047,
      "epoch": 0.0022463076318301792,
      "grad_norm": 6.15625,
      "learning_rate": 9.5e-06,
      "loss": 10.1375,
      "mean_token_accuracy": 0.03672755789011717,
      "num_tokens": 46019.0,
      "step": 20
    },
    {
      "entropy": 10.640995311737061,
      "epoch": 0.002807884539787724,
      "grad_norm": 4.375,
      "learning_rate": 1.2e-05,
      "loss": 9.8065,
      "mean_token_accuracy": 0.0391056377440691,
      "num_tokens": 57272.0,
      "step": 25
    },
    {
      "entropy": 10.578758907318115,
      "epoch": 0.0033694614477452686,
      "grad_norm": 3.28125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 9.6789,
      "mean_token_accuracy": 0.036402178928256035,
      "num_tokens": 68686.0,
      "step": 30
    },
    {
      "entropy": 10.573095703125,
      "epoch": 0.003931038355702813,
      "grad_norm": 2.90625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 9.5714,
      "mean_token_accuracy": 0.04141928404569626,
      "num_tokens": 79187.0,
      "step": 35
    },
    {
      "entropy": 10.575594806671143,
      "epoch": 0.0044926152636603585,
      "grad_norm": 2.75,
      "learning_rate": 1.95e-05,
      "loss": 9.5576,
      "mean_token_accuracy": 0.044133822247385976,
      "num_tokens": 90293.0,
      "step": 40
    },
    {
      "entropy": 10.556080150604249,
      "epoch": 0.005054192171617903,
      "grad_norm": 2.578125,
      "learning_rate": 2.2e-05,
      "loss": 9.4673,
      "mean_token_accuracy": 0.04080899097025394,
      "num_tokens": 100420.0,
      "step": 45
    },
    {
      "entropy": 10.538790798187256,
      "epoch": 0.005615769079575448,
      "grad_norm": 2.515625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.4858,
      "mean_token_accuracy": 0.03646625243127346,
      "num_tokens": 111198.0,
      "step": 50
    },
    {
      "entropy": 10.547248840332031,
      "epoch": 0.006177345987532993,
      "grad_norm": 2.34375,
      "learning_rate": 2.7e-05,
      "loss": 9.3988,
      "mean_token_accuracy": 0.05780095718801022,
      "num_tokens": 124723.0,
      "step": 55
    },
    {
      "entropy": 10.461988258361817,
      "epoch": 0.006738922895490537,
      "grad_norm": 2.46875,
      "learning_rate": 2.95e-05,
      "loss": 9.2587,
      "mean_token_accuracy": 0.061974577233195306,
      "num_tokens": 136366.0,
      "step": 60
    },
    {
      "entropy": 10.363299369812012,
      "epoch": 0.007300499803448083,
      "grad_norm": 2.4375,
      "learning_rate": 3.2e-05,
      "loss": 9.1817,
      "mean_token_accuracy": 0.06283607967197895,
      "num_tokens": 146536.0,
      "step": 65
    },
    {
      "entropy": 10.396018791198731,
      "epoch": 0.007862076711405626,
      "grad_norm": 2.21875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.097,
      "mean_token_accuracy": 0.0707736562937498,
      "num_tokens": 157005.0,
      "step": 70
    },
    {
      "entropy": 10.386055850982666,
      "epoch": 0.008423653619363172,
      "grad_norm": 2.328125,
      "learning_rate": 3.7e-05,
      "loss": 9.017,
      "mean_token_accuracy": 0.07136933580040931,
      "num_tokens": 167864.0,
      "step": 75
    },
    {
      "entropy": 10.352816581726074,
      "epoch": 0.008985230527320717,
      "grad_norm": 2.5625,
      "learning_rate": 3.95e-05,
      "loss": 8.9484,
      "mean_token_accuracy": 0.06433885768055916,
      "num_tokens": 179635.0,
      "step": 80
    },
    {
      "entropy": 10.24338674545288,
      "epoch": 0.00954680743527826,
      "grad_norm": 2.671875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 8.8572,
      "mean_token_accuracy": 0.07211834266781807,
      "num_tokens": 190614.0,
      "step": 85
    },
    {
      "entropy": 10.245021152496339,
      "epoch": 0.010108384343235806,
      "grad_norm": 2.21875,
      "learning_rate": 4.45e-05,
      "loss": 8.7198,
      "mean_token_accuracy": 0.07616456523537636,
      "num_tokens": 201496.0,
      "step": 90
    },
    {
      "entropy": 10.094779968261719,
      "epoch": 0.010669961251193351,
      "grad_norm": 2.46875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.6055,
      "mean_token_accuracy": 0.0780682846903801,
      "num_tokens": 211606.0,
      "step": 95
    },
    {
      "entropy": 10.115865230560303,
      "epoch": 0.011231538159150897,
      "grad_norm": 2.625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.4856,
      "mean_token_accuracy": 0.07667104676365852,
      "num_tokens": 223503.0,
      "step": 100
    },
    {
      "entropy": 10.032721042633057,
      "epoch": 0.01179311506710844,
      "grad_norm": 2.421875,
      "learning_rate": 5.2e-05,
      "loss": 8.4196,
      "mean_token_accuracy": 0.0808894768357277,
      "num_tokens": 234744.0,
      "step": 105
    },
    {
      "entropy": 9.90083703994751,
      "epoch": 0.012354691975065986,
      "grad_norm": 1.7890625,
      "learning_rate": 5.45e-05,
      "loss": 8.3419,
      "mean_token_accuracy": 0.08203728273510932,
      "num_tokens": 247663.0,
      "step": 110
    },
    {
      "entropy": 9.945332717895507,
      "epoch": 0.012916268883023531,
      "grad_norm": 1.96875,
      "learning_rate": 5.7e-05,
      "loss": 8.2013,
      "mean_token_accuracy": 0.08452324643731117,
      "num_tokens": 260326.0,
      "step": 115
    },
    {
      "entropy": 9.736694145202637,
      "epoch": 0.013477845790981074,
      "grad_norm": 2.1875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.0563,
      "mean_token_accuracy": 0.0793469488620758,
      "num_tokens": 271568.0,
      "step": 120
    },
    {
      "entropy": 9.62724199295044,
      "epoch": 0.01403942269893862,
      "grad_norm": 2.359375,
      "learning_rate": 6.2e-05,
      "loss": 7.8825,
      "mean_token_accuracy": 0.08820232152938842,
      "num_tokens": 282651.0,
      "step": 125
    },
    {
      "entropy": 9.494030666351318,
      "epoch": 0.014600999606896165,
      "grad_norm": 1.96875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.719,
      "mean_token_accuracy": 0.08837165534496308,
      "num_tokens": 292753.0,
      "step": 130
    },
    {
      "entropy": 9.272327136993407,
      "epoch": 0.015162576514853709,
      "grad_norm": 1.609375,
      "learning_rate": 6.7e-05,
      "loss": 7.7167,
      "mean_token_accuracy": 0.0898565836250782,
      "num_tokens": 302981.0,
      "step": 135
    },
    {
      "entropy": 9.16934871673584,
      "epoch": 0.015724153422811252,
      "grad_norm": 1.5,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.6775,
      "mean_token_accuracy": 0.08807501792907715,
      "num_tokens": 314461.0,
      "step": 140
    },
    {
      "entropy": 8.914916038513184,
      "epoch": 0.0162857303307688,
      "grad_norm": 1.6796875,
      "learning_rate": 7.2e-05,
      "loss": 7.6052,
      "mean_token_accuracy": 0.09032118767499923,
      "num_tokens": 326877.0,
      "step": 145
    },
    {
      "entropy": 8.733859825134278,
      "epoch": 0.016847307238726343,
      "grad_norm": 1.5546875,
      "learning_rate": 7.45e-05,
      "loss": 7.3998,
      "mean_token_accuracy": 0.09276190958917141,
      "num_tokens": 337831.0,
      "step": 150
    },
    {
      "entropy": 8.528983402252198,
      "epoch": 0.017408884146683887,
      "grad_norm": 1.5703125,
      "learning_rate": 7.7e-05,
      "loss": 7.337,
      "mean_token_accuracy": 0.09177347123622895,
      "num_tokens": 348233.0,
      "step": 155
    },
    {
      "entropy": 8.35361089706421,
      "epoch": 0.017970461054641434,
      "grad_norm": 1.4609375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.2213,
      "mean_token_accuracy": 0.09552499204874039,
      "num_tokens": 359113.0,
      "step": 160
    },
    {
      "entropy": 8.23711051940918,
      "epoch": 0.018532037962598977,
      "grad_norm": 2.703125,
      "learning_rate": 8.2e-05,
      "loss": 7.336,
      "mean_token_accuracy": 0.10098176226019859,
      "num_tokens": 370284.0,
      "step": 165
    },
    {
      "entropy": 8.123040294647216,
      "epoch": 0.01909361487055652,
      "grad_norm": 1.7734375,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.3538,
      "mean_token_accuracy": 0.08904306292533874,
      "num_tokens": 381415.0,
      "step": 170
    },
    {
      "entropy": 7.9991672992706295,
      "epoch": 0.019655191778514068,
      "grad_norm": 1.34375,
      "learning_rate": 8.7e-05,
      "loss": 7.2764,
      "mean_token_accuracy": 0.09914608523249627,
      "num_tokens": 392080.0,
      "step": 175
    },
    {
      "entropy": 7.944419240951538,
      "epoch": 0.020216768686471612,
      "grad_norm": 1.3359375,
      "learning_rate": 8.95e-05,
      "loss": 7.2597,
      "mean_token_accuracy": 0.09534113705158234,
      "num_tokens": 402551.0,
      "step": 180
    },
    {
      "entropy": 7.984531497955322,
      "epoch": 0.020778345594429155,
      "grad_norm": 1.296875,
      "learning_rate": 9.2e-05,
      "loss": 7.2418,
      "mean_token_accuracy": 0.09288135915994644,
      "num_tokens": 414670.0,
      "step": 185
    },
    {
      "entropy": 7.863457107543946,
      "epoch": 0.021339922502386702,
      "grad_norm": 1.1484375,
      "learning_rate": 9.45e-05,
      "loss": 7.2269,
      "mean_token_accuracy": 0.09570453688502312,
      "num_tokens": 427162.0,
      "step": 190
    },
    {
      "entropy": 7.844120883941651,
      "epoch": 0.021901499410344246,
      "grad_norm": 1.4296875,
      "learning_rate": 9.7e-05,
      "loss": 7.125,
      "mean_token_accuracy": 0.09662900492548943,
      "num_tokens": 438628.0,
      "step": 195
    },
    {
      "entropy": 7.827740716934204,
      "epoch": 0.022463076318301793,
      "grad_norm": 1.3984375,
      "learning_rate": 9.95e-05,
      "loss": 7.0534,
      "mean_token_accuracy": 0.10476038753986358,
      "num_tokens": 449248.0,
      "step": 200
    },
    {
      "entropy": 7.613120603561401,
      "epoch": 0.023024653226259337,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000102,
      "loss": 7.0291,
      "mean_token_accuracy": 0.10421562120318413,
      "num_tokens": 460815.0,
      "step": 205
    },
    {
      "entropy": 7.653258895874023,
      "epoch": 0.02358623013421688,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.1939,
      "mean_token_accuracy": 0.09638625457882881,
      "num_tokens": 471972.0,
      "step": 210
    },
    {
      "entropy": 7.694521474838257,
      "epoch": 0.024147807042174427,
      "grad_norm": 1.265625,
      "learning_rate": 0.000107,
      "loss": 7.1589,
      "mean_token_accuracy": 0.1014483891427517,
      "num_tokens": 482775.0,
      "step": 215
    },
    {
      "entropy": 7.650409984588623,
      "epoch": 0.02470938395013197,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0001095,
      "loss": 7.087,
      "mean_token_accuracy": 0.10187960937619209,
      "num_tokens": 495151.0,
      "step": 220
    },
    {
      "entropy": 7.605823087692261,
      "epoch": 0.025270960858089515,
      "grad_norm": 1.328125,
      "learning_rate": 0.000112,
      "loss": 7.0608,
      "mean_token_accuracy": 0.10176030248403549,
      "num_tokens": 506947.0,
      "step": 225
    },
    {
      "entropy": 7.604726552963257,
      "epoch": 0.025832537766047062,
      "grad_norm": 1.421875,
      "learning_rate": 0.0001145,
      "loss": 7.0892,
      "mean_token_accuracy": 0.10170613378286361,
      "num_tokens": 519153.0,
      "step": 230
    },
    {
      "entropy": 7.570750951766968,
      "epoch": 0.026394114674004605,
      "grad_norm": 1.21875,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.07,
      "mean_token_accuracy": 0.10081414431333542,
      "num_tokens": 531050.0,
      "step": 235
    },
    {
      "entropy": 7.4688334465026855,
      "epoch": 0.02695569158196215,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.006,
      "mean_token_accuracy": 0.10289591774344445,
      "num_tokens": 542744.0,
      "step": 240
    },
    {
      "entropy": 7.563732862472534,
      "epoch": 0.027517268489919696,
      "grad_norm": 1.453125,
      "learning_rate": 0.000122,
      "loss": 7.0923,
      "mean_token_accuracy": 0.0983924463391304,
      "num_tokens": 554748.0,
      "step": 245
    },
    {
      "entropy": 7.534366416931152,
      "epoch": 0.02807884539787724,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0001245,
      "loss": 7.0106,
      "mean_token_accuracy": 0.10552474185824394,
      "num_tokens": 564868.0,
      "step": 250
    },
    {
      "entropy": 7.530099248886108,
      "epoch": 0.028640422305834783,
      "grad_norm": 1.375,
      "learning_rate": 0.000127,
      "loss": 7.0091,
      "mean_token_accuracy": 0.11061361208558082,
      "num_tokens": 575668.0,
      "step": 255
    },
    {
      "entropy": 7.455753898620605,
      "epoch": 0.02920199921379233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0001295,
      "loss": 7.1352,
      "mean_token_accuracy": 0.10184136554598808,
      "num_tokens": 587956.0,
      "step": 260
    },
    {
      "entropy": 7.4932270526885985,
      "epoch": 0.029763576121749874,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000132,
      "loss": 6.9785,
      "mean_token_accuracy": 0.10903615280985832,
      "num_tokens": 598246.0,
      "step": 265
    },
    {
      "entropy": 7.416919708251953,
      "epoch": 0.030325153029707418,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.1108,
      "mean_token_accuracy": 0.10414381325244904,
      "num_tokens": 610628.0,
      "step": 270
    },
    {
      "entropy": 7.3690831661224365,
      "epoch": 0.030886729937664965,
      "grad_norm": 1.453125,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.8616,
      "mean_token_accuracy": 0.1097691647708416,
      "num_tokens": 621214.0,
      "step": 275
    },
    {
      "entropy": 7.360501384735107,
      "epoch": 0.031448306845622505,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0001395,
      "loss": 6.957,
      "mean_token_accuracy": 0.1151120163500309,
      "num_tokens": 631123.0,
      "step": 280
    },
    {
      "entropy": 7.397417736053467,
      "epoch": 0.032009883753580055,
      "grad_norm": 1.296875,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.9474,
      "mean_token_accuracy": 0.10703005790710449,
      "num_tokens": 643434.0,
      "step": 285
    },
    {
      "entropy": 7.342480707168579,
      "epoch": 0.0325714606615376,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0001445,
      "loss": 6.9321,
      "mean_token_accuracy": 0.11211270317435265,
      "num_tokens": 655336.0,
      "step": 290
    },
    {
      "entropy": 7.372163486480713,
      "epoch": 0.03313303756949514,
      "grad_norm": 1.5,
      "learning_rate": 0.000147,
      "loss": 6.9245,
      "mean_token_accuracy": 0.10963715985417366,
      "num_tokens": 666591.0,
      "step": 295
    },
    {
      "entropy": 7.286645698547363,
      "epoch": 0.033694614477452686,
      "grad_norm": 1.328125,
      "learning_rate": 0.0001495,
      "loss": 6.9458,
      "mean_token_accuracy": 0.10590762570500374,
      "num_tokens": 678288.0,
      "step": 300
    },
    {
      "entropy": 7.292938613891602,
      "epoch": 0.03425619138541023,
      "grad_norm": 1.21875,
      "learning_rate": 0.000152,
      "loss": 6.8788,
      "mean_token_accuracy": 0.11080723032355308,
      "num_tokens": 688526.0,
      "step": 305
    },
    {
      "entropy": 7.20158314704895,
      "epoch": 0.03481776829336777,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.7955,
      "mean_token_accuracy": 0.12063507586717606,
      "num_tokens": 698727.0,
      "step": 310
    },
    {
      "entropy": 7.281469917297363,
      "epoch": 0.035379345201325324,
      "grad_norm": 1.359375,
      "learning_rate": 0.000157,
      "loss": 6.8411,
      "mean_token_accuracy": 0.1111941859126091,
      "num_tokens": 710042.0,
      "step": 315
    },
    {
      "entropy": 7.2135881900787355,
      "epoch": 0.03594092210928287,
      "grad_norm": 1.3125,
      "learning_rate": 0.0001595,
      "loss": 6.7838,
      "mean_token_accuracy": 0.11598038524389268,
      "num_tokens": 721390.0,
      "step": 320
    },
    {
      "entropy": 7.166607904434204,
      "epoch": 0.03650249901724041,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000162,
      "loss": 6.9676,
      "mean_token_accuracy": 0.11324010118842125,
      "num_tokens": 732949.0,
      "step": 325
    },
    {
      "entropy": 7.382533311843872,
      "epoch": 0.037064075925197955,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.983,
      "mean_token_accuracy": 0.11112697571516036,
      "num_tokens": 744943.0,
      "step": 330
    },
    {
      "entropy": 7.1175624370574955,
      "epoch": 0.0376256528331555,
      "grad_norm": 1.265625,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.7549,
      "mean_token_accuracy": 0.11743907704949379,
      "num_tokens": 756579.0,
      "step": 335
    },
    {
      "entropy": 7.059293413162232,
      "epoch": 0.03818722974111304,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.7064,
      "mean_token_accuracy": 0.11693573296070099,
      "num_tokens": 767260.0,
      "step": 340
    },
    {
      "entropy": 7.129382991790772,
      "epoch": 0.03874880664907059,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8055,
      "mean_token_accuracy": 0.11503136828541756,
      "num_tokens": 778163.0,
      "step": 345
    },
    {
      "entropy": 7.21411828994751,
      "epoch": 0.039310383557028136,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.841,
      "mean_token_accuracy": 0.11127241551876069,
      "num_tokens": 788912.0,
      "step": 350
    },
    {
      "entropy": 7.170050191879272,
      "epoch": 0.03987196046498568,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000177,
      "loss": 6.9508,
      "mean_token_accuracy": 0.1135722354054451,
      "num_tokens": 801405.0,
      "step": 355
    },
    {
      "entropy": 7.058458709716797,
      "epoch": 0.040433537372943223,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001795,
      "loss": 6.7437,
      "mean_token_accuracy": 0.11591695919632912,
      "num_tokens": 813023.0,
      "step": 360
    },
    {
      "entropy": 7.132532978057862,
      "epoch": 0.04099511428090077,
      "grad_norm": 1.15625,
      "learning_rate": 0.000182,
      "loss": 6.7692,
      "mean_token_accuracy": 0.1168873094022274,
      "num_tokens": 823993.0,
      "step": 365
    },
    {
      "entropy": 7.123906469345092,
      "epoch": 0.04155669118885831,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0001845,
      "loss": 6.8202,
      "mean_token_accuracy": 0.1134017400443554,
      "num_tokens": 836642.0,
      "step": 370
    },
    {
      "entropy": 7.020151281356812,
      "epoch": 0.04211826809681586,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000187,
      "loss": 6.6249,
      "mean_token_accuracy": 0.12558069303631783,
      "num_tokens": 847475.0,
      "step": 375
    },
    {
      "entropy": 7.032984447479248,
      "epoch": 0.042679845004773405,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001895,
      "loss": 6.6947,
      "mean_token_accuracy": 0.12115299850702285,
      "num_tokens": 857949.0,
      "step": 380
    },
    {
      "entropy": 7.027848196029663,
      "epoch": 0.04324142191273095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000192,
      "loss": 6.7528,
      "mean_token_accuracy": 0.11466690823435784,
      "num_tokens": 868661.0,
      "step": 385
    },
    {
      "entropy": 7.095096111297607,
      "epoch": 0.04380299882068849,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0001945,
      "loss": 6.7906,
      "mean_token_accuracy": 0.11758541241288185,
      "num_tokens": 879651.0,
      "step": 390
    },
    {
      "entropy": 7.0685296058654785,
      "epoch": 0.044364575728646036,
      "grad_norm": 1.046875,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.7722,
      "mean_token_accuracy": 0.10885974839329719,
      "num_tokens": 891189.0,
      "step": 395
    },
    {
      "entropy": 6.994202566146851,
      "epoch": 0.044926152636603586,
      "grad_norm": 1.125,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.6748,
      "mean_token_accuracy": 0.119440957903862,
      "num_tokens": 902915.0,
      "step": 400
    },
    {
      "entropy": 7.045473241806031,
      "epoch": 0.04548772954456113,
      "grad_norm": 1.453125,
      "learning_rate": 0.000202,
      "loss": 6.7649,
      "mean_token_accuracy": 0.11777753084897995,
      "num_tokens": 914108.0,
      "step": 405
    },
    {
      "entropy": 7.049249458312988,
      "epoch": 0.046049306452518673,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.7918,
      "mean_token_accuracy": 0.11295270174741745,
      "num_tokens": 925729.0,
      "step": 410
    },
    {
      "entropy": 7.0732354640960695,
      "epoch": 0.04661088336047622,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000207,
      "loss": 6.7993,
      "mean_token_accuracy": 0.11720411777496338,
      "num_tokens": 936912.0,
      "step": 415
    },
    {
      "entropy": 6.98739104270935,
      "epoch": 0.04717246026843376,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002095,
      "loss": 6.7739,
      "mean_token_accuracy": 0.11668220460414887,
      "num_tokens": 948576.0,
      "step": 420
    },
    {
      "entropy": 6.989003896713257,
      "epoch": 0.047734037176391304,
      "grad_norm": 1.15625,
      "learning_rate": 0.000212,
      "loss": 6.6228,
      "mean_token_accuracy": 0.12309176921844482,
      "num_tokens": 959346.0,
      "step": 425
    },
    {
      "entropy": 6.867386531829834,
      "epoch": 0.048295614084348855,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002145,
      "loss": 6.6393,
      "mean_token_accuracy": 0.12302001640200615,
      "num_tokens": 970124.0,
      "step": 430
    },
    {
      "entropy": 7.04110198020935,
      "epoch": 0.0488571909923064,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.7167,
      "mean_token_accuracy": 0.12065900415182114,
      "num_tokens": 981628.0,
      "step": 435
    },
    {
      "entropy": 6.839612150192261,
      "epoch": 0.04941876790026394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002195,
      "loss": 6.6603,
      "mean_token_accuracy": 0.12483339831233024,
      "num_tokens": 992423.0,
      "step": 440
    },
    {
      "entropy": 7.010668039321899,
      "epoch": 0.049980344808221486,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000222,
      "loss": 6.6703,
      "mean_token_accuracy": 0.11580908223986626,
      "num_tokens": 1002954.0,
      "step": 445
    },
    {
      "entropy": 6.892636823654175,
      "epoch": 0.05054192171617903,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002245,
      "loss": 6.5524,
      "mean_token_accuracy": 0.12802066281437874,
      "num_tokens": 1013351.0,
      "step": 450
    },
    {
      "entropy": 6.97490382194519,
      "epoch": 0.05110349862413657,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.6699,
      "mean_token_accuracy": 0.11892755851149558,
      "num_tokens": 1023781.0,
      "step": 455
    },
    {
      "entropy": 7.029914331436157,
      "epoch": 0.051665075532094124,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.7291,
      "mean_token_accuracy": 0.12224347591400146,
      "num_tokens": 1034565.0,
      "step": 460
    },
    {
      "entropy": 6.768797492980957,
      "epoch": 0.05222665244005167,
      "grad_norm": 1.4375,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.5403,
      "mean_token_accuracy": 0.12486013248562813,
      "num_tokens": 1045595.0,
      "step": 465
    },
    {
      "entropy": 6.811560869216919,
      "epoch": 0.05278822934800921,
      "grad_norm": 1.25,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.5365,
      "mean_token_accuracy": 0.12281112596392632,
      "num_tokens": 1056084.0,
      "step": 470
    },
    {
      "entropy": 6.891090345382691,
      "epoch": 0.053349806255966754,
      "grad_norm": 1.125,
      "learning_rate": 0.000237,
      "loss": 6.5832,
      "mean_token_accuracy": 0.12444683387875558,
      "num_tokens": 1067014.0,
      "step": 475
    },
    {
      "entropy": 6.7909644603729244,
      "epoch": 0.0539113831639243,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002395,
      "loss": 6.549,
      "mean_token_accuracy": 0.1283904306590557,
      "num_tokens": 1078557.0,
      "step": 480
    },
    {
      "entropy": 6.747442579269409,
      "epoch": 0.05447296007188184,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000242,
      "loss": 6.6193,
      "mean_token_accuracy": 0.1250327579677105,
      "num_tokens": 1090541.0,
      "step": 485
    },
    {
      "entropy": 6.929131269454956,
      "epoch": 0.05503453697983939,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002445,
      "loss": 6.6901,
      "mean_token_accuracy": 0.12543862983584403,
      "num_tokens": 1101851.0,
      "step": 490
    },
    {
      "entropy": 6.778433561325073,
      "epoch": 0.055596113887796936,
      "grad_norm": 1.421875,
      "learning_rate": 0.000247,
      "loss": 6.5898,
      "mean_token_accuracy": 0.12135263979434967,
      "num_tokens": 1112728.0,
      "step": 495
    },
    {
      "entropy": 6.916776704788208,
      "epoch": 0.05615769079575448,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002495,
      "loss": 6.7185,
      "mean_token_accuracy": 0.12112508118152618,
      "num_tokens": 1124537.0,
      "step": 500
    },
    {
      "entropy": 6.7057990550994875,
      "epoch": 0.05671926770371202,
      "grad_norm": 1.3125,
      "learning_rate": 0.000252,
      "loss": 6.6177,
      "mean_token_accuracy": 0.11861735060811043,
      "num_tokens": 1135749.0,
      "step": 505
    },
    {
      "entropy": 6.887033128738404,
      "epoch": 0.05728084461166957,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002545,
      "loss": 6.523,
      "mean_token_accuracy": 0.12177179381251335,
      "num_tokens": 1147597.0,
      "step": 510
    },
    {
      "entropy": 6.736017560958862,
      "epoch": 0.05784242151962711,
      "grad_norm": 1.125,
      "learning_rate": 0.000257,
      "loss": 6.483,
      "mean_token_accuracy": 0.13104297444224358,
      "num_tokens": 1158975.0,
      "step": 515
    },
    {
      "entropy": 6.74561619758606,
      "epoch": 0.05840399842758466,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002595,
      "loss": 6.6127,
      "mean_token_accuracy": 0.12473157271742821,
      "num_tokens": 1170556.0,
      "step": 520
    },
    {
      "entropy": 6.734094333648682,
      "epoch": 0.058965575335542204,
      "grad_norm": 1.25,
      "learning_rate": 0.000262,
      "loss": 6.4825,
      "mean_token_accuracy": 0.12628087177872657,
      "num_tokens": 1180757.0,
      "step": 525
    },
    {
      "entropy": 6.786943149566651,
      "epoch": 0.05952715224349975,
      "grad_norm": 1.25,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.5112,
      "mean_token_accuracy": 0.12826752737164498,
      "num_tokens": 1192428.0,
      "step": 530
    },
    {
      "entropy": 6.773551893234253,
      "epoch": 0.06008872915145729,
      "grad_norm": 1.171875,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.5812,
      "mean_token_accuracy": 0.1194170169532299,
      "num_tokens": 1203867.0,
      "step": 535
    },
    {
      "entropy": 6.7591596126556395,
      "epoch": 0.060650306059414835,
      "grad_norm": 1.1875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.5734,
      "mean_token_accuracy": 0.1305850028991699,
      "num_tokens": 1214059.0,
      "step": 540
    },
    {
      "entropy": 6.792087364196777,
      "epoch": 0.06121188296737238,
      "grad_norm": 1.375,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6338,
      "mean_token_accuracy": 0.13109224960207938,
      "num_tokens": 1224750.0,
      "step": 545
    },
    {
      "entropy": 6.816738271713257,
      "epoch": 0.06177345987532993,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002745,
      "loss": 6.7109,
      "mean_token_accuracy": 0.11931365355849266,
      "num_tokens": 1235936.0,
      "step": 550
    },
    {
      "entropy": 6.697342395782471,
      "epoch": 0.06233503678328747,
      "grad_norm": 1.0625,
      "learning_rate": 0.000277,
      "loss": 6.5229,
      "mean_token_accuracy": 0.12435328289866447,
      "num_tokens": 1247619.0,
      "step": 555
    },
    {
      "entropy": 6.797955465316773,
      "epoch": 0.06289661369124501,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002795,
      "loss": 6.5973,
      "mean_token_accuracy": 0.12734515964984894,
      "num_tokens": 1258089.0,
      "step": 560
    },
    {
      "entropy": 6.7907819747924805,
      "epoch": 0.06345819059920256,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.6714,
      "mean_token_accuracy": 0.12250098884105683,
      "num_tokens": 1270727.0,
      "step": 565
    },
    {
      "entropy": 6.718038558959961,
      "epoch": 0.06401976750716011,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002845,
      "loss": 6.5278,
      "mean_token_accuracy": 0.12575457990169525,
      "num_tokens": 1280886.0,
      "step": 570
    },
    {
      "entropy": 6.816065120697021,
      "epoch": 0.06458134441511765,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000287,
      "loss": 6.5532,
      "mean_token_accuracy": 0.13024218007922173,
      "num_tokens": 1292205.0,
      "step": 575
    },
    {
      "entropy": 6.724169111251831,
      "epoch": 0.0651429213230752,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002895,
      "loss": 6.5387,
      "mean_token_accuracy": 0.12204434722661972,
      "num_tokens": 1304621.0,
      "step": 580
    },
    {
      "entropy": 6.828542423248291,
      "epoch": 0.06570449823103273,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000292,
      "loss": 6.6198,
      "mean_token_accuracy": 0.12465316653251649,
      "num_tokens": 1315481.0,
      "step": 585
    },
    {
      "entropy": 6.641292333602905,
      "epoch": 0.06626607513899029,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002945,
      "loss": 6.6021,
      "mean_token_accuracy": 0.12428974732756615,
      "num_tokens": 1328726.0,
      "step": 590
    },
    {
      "entropy": 6.790183639526367,
      "epoch": 0.06682765204694784,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000297,
      "loss": 6.6603,
      "mean_token_accuracy": 0.12199008911848068,
      "num_tokens": 1340151.0,
      "step": 595
    },
    {
      "entropy": 6.6274354457855225,
      "epoch": 0.06738922895490537,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002995,
      "loss": 6.4411,
      "mean_token_accuracy": 0.12323693186044693,
      "num_tokens": 1350833.0,
      "step": 600
    },
    {
      "entropy": 6.738854312896729,
      "epoch": 0.06795080586286292,
      "grad_norm": 1.234375,
      "learning_rate": 0.000302,
      "loss": 6.6111,
      "mean_token_accuracy": 0.12676038667559625,
      "num_tokens": 1362332.0,
      "step": 605
    },
    {
      "entropy": 6.6895301818847654,
      "epoch": 0.06851238277082046,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003045,
      "loss": 6.5002,
      "mean_token_accuracy": 0.13042355701327324,
      "num_tokens": 1373696.0,
      "step": 610
    },
    {
      "entropy": 6.526501417160034,
      "epoch": 0.06907395967877801,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000307,
      "loss": 6.4218,
      "mean_token_accuracy": 0.1356612041592598,
      "num_tokens": 1383553.0,
      "step": 615
    },
    {
      "entropy": 6.660169887542724,
      "epoch": 0.06963553658673555,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003095,
      "loss": 6.5499,
      "mean_token_accuracy": 0.1267920345067978,
      "num_tokens": 1394487.0,
      "step": 620
    },
    {
      "entropy": 6.846566820144654,
      "epoch": 0.0701971134946931,
      "grad_norm": 1.015625,
      "learning_rate": 0.000312,
      "loss": 6.5606,
      "mean_token_accuracy": 0.13110520616173743,
      "num_tokens": 1404928.0,
      "step": 625
    },
    {
      "entropy": 6.503131103515625,
      "epoch": 0.07075869040265065,
      "grad_norm": 1.25,
      "learning_rate": 0.0003145,
      "loss": 6.4294,
      "mean_token_accuracy": 0.1311935856938362,
      "num_tokens": 1415854.0,
      "step": 630
    },
    {
      "entropy": 6.592182874679565,
      "epoch": 0.07132026731060818,
      "grad_norm": 1.171875,
      "learning_rate": 0.000317,
      "loss": 6.4816,
      "mean_token_accuracy": 0.1329976461827755,
      "num_tokens": 1427153.0,
      "step": 635
    },
    {
      "entropy": 6.661157464981079,
      "epoch": 0.07188184421856574,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003195,
      "loss": 6.5885,
      "mean_token_accuracy": 0.12770819365978242,
      "num_tokens": 1438907.0,
      "step": 640
    },
    {
      "entropy": 6.7143834114074705,
      "epoch": 0.07244342112652327,
      "grad_norm": 1.046875,
      "learning_rate": 0.000322,
      "loss": 6.4669,
      "mean_token_accuracy": 0.12574334740638732,
      "num_tokens": 1449975.0,
      "step": 645
    },
    {
      "entropy": 6.589789390563965,
      "epoch": 0.07300499803448082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.4557,
      "mean_token_accuracy": 0.12991306632757188,
      "num_tokens": 1460815.0,
      "step": 650
    },
    {
      "entropy": 6.6092170715332035,
      "epoch": 0.07356657494243837,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.5044,
      "mean_token_accuracy": 0.1279452659189701,
      "num_tokens": 1472403.0,
      "step": 655
    },
    {
      "entropy": 6.639557504653931,
      "epoch": 0.07412815185039591,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5141,
      "mean_token_accuracy": 0.12161320745944977,
      "num_tokens": 1484015.0,
      "step": 660
    },
    {
      "entropy": 6.653767776489258,
      "epoch": 0.07468972875835346,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.5425,
      "mean_token_accuracy": 0.12279097139835357,
      "num_tokens": 1495383.0,
      "step": 665
    },
    {
      "entropy": 6.608673715591431,
      "epoch": 0.075251305666311,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.468,
      "mean_token_accuracy": 0.1286981873214245,
      "num_tokens": 1505893.0,
      "step": 670
    },
    {
      "entropy": 6.724495363235474,
      "epoch": 0.07581288257426855,
      "grad_norm": 1.15625,
      "learning_rate": 0.000337,
      "loss": 6.6061,
      "mean_token_accuracy": 0.12201442718505859,
      "num_tokens": 1518933.0,
      "step": 675
    },
    {
      "entropy": 6.7100588321685795,
      "epoch": 0.07637445948222608,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003395,
      "loss": 6.6149,
      "mean_token_accuracy": 0.1266277812421322,
      "num_tokens": 1530603.0,
      "step": 680
    },
    {
      "entropy": 6.643559837341309,
      "epoch": 0.07693603639018363,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000342,
      "loss": 6.4551,
      "mean_token_accuracy": 0.131996688246727,
      "num_tokens": 1540977.0,
      "step": 685
    },
    {
      "entropy": 6.6613860607147215,
      "epoch": 0.07749761329814119,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.5621,
      "mean_token_accuracy": 0.12197319865226745,
      "num_tokens": 1553004.0,
      "step": 690
    },
    {
      "entropy": 6.604646968841553,
      "epoch": 0.07805919020609872,
      "grad_norm": 1.140625,
      "learning_rate": 0.000347,
      "loss": 6.4726,
      "mean_token_accuracy": 0.12990233227610587,
      "num_tokens": 1563446.0,
      "step": 695
    },
    {
      "entropy": 6.59793119430542,
      "epoch": 0.07862076711405627,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003495,
      "loss": 6.5153,
      "mean_token_accuracy": 0.1260698601603508,
      "num_tokens": 1573884.0,
      "step": 700
    },
    {
      "entropy": 6.554755544662475,
      "epoch": 0.07918234402201381,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000352,
      "loss": 6.3739,
      "mean_token_accuracy": 0.13849396333098413,
      "num_tokens": 1585396.0,
      "step": 705
    },
    {
      "entropy": 6.4631202697753904,
      "epoch": 0.07974392092997136,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003545,
      "loss": 6.3757,
      "mean_token_accuracy": 0.1374616429209709,
      "num_tokens": 1596934.0,
      "step": 710
    },
    {
      "entropy": 6.508552122116089,
      "epoch": 0.08030549783792891,
      "grad_norm": 1.0625,
      "learning_rate": 0.000357,
      "loss": 6.4243,
      "mean_token_accuracy": 0.13476341739296913,
      "num_tokens": 1608069.0,
      "step": 715
    },
    {
      "entropy": 6.504626750946045,
      "epoch": 0.08086707474588645,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003595,
      "loss": 6.345,
      "mean_token_accuracy": 0.1367340750992298,
      "num_tokens": 1619532.0,
      "step": 720
    },
    {
      "entropy": 6.577574253082275,
      "epoch": 0.081428651653844,
      "grad_norm": 1.15625,
      "learning_rate": 0.000362,
      "loss": 6.5057,
      "mean_token_accuracy": 0.12562305480241776,
      "num_tokens": 1630697.0,
      "step": 725
    },
    {
      "entropy": 6.569473791122436,
      "epoch": 0.08199022856180153,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003645,
      "loss": 6.483,
      "mean_token_accuracy": 0.12953646928071977,
      "num_tokens": 1641854.0,
      "step": 730
    },
    {
      "entropy": 6.676255798339843,
      "epoch": 0.08255180546975908,
      "grad_norm": 1.078125,
      "learning_rate": 0.000367,
      "loss": 6.6127,
      "mean_token_accuracy": 0.1251482017338276,
      "num_tokens": 1653036.0,
      "step": 735
    },
    {
      "entropy": 6.500600099563599,
      "epoch": 0.08311338237771662,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003695,
      "loss": 6.3686,
      "mean_token_accuracy": 0.12898936495184898,
      "num_tokens": 1664096.0,
      "step": 740
    },
    {
      "entropy": 6.575688219070434,
      "epoch": 0.08367495928567417,
      "grad_norm": 0.96875,
      "learning_rate": 0.000372,
      "loss": 6.5405,
      "mean_token_accuracy": 0.12776357308030128,
      "num_tokens": 1677890.0,
      "step": 745
    },
    {
      "entropy": 6.513050270080567,
      "epoch": 0.08423653619363172,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003745,
      "loss": 6.4039,
      "mean_token_accuracy": 0.13092774003744126,
      "num_tokens": 1687627.0,
      "step": 750
    },
    {
      "entropy": 6.6496001243591305,
      "epoch": 0.08479811310158926,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000377,
      "loss": 6.6299,
      "mean_token_accuracy": 0.12669486477971076,
      "num_tokens": 1700843.0,
      "step": 755
    },
    {
      "entropy": 6.54553918838501,
      "epoch": 0.08535969000954681,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003795,
      "loss": 6.4485,
      "mean_token_accuracy": 0.12898636609315872,
      "num_tokens": 1712356.0,
      "step": 760
    },
    {
      "entropy": 6.5033485889434814,
      "epoch": 0.08592126691750435,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000382,
      "loss": 6.44,
      "mean_token_accuracy": 0.1329922027885914,
      "num_tokens": 1722041.0,
      "step": 765
    },
    {
      "entropy": 6.540292739868164,
      "epoch": 0.0864828438254619,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003845,
      "loss": 6.5334,
      "mean_token_accuracy": 0.12715438082814218,
      "num_tokens": 1733628.0,
      "step": 770
    },
    {
      "entropy": 6.52957124710083,
      "epoch": 0.08704442073341945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.3951,
      "mean_token_accuracy": 0.13516492396593094,
      "num_tokens": 1745174.0,
      "step": 775
    },
    {
      "entropy": 6.472682189941406,
      "epoch": 0.08760599764137698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.3261,
      "mean_token_accuracy": 0.13572794049978257,
      "num_tokens": 1757117.0,
      "step": 780
    },
    {
      "entropy": 6.4541521072387695,
      "epoch": 0.08816757454933453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.3477,
      "mean_token_accuracy": 0.13643818646669387,
      "num_tokens": 1767343.0,
      "step": 785
    },
    {
      "entropy": 6.52891354560852,
      "epoch": 0.08872915145729207,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.4341,
      "mean_token_accuracy": 0.1295755423605442,
      "num_tokens": 1778500.0,
      "step": 790
    },
    {
      "entropy": 6.495887565612793,
      "epoch": 0.08929072836524962,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.4257,
      "mean_token_accuracy": 0.13391977846622466,
      "num_tokens": 1788474.0,
      "step": 795
    },
    {
      "entropy": 6.669151258468628,
      "epoch": 0.08985230527320717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003995,
      "loss": 6.6066,
      "mean_token_accuracy": 0.12243209406733513,
      "num_tokens": 1800756.0,
      "step": 800
    },
    {
      "entropy": 6.264161491394043,
      "epoch": 0.09041388218116471,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000402,
      "loss": 6.3051,
      "mean_token_accuracy": 0.13255022764205932,
      "num_tokens": 1812172.0,
      "step": 805
    },
    {
      "entropy": 6.581796360015869,
      "epoch": 0.09097545908912226,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004045,
      "loss": 6.4369,
      "mean_token_accuracy": 0.13564810305833816,
      "num_tokens": 1822057.0,
      "step": 810
    },
    {
      "entropy": 6.388931703567505,
      "epoch": 0.0915370359970798,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.3692,
      "mean_token_accuracy": 0.13979063034057618,
      "num_tokens": 1833074.0,
      "step": 815
    },
    {
      "entropy": 6.4439905166625975,
      "epoch": 0.09209861290503735,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004095,
      "loss": 6.3689,
      "mean_token_accuracy": 0.1369178406894207,
      "num_tokens": 1844932.0,
      "step": 820
    },
    {
      "entropy": 6.451093339920044,
      "epoch": 0.09266018981299488,
      "grad_norm": 1.09375,
      "learning_rate": 0.000412,
      "loss": 6.3812,
      "mean_token_accuracy": 0.13171854838728905,
      "num_tokens": 1855591.0,
      "step": 825
    },
    {
      "entropy": 6.412005949020386,
      "epoch": 0.09322176672095243,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004145,
      "loss": 6.3368,
      "mean_token_accuracy": 0.13641790226101874,
      "num_tokens": 1866726.0,
      "step": 830
    },
    {
      "entropy": 6.4959104537963865,
      "epoch": 0.09378334362890998,
      "grad_norm": 1.140625,
      "learning_rate": 0.000417,
      "loss": 6.3763,
      "mean_token_accuracy": 0.13648990392684937,
      "num_tokens": 1877385.0,
      "step": 835
    },
    {
      "entropy": 6.375614738464355,
      "epoch": 0.09434492053686752,
      "grad_norm": 1.125,
      "learning_rate": 0.0004195,
      "loss": 6.2828,
      "mean_token_accuracy": 0.1361901879310608,
      "num_tokens": 1887905.0,
      "step": 840
    },
    {
      "entropy": 6.32490758895874,
      "epoch": 0.09490649744482507,
      "grad_norm": 1.09375,
      "learning_rate": 0.000422,
      "loss": 6.3418,
      "mean_token_accuracy": 0.13226142525672913,
      "num_tokens": 1898649.0,
      "step": 845
    },
    {
      "entropy": 6.440097618103027,
      "epoch": 0.09546807435278261,
      "grad_norm": 1.25,
      "learning_rate": 0.0004245,
      "loss": 6.3581,
      "mean_token_accuracy": 0.12981755211949347,
      "num_tokens": 1908808.0,
      "step": 850
    },
    {
      "entropy": 6.449653148651123,
      "epoch": 0.09602965126074016,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000427,
      "loss": 6.4636,
      "mean_token_accuracy": 0.13088747188448907,
      "num_tokens": 1920360.0,
      "step": 855
    },
    {
      "entropy": 6.425679159164429,
      "epoch": 0.09659122816869771,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004295,
      "loss": 6.3601,
      "mean_token_accuracy": 0.13659372106194495,
      "num_tokens": 1931082.0,
      "step": 860
    },
    {
      "entropy": 6.450051259994507,
      "epoch": 0.09715280507665525,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000432,
      "loss": 6.5104,
      "mean_token_accuracy": 0.13292263224720954,
      "num_tokens": 1942712.0,
      "step": 865
    },
    {
      "entropy": 6.353779935836792,
      "epoch": 0.0977143819846128,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004345,
      "loss": 6.2649,
      "mean_token_accuracy": 0.13690216466784477,
      "num_tokens": 1953316.0,
      "step": 870
    },
    {
      "entropy": 6.438309907913208,
      "epoch": 0.09827595889257033,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000437,
      "loss": 6.3389,
      "mean_token_accuracy": 0.13307465687394143,
      "num_tokens": 1963936.0,
      "step": 875
    },
    {
      "entropy": 6.4518403053283695,
      "epoch": 0.09883753580052788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004395,
      "loss": 6.3619,
      "mean_token_accuracy": 0.1389799363911152,
      "num_tokens": 1974502.0,
      "step": 880
    },
    {
      "entropy": 6.3525231838226315,
      "epoch": 0.09939911270848542,
      "grad_norm": 0.90625,
      "learning_rate": 0.000442,
      "loss": 6.2961,
      "mean_token_accuracy": 0.14234573692083358,
      "num_tokens": 1985746.0,
      "step": 885
    },
    {
      "entropy": 6.370750093460083,
      "epoch": 0.09996068961644297,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004445,
      "loss": 6.3998,
      "mean_token_accuracy": 0.1339979127049446,
      "num_tokens": 1997696.0,
      "step": 890
    },
    {
      "entropy": 6.476981210708618,
      "epoch": 0.10052226652440052,
      "grad_norm": 1.078125,
      "learning_rate": 0.000447,
      "loss": 6.3688,
      "mean_token_accuracy": 0.13148148730397224,
      "num_tokens": 2009462.0,
      "step": 895
    },
    {
      "entropy": 6.306058502197265,
      "epoch": 0.10108384343235806,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3004,
      "mean_token_accuracy": 0.13532110527157784,
      "num_tokens": 2019547.0,
      "step": 900
    },
    {
      "entropy": 6.276200008392334,
      "epoch": 0.10164542034031561,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.3087,
      "mean_token_accuracy": 0.1303063564002514,
      "num_tokens": 2030990.0,
      "step": 905
    },
    {
      "entropy": 6.328003215789795,
      "epoch": 0.10220699724827315,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.379,
      "mean_token_accuracy": 0.13275872617959977,
      "num_tokens": 2042488.0,
      "step": 910
    },
    {
      "entropy": 6.5090795993804935,
      "epoch": 0.1027685741562307,
      "grad_norm": 1.125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.4124,
      "mean_token_accuracy": 0.13137131109833716,
      "num_tokens": 2052822.0,
      "step": 915
    },
    {
      "entropy": 6.463271427154541,
      "epoch": 0.10333015106418825,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3703,
      "mean_token_accuracy": 0.1341596059501171,
      "num_tokens": 2063274.0,
      "step": 920
    },
    {
      "entropy": 6.4825742721557615,
      "epoch": 0.10389172797214578,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000462,
      "loss": 6.4069,
      "mean_token_accuracy": 0.13037585765123366,
      "num_tokens": 2074424.0,
      "step": 925
    },
    {
      "entropy": 6.314002275466919,
      "epoch": 0.10445330488010333,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004645,
      "loss": 6.4013,
      "mean_token_accuracy": 0.13306034207344056,
      "num_tokens": 2085776.0,
      "step": 930
    },
    {
      "entropy": 6.591546201705933,
      "epoch": 0.10501488178806087,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000467,
      "loss": 6.4721,
      "mean_token_accuracy": 0.13484944850206376,
      "num_tokens": 2096789.0,
      "step": 935
    },
    {
      "entropy": 6.330018711090088,
      "epoch": 0.10557645869601842,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004695,
      "loss": 6.3848,
      "mean_token_accuracy": 0.13674520328640938,
      "num_tokens": 2108335.0,
      "step": 940
    },
    {
      "entropy": 6.46536111831665,
      "epoch": 0.10613803560397596,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000472,
      "loss": 6.4749,
      "mean_token_accuracy": 0.13200692012906073,
      "num_tokens": 2121063.0,
      "step": 945
    },
    {
      "entropy": 6.343993902206421,
      "epoch": 0.10669961251193351,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004745,
      "loss": 6.3351,
      "mean_token_accuracy": 0.1397795021533966,
      "num_tokens": 2132442.0,
      "step": 950
    },
    {
      "entropy": 6.362326002120971,
      "epoch": 0.10726118941989106,
      "grad_norm": 1.0625,
      "learning_rate": 0.000477,
      "loss": 6.2455,
      "mean_token_accuracy": 0.1406833589076996,
      "num_tokens": 2142155.0,
      "step": 955
    },
    {
      "entropy": 6.290420150756836,
      "epoch": 0.1078227663278486,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004795,
      "loss": 6.3118,
      "mean_token_accuracy": 0.13632866069674493,
      "num_tokens": 2152923.0,
      "step": 960
    },
    {
      "entropy": 6.347664928436279,
      "epoch": 0.10838434323580615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000482,
      "loss": 6.3058,
      "mean_token_accuracy": 0.13932796120643615,
      "num_tokens": 2164640.0,
      "step": 965
    },
    {
      "entropy": 6.229414701461792,
      "epoch": 0.10894592014376368,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004845,
      "loss": 6.2222,
      "mean_token_accuracy": 0.13819631040096284,
      "num_tokens": 2175403.0,
      "step": 970
    },
    {
      "entropy": 6.34880313873291,
      "epoch": 0.10950749705172123,
      "grad_norm": 1.0625,
      "learning_rate": 0.000487,
      "loss": 6.3313,
      "mean_token_accuracy": 0.1353020928800106,
      "num_tokens": 2186031.0,
      "step": 975
    },
    {
      "entropy": 6.380644750595093,
      "epoch": 0.11006907395967878,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004895,
      "loss": 6.3329,
      "mean_token_accuracy": 0.13262429535388948,
      "num_tokens": 2198152.0,
      "step": 980
    },
    {
      "entropy": 6.278173971176147,
      "epoch": 0.11063065086763632,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000492,
      "loss": 6.338,
      "mean_token_accuracy": 0.13674701675772666,
      "num_tokens": 2209435.0,
      "step": 985
    },
    {
      "entropy": 6.350253009796143,
      "epoch": 0.11119222777559387,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004945,
      "loss": 6.2994,
      "mean_token_accuracy": 0.14112076461315154,
      "num_tokens": 2219476.0,
      "step": 990
    },
    {
      "entropy": 6.3657073974609375,
      "epoch": 0.11175380468355141,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000497,
      "loss": 6.3837,
      "mean_token_accuracy": 0.13610237762331961,
      "num_tokens": 2231131.0,
      "step": 995
    },
    {
      "entropy": 6.300810956954956,
      "epoch": 0.11231538159150896,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995,
      "loss": 6.2892,
      "mean_token_accuracy": 0.1371823087334633,
      "num_tokens": 2242309.0,
      "step": 1000
    },
    {
      "entropy": 6.2305169105529785,
      "epoch": 0.1128769584994665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499999997707492,
      "loss": 6.1273,
      "mean_token_accuracy": 0.1387334682047367,
      "num_tokens": 2252236.0,
      "step": 1005
    },
    {
      "entropy": 6.3840419292449955,
      "epoch": 0.11343853540742405,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999999883941787,
      "loss": 6.4504,
      "mean_token_accuracy": 0.1293959580361843,
      "num_tokens": 2263529.0,
      "step": 1010
    },
    {
      "entropy": 6.223633098602295,
      "epoch": 0.1140001123153816,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999999719167785,
      "loss": 6.2809,
      "mean_token_accuracy": 0.13737061843276024,
      "num_tokens": 2275318.0,
      "step": 1015
    },
    {
      "entropy": 6.3728053092956545,
      "epoch": 0.11456168922333913,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999999482752917,
      "loss": 6.3265,
      "mean_token_accuracy": 0.13787831887602806,
      "num_tokens": 2287434.0,
      "step": 1020
    },
    {
      "entropy": 6.254129695892334,
      "epoch": 0.11512326613129668,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999999174697195,
      "loss": 6.2315,
      "mean_token_accuracy": 0.13839245587587357,
      "num_tokens": 2299145.0,
      "step": 1025
    },
    {
      "entropy": 6.261599349975586,
      "epoch": 0.11568484303925422,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999998795000627,
      "loss": 6.3257,
      "mean_token_accuracy": 0.13926222771406174,
      "num_tokens": 2309841.0,
      "step": 1030
    },
    {
      "entropy": 6.315525674819947,
      "epoch": 0.11624641994721177,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999998343663223,
      "loss": 6.3193,
      "mean_token_accuracy": 0.1403449870646,
      "num_tokens": 2322143.0,
      "step": 1035
    },
    {
      "entropy": 6.261302518844604,
      "epoch": 0.11680799685516932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999997820685,
      "loss": 6.2402,
      "mean_token_accuracy": 0.13240669071674346,
      "num_tokens": 2333207.0,
      "step": 1040
    },
    {
      "entropy": 6.1786919116973875,
      "epoch": 0.11736957376312686,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999997226065974,
      "loss": 6.3105,
      "mean_token_accuracy": 0.13664863780140876,
      "num_tokens": 2345542.0,
      "step": 1045
    },
    {
      "entropy": 6.371749353408814,
      "epoch": 0.11793115067108441,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999996559806163,
      "loss": 6.3079,
      "mean_token_accuracy": 0.14061758145689965,
      "num_tokens": 2356612.0,
      "step": 1050
    },
    {
      "entropy": 6.225208377838134,
      "epoch": 0.11849272757904195,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999995821905589,
      "loss": 6.2306,
      "mean_token_accuracy": 0.13814162313938141,
      "num_tokens": 2368204.0,
      "step": 1055
    },
    {
      "entropy": 6.262781095504761,
      "epoch": 0.1190543044869995,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999995012364276,
      "loss": 6.264,
      "mean_token_accuracy": 0.14789239913225175,
      "num_tokens": 2379114.0,
      "step": 1060
    },
    {
      "entropy": 6.258517122268676,
      "epoch": 0.11961588139495703,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999994131182248,
      "loss": 6.3024,
      "mean_token_accuracy": 0.1361288011074066,
      "num_tokens": 2390562.0,
      "step": 1065
    },
    {
      "entropy": 6.30267333984375,
      "epoch": 0.12017745830291458,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999993178359535,
      "loss": 6.3561,
      "mean_token_accuracy": 0.13576339259743692,
      "num_tokens": 2402352.0,
      "step": 1070
    },
    {
      "entropy": 6.422991371154785,
      "epoch": 0.12073903521087213,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999992153896166,
      "loss": 6.3906,
      "mean_token_accuracy": 0.1379263438284397,
      "num_tokens": 2415140.0,
      "step": 1075
    },
    {
      "entropy": 6.226438283920288,
      "epoch": 0.12130061211882967,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999991057792175,
      "loss": 6.2828,
      "mean_token_accuracy": 0.13986248895525932,
      "num_tokens": 2426679.0,
      "step": 1080
    },
    {
      "entropy": 6.307008361816406,
      "epoch": 0.12186218902678722,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999989890047596,
      "loss": 6.3213,
      "mean_token_accuracy": 0.14080586433410644,
      "num_tokens": 2437735.0,
      "step": 1085
    },
    {
      "entropy": 6.244773292541504,
      "epoch": 0.12242376593474476,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999988650662467,
      "loss": 6.2433,
      "mean_token_accuracy": 0.13999149277806283,
      "num_tokens": 2448611.0,
      "step": 1090
    },
    {
      "entropy": 6.176812791824341,
      "epoch": 0.12298534284270231,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999987339636825,
      "loss": 6.1729,
      "mean_token_accuracy": 0.14464567825198174,
      "num_tokens": 2461101.0,
      "step": 1095
    },
    {
      "entropy": 6.265973329544067,
      "epoch": 0.12354691975065986,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999985956970713,
      "loss": 6.3176,
      "mean_token_accuracy": 0.13885853216052055,
      "num_tokens": 2473569.0,
      "step": 1100
    },
    {
      "entropy": 6.171710395812989,
      "epoch": 0.1241084966586174,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999984502664178,
      "loss": 6.2382,
      "mean_token_accuracy": 0.13876865357160567,
      "num_tokens": 2483970.0,
      "step": 1105
    },
    {
      "entropy": 6.363559436798096,
      "epoch": 0.12467007356657495,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999982976717261,
      "loss": 6.2895,
      "mean_token_accuracy": 0.14146780967712402,
      "num_tokens": 2494184.0,
      "step": 1110
    },
    {
      "entropy": 6.217218828201294,
      "epoch": 0.1252316504745325,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999981379130016,
      "loss": 6.3041,
      "mean_token_accuracy": 0.14307402372360228,
      "num_tokens": 2505715.0,
      "step": 1115
    },
    {
      "entropy": 6.386610078811645,
      "epoch": 0.12579322738249002,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499997970990249,
      "loss": 6.3412,
      "mean_token_accuracy": 0.14221135526895523,
      "num_tokens": 2517168.0,
      "step": 1120
    },
    {
      "entropy": 6.303906965255737,
      "epoch": 0.12635480429044757,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999977969034737,
      "loss": 6.3674,
      "mean_token_accuracy": 0.1398570343852043,
      "num_tokens": 2528296.0,
      "step": 1125
    },
    {
      "entropy": 6.249405193328857,
      "epoch": 0.12691638119840512,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999976156526814,
      "loss": 6.3056,
      "mean_token_accuracy": 0.14569928944110871,
      "num_tokens": 2540648.0,
      "step": 1130
    },
    {
      "entropy": 6.177938985824585,
      "epoch": 0.12747795810636267,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999974272378776,
      "loss": 6.1885,
      "mean_token_accuracy": 0.14313262403011323,
      "num_tokens": 2551840.0,
      "step": 1135
    },
    {
      "entropy": 6.324184131622315,
      "epoch": 0.12803953501432022,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999972316590685,
      "loss": 6.2524,
      "mean_token_accuracy": 0.1379991091787815,
      "num_tokens": 2561362.0,
      "step": 1140
    },
    {
      "entropy": 6.31357970237732,
      "epoch": 0.12860111192227774,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999970289162604,
      "loss": 6.298,
      "mean_token_accuracy": 0.14433052241802216,
      "num_tokens": 2573584.0,
      "step": 1145
    },
    {
      "entropy": 6.446721315383911,
      "epoch": 0.1291626888302353,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999968190094595,
      "loss": 6.4204,
      "mean_token_accuracy": 0.1336354859173298,
      "num_tokens": 2585255.0,
      "step": 1150
    },
    {
      "entropy": 6.054596328735352,
      "epoch": 0.12972426573819285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999966019386727,
      "loss": 6.226,
      "mean_token_accuracy": 0.14078792110085486,
      "num_tokens": 2596849.0,
      "step": 1155
    },
    {
      "entropy": 6.258234024047852,
      "epoch": 0.1302858426461504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999963777039067,
      "loss": 6.2622,
      "mean_token_accuracy": 0.13961924761533737,
      "num_tokens": 2606918.0,
      "step": 1160
    },
    {
      "entropy": 6.1849829196929935,
      "epoch": 0.13084741955410795,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499996146305169,
      "loss": 6.1952,
      "mean_token_accuracy": 0.14556236416101456,
      "num_tokens": 2617779.0,
      "step": 1165
    },
    {
      "entropy": 6.273143911361695,
      "epoch": 0.13140899646206547,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999959077424665,
      "loss": 6.2513,
      "mean_token_accuracy": 0.13997638300061227,
      "num_tokens": 2630294.0,
      "step": 1170
    },
    {
      "entropy": 6.12783145904541,
      "epoch": 0.13197057337002302,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999956620158072,
      "loss": 6.1548,
      "mean_token_accuracy": 0.14402848556637765,
      "num_tokens": 2641002.0,
      "step": 1175
    },
    {
      "entropy": 6.1911145687103275,
      "epoch": 0.13253215027798057,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999954091251987,
      "loss": 6.1448,
      "mean_token_accuracy": 0.14548660516738893,
      "num_tokens": 2652188.0,
      "step": 1180
    },
    {
      "entropy": 6.125283575057983,
      "epoch": 0.13309372718593812,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999951490706491,
      "loss": 6.1865,
      "mean_token_accuracy": 0.14291539043188095,
      "num_tokens": 2662542.0,
      "step": 1185
    },
    {
      "entropy": 6.296076965332031,
      "epoch": 0.13365530409389567,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999948818521667,
      "loss": 6.2038,
      "mean_token_accuracy": 0.1425152152776718,
      "num_tokens": 2673292.0,
      "step": 1190
    },
    {
      "entropy": 6.177773904800415,
      "epoch": 0.1342168810018532,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049999460746976,
      "loss": 6.2022,
      "mean_token_accuracy": 0.1450867235660553,
      "num_tokens": 2684042.0,
      "step": 1195
    },
    {
      "entropy": 6.199269771575928,
      "epoch": 0.13477845790981074,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999943259234377,
      "loss": 6.1304,
      "mean_token_accuracy": 0.1481417030096054,
      "num_tokens": 2694716.0,
      "step": 1200
    },
    {
      "entropy": 6.261506462097168,
      "epoch": 0.1353400348177683,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999940372132089,
      "loss": 6.2723,
      "mean_token_accuracy": 0.1427570842206478,
      "num_tokens": 2706200.0,
      "step": 1205
    },
    {
      "entropy": 6.097541093826294,
      "epoch": 0.13590161172572585,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999937413390826,
      "loss": 6.0978,
      "mean_token_accuracy": 0.14956343919038773,
      "num_tokens": 2716814.0,
      "step": 1210
    },
    {
      "entropy": 6.26879563331604,
      "epoch": 0.1364631886336834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999934383010683,
      "loss": 6.2646,
      "mean_token_accuracy": 0.13633433505892753,
      "num_tokens": 2728042.0,
      "step": 1215
    },
    {
      "entropy": 6.117309093475342,
      "epoch": 0.13702476554164092,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999931280991757,
      "loss": 6.2276,
      "mean_token_accuracy": 0.13767654895782472,
      "num_tokens": 2739093.0,
      "step": 1220
    },
    {
      "entropy": 6.323478746414184,
      "epoch": 0.13758634244959847,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999928107334146,
      "loss": 6.2653,
      "mean_token_accuracy": 0.14359876736998559,
      "num_tokens": 2750307.0,
      "step": 1225
    },
    {
      "entropy": 6.146844863891602,
      "epoch": 0.13814791935755602,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999924862037952,
      "loss": 6.1785,
      "mean_token_accuracy": 0.142026449739933,
      "num_tokens": 2760534.0,
      "step": 1230
    },
    {
      "entropy": 6.146856021881104,
      "epoch": 0.13870949626551357,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999921545103278,
      "loss": 6.2017,
      "mean_token_accuracy": 0.1402510516345501,
      "num_tokens": 2772459.0,
      "step": 1235
    },
    {
      "entropy": 6.118765497207642,
      "epoch": 0.1392710731734711,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999918156530229,
      "loss": 6.1162,
      "mean_token_accuracy": 0.15379080772399903,
      "num_tokens": 2784038.0,
      "step": 1240
    },
    {
      "entropy": 6.237395238876343,
      "epoch": 0.13983265008142864,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999914696318913,
      "loss": 6.24,
      "mean_token_accuracy": 0.14800693467259407,
      "num_tokens": 2795756.0,
      "step": 1245
    },
    {
      "entropy": 6.129542446136474,
      "epoch": 0.1403942269893862,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999911164469441,
      "loss": 6.1721,
      "mean_token_accuracy": 0.14674090445041657,
      "num_tokens": 2806242.0,
      "step": 1250
    },
    {
      "entropy": 6.290287971496582,
      "epoch": 0.14095580389734375,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999907560981925,
      "loss": 6.2556,
      "mean_token_accuracy": 0.14589913040399552,
      "num_tokens": 2817017.0,
      "step": 1255
    },
    {
      "entropy": 6.189040660858154,
      "epoch": 0.1415173808053013,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999903885856479,
      "loss": 6.1629,
      "mean_token_accuracy": 0.15110919922590255,
      "num_tokens": 2827415.0,
      "step": 1260
    },
    {
      "entropy": 6.141000080108642,
      "epoch": 0.14207895771325882,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499990013909322,
      "loss": 6.1726,
      "mean_token_accuracy": 0.14932918325066566,
      "num_tokens": 2837595.0,
      "step": 1265
    },
    {
      "entropy": 6.088430595397949,
      "epoch": 0.14264053462121637,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499989632069227,
      "loss": 6.1339,
      "mean_token_accuracy": 0.14398894980549812,
      "num_tokens": 2847906.0,
      "step": 1270
    },
    {
      "entropy": 6.23743634223938,
      "epoch": 0.14320211152917392,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999892430653747,
      "loss": 6.2346,
      "mean_token_accuracy": 0.13501434922218322,
      "num_tokens": 2859145.0,
      "step": 1275
    },
    {
      "entropy": 6.056405162811279,
      "epoch": 0.14376368843713147,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999888468977776,
      "loss": 6.1648,
      "mean_token_accuracy": 0.14927329123020172,
      "num_tokens": 2870704.0,
      "step": 1280
    },
    {
      "entropy": 6.249700832366943,
      "epoch": 0.14432526534508902,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999884435664485,
      "loss": 6.2647,
      "mean_token_accuracy": 0.14412776455283166,
      "num_tokens": 2881861.0,
      "step": 1285
    },
    {
      "entropy": 6.240557241439819,
      "epoch": 0.14488684225304654,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004999880330714,
      "loss": 6.1861,
      "mean_token_accuracy": 0.14692625999450684,
      "num_tokens": 2893247.0,
      "step": 1290
    },
    {
      "entropy": 6.075284337997436,
      "epoch": 0.1454484191610041,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999876154126453,
      "loss": 6.0185,
      "mean_token_accuracy": 0.1487583711743355,
      "num_tokens": 2905351.0,
      "step": 1295
    },
    {
      "entropy": 6.0624950408935545,
      "epoch": 0.14600999606896164,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999871905901977,
      "loss": 6.1821,
      "mean_token_accuracy": 0.14871341213583947,
      "num_tokens": 2915153.0,
      "step": 1300
    },
    {
      "entropy": 6.169287776947021,
      "epoch": 0.1465715729769192,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999867586040706,
      "loss": 6.1416,
      "mean_token_accuracy": 0.13760501518845558,
      "num_tokens": 2926234.0,
      "step": 1305
    },
    {
      "entropy": 6.069522142410278,
      "epoch": 0.14713314988487675,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999863194542779,
      "loss": 6.0943,
      "mean_token_accuracy": 0.14693765938282013,
      "num_tokens": 2937033.0,
      "step": 1310
    },
    {
      "entropy": 6.068335819244385,
      "epoch": 0.14769472679283427,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999858731408336,
      "loss": 6.0518,
      "mean_token_accuracy": 0.14859263896942138,
      "num_tokens": 2945904.0,
      "step": 1315
    },
    {
      "entropy": 6.208248567581177,
      "epoch": 0.14825630370079182,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999854196637518,
      "loss": 6.253,
      "mean_token_accuracy": 0.14146910086274148,
      "num_tokens": 2956298.0,
      "step": 1320
    },
    {
      "entropy": 6.151482725143433,
      "epoch": 0.14881788060874937,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999849590230469,
      "loss": 6.1522,
      "mean_token_accuracy": 0.14781234338879584,
      "num_tokens": 2967660.0,
      "step": 1325
    },
    {
      "entropy": 6.198898839950561,
      "epoch": 0.14937945751670692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999844912187336,
      "loss": 6.2205,
      "mean_token_accuracy": 0.1416272446513176,
      "num_tokens": 2978616.0,
      "step": 1330
    },
    {
      "entropy": 6.0869788646698,
      "epoch": 0.14994103442466447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499984016250827,
      "loss": 6.0738,
      "mean_token_accuracy": 0.15207655876874923,
      "num_tokens": 2989731.0,
      "step": 1335
    },
    {
      "entropy": 6.171562576293946,
      "epoch": 0.150502611332622,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499983534119342,
      "loss": 6.177,
      "mean_token_accuracy": 0.14310620874166488,
      "num_tokens": 3002110.0,
      "step": 1340
    },
    {
      "entropy": 5.928503322601318,
      "epoch": 0.15106418824057954,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999830448242941,
      "loss": 6.0135,
      "mean_token_accuracy": 0.15312863737344742,
      "num_tokens": 3012709.0,
      "step": 1345
    },
    {
      "entropy": 6.257901430130005,
      "epoch": 0.1516257651485371,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999825483656986,
      "loss": 6.3139,
      "mean_token_accuracy": 0.13134505301713945,
      "num_tokens": 3025738.0,
      "step": 1350
    },
    {
      "entropy": 6.151828098297119,
      "epoch": 0.15218734205649465,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999820447435716,
      "loss": 6.1416,
      "mean_token_accuracy": 0.14338411912322044,
      "num_tokens": 3037711.0,
      "step": 1355
    },
    {
      "entropy": 6.239610385894776,
      "epoch": 0.15274891896445217,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999815339579291,
      "loss": 6.3286,
      "mean_token_accuracy": 0.13714671283960342,
      "num_tokens": 3048929.0,
      "step": 1360
    },
    {
      "entropy": 6.117274618148803,
      "epoch": 0.15331049587240972,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999810160087872,
      "loss": 6.0739,
      "mean_token_accuracy": 0.15920419096946717,
      "num_tokens": 3059717.0,
      "step": 1365
    },
    {
      "entropy": 6.133962488174438,
      "epoch": 0.15387207278036727,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999804908961626,
      "loss": 6.0778,
      "mean_token_accuracy": 0.15217963755130767,
      "num_tokens": 3069530.0,
      "step": 1370
    },
    {
      "entropy": 6.133673429489136,
      "epoch": 0.15443364968832482,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999799586200719,
      "loss": 6.2715,
      "mean_token_accuracy": 0.14009033963084222,
      "num_tokens": 3080705.0,
      "step": 1375
    },
    {
      "entropy": 6.1995704650878904,
      "epoch": 0.15499522659628237,
      "grad_norm": 1.0,
      "learning_rate": 0.000499979419180532,
      "loss": 6.1198,
      "mean_token_accuracy": 0.1425781100988388,
      "num_tokens": 3092553.0,
      "step": 1380
    },
    {
      "entropy": 6.193589210510254,
      "epoch": 0.1555568035042399,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999788725775601,
      "loss": 6.2416,
      "mean_token_accuracy": 0.1423923633992672,
      "num_tokens": 3102947.0,
      "step": 1385
    },
    {
      "entropy": 6.08302493095398,
      "epoch": 0.15611838041219744,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999783188111737,
      "loss": 6.0833,
      "mean_token_accuracy": 0.15142102390527726,
      "num_tokens": 3114157.0,
      "step": 1390
    },
    {
      "entropy": 6.019034147262573,
      "epoch": 0.156679957320155,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999777578813904,
      "loss": 5.9512,
      "mean_token_accuracy": 0.14669259935617446,
      "num_tokens": 3125739.0,
      "step": 1395
    },
    {
      "entropy": 6.063040637969971,
      "epoch": 0.15724153422811254,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499977189788228,
      "loss": 6.1128,
      "mean_token_accuracy": 0.15026577562093735,
      "num_tokens": 3136970.0,
      "step": 1400
    },
    {
      "entropy": 6.108362627029419,
      "epoch": 0.1578031111360701,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999766145317046,
      "loss": 6.0792,
      "mean_token_accuracy": 0.14896431863307952,
      "num_tokens": 3148512.0,
      "step": 1405
    },
    {
      "entropy": 6.111115980148315,
      "epoch": 0.15836468804402762,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999760321118385,
      "loss": 6.1135,
      "mean_token_accuracy": 0.15590201914310456,
      "num_tokens": 3159477.0,
      "step": 1410
    },
    {
      "entropy": 6.05187873840332,
      "epoch": 0.15892626495198517,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999754425286483,
      "loss": 6.128,
      "mean_token_accuracy": 0.15035361796617508,
      "num_tokens": 3171361.0,
      "step": 1415
    },
    {
      "entropy": 6.220038843154907,
      "epoch": 0.15948784185994272,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999748457821529,
      "loss": 6.2881,
      "mean_token_accuracy": 0.13868645653128625,
      "num_tokens": 3184652.0,
      "step": 1420
    },
    {
      "entropy": 6.1836763381958,
      "epoch": 0.16004941876790027,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499974241872371,
      "loss": 6.1675,
      "mean_token_accuracy": 0.14523480385541915,
      "num_tokens": 3195557.0,
      "step": 1425
    },
    {
      "entropy": 6.168182754516602,
      "epoch": 0.16061099567585782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999736307993221,
      "loss": 6.1987,
      "mean_token_accuracy": 0.1331624984741211,
      "num_tokens": 3207139.0,
      "step": 1430
    },
    {
      "entropy": 6.287556171417236,
      "epoch": 0.16117257258381534,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999730125630255,
      "loss": 6.2968,
      "mean_token_accuracy": 0.13671731874346732,
      "num_tokens": 3218967.0,
      "step": 1435
    },
    {
      "entropy": 6.173974084854126,
      "epoch": 0.1617341494917729,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999723871635008,
      "loss": 6.1019,
      "mean_token_accuracy": 0.1492471754550934,
      "num_tokens": 3230168.0,
      "step": 1440
    },
    {
      "entropy": 6.022477769851685,
      "epoch": 0.16229572639973044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999717546007681,
      "loss": 6.1451,
      "mean_token_accuracy": 0.1437738373875618,
      "num_tokens": 3241829.0,
      "step": 1445
    },
    {
      "entropy": 6.239676189422608,
      "epoch": 0.162857303307688,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999711148748476,
      "loss": 6.1597,
      "mean_token_accuracy": 0.14430854842066765,
      "num_tokens": 3253195.0,
      "step": 1450
    },
    {
      "entropy": 6.084852933883667,
      "epoch": 0.16341888021564555,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999704679857594,
      "loss": 6.2113,
      "mean_token_accuracy": 0.14120337069034578,
      "num_tokens": 3264951.0,
      "step": 1455
    },
    {
      "entropy": 6.022989225387573,
      "epoch": 0.16398045712360307,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999698139335243,
      "loss": 6.0438,
      "mean_token_accuracy": 0.14825797379016875,
      "num_tokens": 3276054.0,
      "step": 1460
    },
    {
      "entropy": 6.0944970607757565,
      "epoch": 0.16454203403156062,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999691527181632,
      "loss": 6.0089,
      "mean_token_accuracy": 0.1612110674381256,
      "num_tokens": 3285540.0,
      "step": 1465
    },
    {
      "entropy": 6.161299562454223,
      "epoch": 0.16510361093951817,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999684843396968,
      "loss": 6.1427,
      "mean_token_accuracy": 0.14661948680877684,
      "num_tokens": 3296869.0,
      "step": 1470
    },
    {
      "entropy": 6.089122295379639,
      "epoch": 0.16566518784747572,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999678087981467,
      "loss": 6.0813,
      "mean_token_accuracy": 0.1550757810473442,
      "num_tokens": 3307554.0,
      "step": 1475
    },
    {
      "entropy": 5.974129676818848,
      "epoch": 0.16622676475543324,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999671260935343,
      "loss": 6.0369,
      "mean_token_accuracy": 0.14905909150838853,
      "num_tokens": 3319685.0,
      "step": 1480
    },
    {
      "entropy": 6.121063327789306,
      "epoch": 0.1667883416633908,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999664362258814,
      "loss": 6.1516,
      "mean_token_accuracy": 0.14900618344545363,
      "num_tokens": 3330742.0,
      "step": 1485
    },
    {
      "entropy": 6.037702798843384,
      "epoch": 0.16734991857134834,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999657391952099,
      "loss": 6.1774,
      "mean_token_accuracy": 0.14220552295446395,
      "num_tokens": 3343613.0,
      "step": 1490
    },
    {
      "entropy": 6.289458322525024,
      "epoch": 0.1679114954793059,
      "grad_norm": 0.875,
      "learning_rate": 0.000499965035001542,
      "loss": 6.2146,
      "mean_token_accuracy": 0.1457619272172451,
      "num_tokens": 3355162.0,
      "step": 1495
    },
    {
      "entropy": 6.107675218582154,
      "epoch": 0.16847307238726345,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999643236449001,
      "loss": 6.1513,
      "mean_token_accuracy": 0.15243639126420022,
      "num_tokens": 3366324.0,
      "step": 1500
    },
    {
      "entropy": 6.264118766784668,
      "epoch": 0.16903464929522097,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999636051253069,
      "loss": 6.292,
      "mean_token_accuracy": 0.13710838407278061,
      "num_tokens": 3378571.0,
      "step": 1505
    },
    {
      "entropy": 6.134381484985352,
      "epoch": 0.16959622620317852,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999628794427854,
      "loss": 6.1696,
      "mean_token_accuracy": 0.13963365852832793,
      "num_tokens": 3390605.0,
      "step": 1510
    },
    {
      "entropy": 6.0607460021972654,
      "epoch": 0.17015780311113607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999621465973583,
      "loss": 6.0866,
      "mean_token_accuracy": 0.14871912077069283,
      "num_tokens": 3402735.0,
      "step": 1515
    },
    {
      "entropy": 6.100008773803711,
      "epoch": 0.17071938001909362,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999614065890495,
      "loss": 6.0991,
      "mean_token_accuracy": 0.15326170325279237,
      "num_tokens": 3413079.0,
      "step": 1520
    },
    {
      "entropy": 6.077723407745362,
      "epoch": 0.17128095692705117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999606594178821,
      "loss": 6.0422,
      "mean_token_accuracy": 0.14912576824426652,
      "num_tokens": 3423098.0,
      "step": 1525
    },
    {
      "entropy": 6.100780582427978,
      "epoch": 0.1718425338350087,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049995990508388,
      "loss": 6.0769,
      "mean_token_accuracy": 0.14803334027528764,
      "num_tokens": 3435794.0,
      "step": 1530
    },
    {
      "entropy": 6.115665483474731,
      "epoch": 0.17240411074296624,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999591435870674,
      "loss": 6.2069,
      "mean_token_accuracy": 0.13976619020104408,
      "num_tokens": 3448125.0,
      "step": 1535
    },
    {
      "entropy": 6.065301561355591,
      "epoch": 0.1729656876509238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999583749274684,
      "loss": 6.0397,
      "mean_token_accuracy": 0.14901299476623536,
      "num_tokens": 3458217.0,
      "step": 1540
    },
    {
      "entropy": 6.002480125427246,
      "epoch": 0.17352726455888134,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999575991051074,
      "loss": 6.0901,
      "mean_token_accuracy": 0.1483938679099083,
      "num_tokens": 3469553.0,
      "step": 1545
    },
    {
      "entropy": 6.146899223327637,
      "epoch": 0.1740888414668389,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999568161200093,
      "loss": 6.0427,
      "mean_token_accuracy": 0.15437329113483428,
      "num_tokens": 3480928.0,
      "step": 1550
    },
    {
      "entropy": 6.065676021575928,
      "epoch": 0.17465041837479642,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999560259721989,
      "loss": 6.1347,
      "mean_token_accuracy": 0.1498277299106121,
      "num_tokens": 3492074.0,
      "step": 1555
    },
    {
      "entropy": 6.13408784866333,
      "epoch": 0.17521199528275397,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999552286617014,
      "loss": 6.1425,
      "mean_token_accuracy": 0.14677510857582093,
      "num_tokens": 3503320.0,
      "step": 1560
    },
    {
      "entropy": 6.015400695800781,
      "epoch": 0.17577357219071152,
      "grad_norm": 1.375,
      "learning_rate": 0.0004999544241885421,
      "loss": 6.136,
      "mean_token_accuracy": 0.15524707436561586,
      "num_tokens": 3515466.0,
      "step": 1565
    },
    {
      "entropy": 6.1630960464477536,
      "epoch": 0.17633514909866907,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999536125527467,
      "loss": 6.0079,
      "mean_token_accuracy": 0.15108589679002762,
      "num_tokens": 3525584.0,
      "step": 1570
    },
    {
      "entropy": 5.98162202835083,
      "epoch": 0.17689672600662662,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499952793754341,
      "loss": 6.0963,
      "mean_token_accuracy": 0.1469439223408699,
      "num_tokens": 3536973.0,
      "step": 1575
    },
    {
      "entropy": 6.10207724571228,
      "epoch": 0.17745830291458414,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999519677933512,
      "loss": 6.1286,
      "mean_token_accuracy": 0.14693209826946257,
      "num_tokens": 3547799.0,
      "step": 1580
    },
    {
      "entropy": 6.159058141708374,
      "epoch": 0.1780198798225417,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999511346698035,
      "loss": 6.1786,
      "mean_token_accuracy": 0.15070472061634063,
      "num_tokens": 3558436.0,
      "step": 1585
    },
    {
      "entropy": 6.149427270889282,
      "epoch": 0.17858145673049924,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999502943837244,
      "loss": 6.1659,
      "mean_token_accuracy": 0.15004989951848985,
      "num_tokens": 3569851.0,
      "step": 1590
    },
    {
      "entropy": 6.135163640975952,
      "epoch": 0.1791430336384568,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999494469351406,
      "loss": 6.0449,
      "mean_token_accuracy": 0.14947315603494643,
      "num_tokens": 3580797.0,
      "step": 1595
    },
    {
      "entropy": 6.05308346748352,
      "epoch": 0.17970461054641435,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999485923240793,
      "loss": 6.0467,
      "mean_token_accuracy": 0.1534357637166977,
      "num_tokens": 3592552.0,
      "step": 1600
    },
    {
      "entropy": 5.9738929748535154,
      "epoch": 0.18026618745437187,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999477305505675,
      "loss": 6.0287,
      "mean_token_accuracy": 0.15556652545928956,
      "num_tokens": 3603932.0,
      "step": 1605
    },
    {
      "entropy": 6.095688581466675,
      "epoch": 0.18082776436232942,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999468616146328,
      "loss": 6.052,
      "mean_token_accuracy": 0.1578151524066925,
      "num_tokens": 3614865.0,
      "step": 1610
    },
    {
      "entropy": 5.861964559555053,
      "epoch": 0.18138934127028697,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999459855163026,
      "loss": 5.9784,
      "mean_token_accuracy": 0.1593295931816101,
      "num_tokens": 3626275.0,
      "step": 1615
    },
    {
      "entropy": 6.140000581741333,
      "epoch": 0.18195091817824452,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499945102255605,
      "loss": 6.1078,
      "mean_token_accuracy": 0.1455901011824608,
      "num_tokens": 3638416.0,
      "step": 1620
    },
    {
      "entropy": 5.946600866317749,
      "epoch": 0.18251249508620204,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999442118325682,
      "loss": 5.9553,
      "mean_token_accuracy": 0.1524277485907078,
      "num_tokens": 3649651.0,
      "step": 1625
    },
    {
      "entropy": 6.132726192474365,
      "epoch": 0.1830740719941596,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999433142472205,
      "loss": 6.192,
      "mean_token_accuracy": 0.15279360264539718,
      "num_tokens": 3661172.0,
      "step": 1630
    },
    {
      "entropy": 6.108269739151001,
      "epoch": 0.18363564890211714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999424094995902,
      "loss": 6.0268,
      "mean_token_accuracy": 0.15440748035907745,
      "num_tokens": 3672086.0,
      "step": 1635
    },
    {
      "entropy": 6.029655122756958,
      "epoch": 0.1841972258100747,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999414975897066,
      "loss": 6.0021,
      "mean_token_accuracy": 0.14995888769626617,
      "num_tokens": 3682579.0,
      "step": 1640
    },
    {
      "entropy": 6.10533299446106,
      "epoch": 0.18475880271803224,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999405785175983,
      "loss": 6.1253,
      "mean_token_accuracy": 0.14912039935588836,
      "num_tokens": 3694616.0,
      "step": 1645
    },
    {
      "entropy": 5.989877414703369,
      "epoch": 0.18532037962598977,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999396522832949,
      "loss": 5.909,
      "mean_token_accuracy": 0.1535977005958557,
      "num_tokens": 3706821.0,
      "step": 1650
    },
    {
      "entropy": 5.970361423492432,
      "epoch": 0.18588195653394732,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999387188868256,
      "loss": 5.9652,
      "mean_token_accuracy": 0.1572886511683464,
      "num_tokens": 3717552.0,
      "step": 1655
    },
    {
      "entropy": 5.978477573394775,
      "epoch": 0.18644353344190487,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999377783282203,
      "loss": 6.043,
      "mean_token_accuracy": 0.15294748544692993,
      "num_tokens": 3728300.0,
      "step": 1660
    },
    {
      "entropy": 6.069652843475342,
      "epoch": 0.18700511034986242,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999368306075089,
      "loss": 6.0383,
      "mean_token_accuracy": 0.15796971321105957,
      "num_tokens": 3738545.0,
      "step": 1665
    },
    {
      "entropy": 6.053223991394043,
      "epoch": 0.18756668725781997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999358757247215,
      "loss": 6.031,
      "mean_token_accuracy": 0.14949157685041428,
      "num_tokens": 3749190.0,
      "step": 1670
    },
    {
      "entropy": 5.973500061035156,
      "epoch": 0.1881282641657775,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999349136798886,
      "loss": 6.0387,
      "mean_token_accuracy": 0.15051646381616593,
      "num_tokens": 3760978.0,
      "step": 1675
    },
    {
      "entropy": 5.94163761138916,
      "epoch": 0.18868984107373504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999339444730409,
      "loss": 5.9304,
      "mean_token_accuracy": 0.1564078763127327,
      "num_tokens": 3771339.0,
      "step": 1680
    },
    {
      "entropy": 6.000215005874634,
      "epoch": 0.1892514179816926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999329681042092,
      "loss": 5.9158,
      "mean_token_accuracy": 0.15469563752412796,
      "num_tokens": 3781550.0,
      "step": 1685
    },
    {
      "entropy": 5.960664081573486,
      "epoch": 0.18981299488965014,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999319845734243,
      "loss": 6.1339,
      "mean_token_accuracy": 0.14286689832806587,
      "num_tokens": 3793700.0,
      "step": 1690
    },
    {
      "entropy": 6.237479496002197,
      "epoch": 0.1903745717976077,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999309938807181,
      "loss": 6.1323,
      "mean_token_accuracy": 0.14619217813014984,
      "num_tokens": 3806114.0,
      "step": 1695
    },
    {
      "entropy": 5.988754749298096,
      "epoch": 0.19093614870556522,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999299960261216,
      "loss": 6.0421,
      "mean_token_accuracy": 0.15423230230808258,
      "num_tokens": 3817251.0,
      "step": 1700
    },
    {
      "entropy": 6.091027498245239,
      "epoch": 0.19149772561352277,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999289910096669,
      "loss": 6.0464,
      "mean_token_accuracy": 0.15106936246156694,
      "num_tokens": 3828544.0,
      "step": 1705
    },
    {
      "entropy": 5.994654273986816,
      "epoch": 0.19205930252148032,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999279788313858,
      "loss": 6.1189,
      "mean_token_accuracy": 0.14720231145620347,
      "num_tokens": 3840423.0,
      "step": 1710
    },
    {
      "entropy": 6.1074625015258786,
      "epoch": 0.19262087942943787,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999269594913107,
      "loss": 6.0199,
      "mean_token_accuracy": 0.15324137806892396,
      "num_tokens": 3852751.0,
      "step": 1715
    },
    {
      "entropy": 5.966987991333008,
      "epoch": 0.19318245633739542,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999259329894739,
      "loss": 6.0043,
      "mean_token_accuracy": 0.15583723783493042,
      "num_tokens": 3863382.0,
      "step": 1720
    },
    {
      "entropy": 5.978365659713745,
      "epoch": 0.19374403324535294,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999248993259081,
      "loss": 6.0004,
      "mean_token_accuracy": 0.1522401787340641,
      "num_tokens": 3874294.0,
      "step": 1725
    },
    {
      "entropy": 6.057207489013672,
      "epoch": 0.1943056101533105,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999238585006463,
      "loss": 6.0364,
      "mean_token_accuracy": 0.14614153355360032,
      "num_tokens": 3886693.0,
      "step": 1730
    },
    {
      "entropy": 5.985201740264893,
      "epoch": 0.19486718706126804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999228105137218,
      "loss": 5.9488,
      "mean_token_accuracy": 0.1587449014186859,
      "num_tokens": 3897755.0,
      "step": 1735
    },
    {
      "entropy": 5.956290817260742,
      "epoch": 0.1954287639692256,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999217553651676,
      "loss": 5.9546,
      "mean_token_accuracy": 0.15806302726268767,
      "num_tokens": 3909418.0,
      "step": 1740
    },
    {
      "entropy": 6.051894617080689,
      "epoch": 0.19599034087718312,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999206930550175,
      "loss": 6.1081,
      "mean_token_accuracy": 0.14818969666957854,
      "num_tokens": 3919992.0,
      "step": 1745
    },
    {
      "entropy": 6.063493871688843,
      "epoch": 0.19655191778514067,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999196235833054,
      "loss": 5.9663,
      "mean_token_accuracy": 0.16147716641426085,
      "num_tokens": 3930648.0,
      "step": 1750
    },
    {
      "entropy": 5.885579061508179,
      "epoch": 0.19711349469309822,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999185469500651,
      "loss": 5.9529,
      "mean_token_accuracy": 0.15516203343868257,
      "num_tokens": 3941883.0,
      "step": 1755
    },
    {
      "entropy": 6.019442367553711,
      "epoch": 0.19767507160105577,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499917463155331,
      "loss": 5.9987,
      "mean_token_accuracy": 0.15911805033683776,
      "num_tokens": 3952005.0,
      "step": 1760
    },
    {
      "entropy": 5.935513496398926,
      "epoch": 0.19823664850901332,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999163721991378,
      "loss": 5.9213,
      "mean_token_accuracy": 0.15570384711027146,
      "num_tokens": 3963720.0,
      "step": 1765
    },
    {
      "entropy": 5.93169755935669,
      "epoch": 0.19879822541697084,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049991527408152,
      "loss": 6.0081,
      "mean_token_accuracy": 0.15529269427061082,
      "num_tokens": 3974998.0,
      "step": 1770
    },
    {
      "entropy": 6.000465297698975,
      "epoch": 0.1993598023249284,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999141688025127,
      "loss": 6.048,
      "mean_token_accuracy": 0.14723881632089614,
      "num_tokens": 3986469.0,
      "step": 1775
    },
    {
      "entropy": 6.037735271453857,
      "epoch": 0.19992137923288594,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499913056362151,
      "loss": 5.8733,
      "mean_token_accuracy": 0.1559750333428383,
      "num_tokens": 3997553.0,
      "step": 1780
    },
    {
      "entropy": 5.848225831985474,
      "epoch": 0.2004829561408435,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999119367604703,
      "loss": 5.958,
      "mean_token_accuracy": 0.15345809012651443,
      "num_tokens": 4008613.0,
      "step": 1785
    },
    {
      "entropy": 6.148313570022583,
      "epoch": 0.20104453304880104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999108099975064,
      "loss": 6.1178,
      "mean_token_accuracy": 0.1494819074869156,
      "num_tokens": 4019860.0,
      "step": 1790
    },
    {
      "entropy": 5.971745538711548,
      "epoch": 0.20160610995675857,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499909676073295,
      "loss": 6.0125,
      "mean_token_accuracy": 0.15912038534879686,
      "num_tokens": 4030602.0,
      "step": 1795
    },
    {
      "entropy": 5.967540597915649,
      "epoch": 0.20216768686471612,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999085349878723,
      "loss": 5.8881,
      "mean_token_accuracy": 0.1551676794886589,
      "num_tokens": 4041878.0,
      "step": 1800
    },
    {
      "entropy": 5.902796602249145,
      "epoch": 0.20272926377267367,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999073867412746,
      "loss": 5.985,
      "mean_token_accuracy": 0.15298098027706147,
      "num_tokens": 4052799.0,
      "step": 1805
    },
    {
      "entropy": 6.052335357666015,
      "epoch": 0.20329084068063122,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999062313335385,
      "loss": 5.977,
      "mean_token_accuracy": 0.1513774089515209,
      "num_tokens": 4064339.0,
      "step": 1810
    },
    {
      "entropy": 5.881051683425904,
      "epoch": 0.20385241758858877,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999050687647008,
      "loss": 5.9321,
      "mean_token_accuracy": 0.15355904400348663,
      "num_tokens": 4076246.0,
      "step": 1815
    },
    {
      "entropy": 5.9952904224395756,
      "epoch": 0.2044139944965463,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999038990347984,
      "loss": 6.0102,
      "mean_token_accuracy": 0.15132154226303102,
      "num_tokens": 4088006.0,
      "step": 1820
    },
    {
      "entropy": 5.986813735961914,
      "epoch": 0.20497557140450384,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999027221438688,
      "loss": 6.0135,
      "mean_token_accuracy": 0.1533118411898613,
      "num_tokens": 4098122.0,
      "step": 1825
    },
    {
      "entropy": 6.179859924316406,
      "epoch": 0.2055371483124614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499901538091949,
      "loss": 6.2072,
      "mean_token_accuracy": 0.14116512686014177,
      "num_tokens": 4109958.0,
      "step": 1830
    },
    {
      "entropy": 6.044462394714356,
      "epoch": 0.20609872522041894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999003468790773,
      "loss": 6.0967,
      "mean_token_accuracy": 0.14568886607885362,
      "num_tokens": 4121262.0,
      "step": 1835
    },
    {
      "entropy": 6.016641092300415,
      "epoch": 0.2066603021283765,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998991485052911,
      "loss": 5.87,
      "mean_token_accuracy": 0.16044027954339982,
      "num_tokens": 4130958.0,
      "step": 1840
    },
    {
      "entropy": 5.966278457641602,
      "epoch": 0.20722187903633402,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499897942970629,
      "loss": 6.0778,
      "mean_token_accuracy": 0.1423511266708374,
      "num_tokens": 4143779.0,
      "step": 1845
    },
    {
      "entropy": 6.1042670726776125,
      "epoch": 0.20778345594429157,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499896730275129,
      "loss": 6.1007,
      "mean_token_accuracy": 0.14468785226345063,
      "num_tokens": 4155367.0,
      "step": 1850
    },
    {
      "entropy": 5.985637760162353,
      "epoch": 0.20834503285224912,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049989551041883,
      "loss": 5.9152,
      "mean_token_accuracy": 0.15388040691614152,
      "num_tokens": 4166657.0,
      "step": 1855
    },
    {
      "entropy": 5.961990070343018,
      "epoch": 0.20890660976020667,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998942834017707,
      "loss": 6.0298,
      "mean_token_accuracy": 0.15152619555592536,
      "num_tokens": 4177517.0,
      "step": 1860
    },
    {
      "entropy": 5.9690758228302006,
      "epoch": 0.2094681866681642,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998930492239901,
      "loss": 5.8913,
      "mean_token_accuracy": 0.15565723478794097,
      "num_tokens": 4188182.0,
      "step": 1865
    },
    {
      "entropy": 6.060051155090332,
      "epoch": 0.21002976357612174,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998918078855277,
      "loss": 6.0221,
      "mean_token_accuracy": 0.1551324889063835,
      "num_tokens": 4199415.0,
      "step": 1870
    },
    {
      "entropy": 5.884836721420288,
      "epoch": 0.2105913404840793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998905593864227,
      "loss": 5.967,
      "mean_token_accuracy": 0.15338498651981353,
      "num_tokens": 4210025.0,
      "step": 1875
    },
    {
      "entropy": 5.985698986053467,
      "epoch": 0.21115291739203684,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998893037267153,
      "loss": 5.8966,
      "mean_token_accuracy": 0.15888059735298157,
      "num_tokens": 4219920.0,
      "step": 1880
    },
    {
      "entropy": 5.979580545425415,
      "epoch": 0.2117144942999944,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998880409064452,
      "loss": 6.0678,
      "mean_token_accuracy": 0.1490791544318199,
      "num_tokens": 4232647.0,
      "step": 1885
    },
    {
      "entropy": 6.082404136657715,
      "epoch": 0.21227607120795192,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998867709256527,
      "loss": 5.9527,
      "mean_token_accuracy": 0.15514377057552337,
      "num_tokens": 4242637.0,
      "step": 1890
    },
    {
      "entropy": 5.876793336868286,
      "epoch": 0.21283764811590947,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499885493784378,
      "loss": 5.9242,
      "mean_token_accuracy": 0.15735216736793517,
      "num_tokens": 4253633.0,
      "step": 1895
    },
    {
      "entropy": 5.9101646900177,
      "epoch": 0.21339922502386702,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998842094826621,
      "loss": 5.8893,
      "mean_token_accuracy": 0.15901758521795273,
      "num_tokens": 4264472.0,
      "step": 1900
    },
    {
      "entropy": 6.013765573501587,
      "epoch": 0.21396080193182457,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998829180205458,
      "loss": 6.0058,
      "mean_token_accuracy": 0.15594724267721177,
      "num_tokens": 4275233.0,
      "step": 1905
    },
    {
      "entropy": 6.0052587509155275,
      "epoch": 0.21452237883978212,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998816193980701,
      "loss": 5.9606,
      "mean_token_accuracy": 0.15338825732469558,
      "num_tokens": 4287063.0,
      "step": 1910
    },
    {
      "entropy": 6.001710557937622,
      "epoch": 0.21508395574773964,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998803136152764,
      "loss": 5.9866,
      "mean_token_accuracy": 0.15025851130485535,
      "num_tokens": 4298761.0,
      "step": 1915
    },
    {
      "entropy": 5.735988998413086,
      "epoch": 0.2156455326556972,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998790006722063,
      "loss": 5.8469,
      "mean_token_accuracy": 0.16698734015226363,
      "num_tokens": 4309650.0,
      "step": 1920
    },
    {
      "entropy": 6.018524694442749,
      "epoch": 0.21620710956365474,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998776805689016,
      "loss": 5.9063,
      "mean_token_accuracy": 0.16533491015434265,
      "num_tokens": 4320403.0,
      "step": 1925
    },
    {
      "entropy": 5.84126296043396,
      "epoch": 0.2167686864716123,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998763533054045,
      "loss": 5.932,
      "mean_token_accuracy": 0.1564387321472168,
      "num_tokens": 4332049.0,
      "step": 1930
    },
    {
      "entropy": 5.947682619094849,
      "epoch": 0.21733026337956984,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998750188817568,
      "loss": 5.8358,
      "mean_token_accuracy": 0.16291145533323287,
      "num_tokens": 4342551.0,
      "step": 1935
    },
    {
      "entropy": 5.899789810180664,
      "epoch": 0.21789184028752737,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998736772980015,
      "loss": 5.9692,
      "mean_token_accuracy": 0.1594847857952118,
      "num_tokens": 4355156.0,
      "step": 1940
    },
    {
      "entropy": 5.908867359161377,
      "epoch": 0.21845341719548492,
      "grad_norm": 1.0,
      "learning_rate": 0.000499872328554181,
      "loss": 6.0218,
      "mean_token_accuracy": 0.15519974678754805,
      "num_tokens": 4367402.0,
      "step": 1945
    },
    {
      "entropy": 6.017113590240479,
      "epoch": 0.21901499410344247,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998709726503383,
      "loss": 5.8973,
      "mean_token_accuracy": 0.16066040098667145,
      "num_tokens": 4378687.0,
      "step": 1950
    },
    {
      "entropy": 5.973746299743652,
      "epoch": 0.21957657101140002,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998696095865168,
      "loss": 6.0371,
      "mean_token_accuracy": 0.1629161447286606,
      "num_tokens": 4388871.0,
      "step": 1955
    },
    {
      "entropy": 5.922299194335937,
      "epoch": 0.22013814791935757,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998682393627595,
      "loss": 5.8576,
      "mean_token_accuracy": 0.16151300221681594,
      "num_tokens": 4399440.0,
      "step": 1960
    },
    {
      "entropy": 5.825509119033813,
      "epoch": 0.2206997248273151,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998668619791104,
      "loss": 5.8555,
      "mean_token_accuracy": 0.16079302430152892,
      "num_tokens": 4409989.0,
      "step": 1965
    },
    {
      "entropy": 6.02225980758667,
      "epoch": 0.22126130173527264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998654774356131,
      "loss": 6.0152,
      "mean_token_accuracy": 0.15131131708621978,
      "num_tokens": 4422149.0,
      "step": 1970
    },
    {
      "entropy": 6.051222705841065,
      "epoch": 0.2218228786432302,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998640857323118,
      "loss": 6.0486,
      "mean_token_accuracy": 0.15000825822353364,
      "num_tokens": 4433157.0,
      "step": 1975
    },
    {
      "entropy": 6.00563907623291,
      "epoch": 0.22238445555118774,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998626868692509,
      "loss": 5.973,
      "mean_token_accuracy": 0.15266723185777664,
      "num_tokens": 4444595.0,
      "step": 1980
    },
    {
      "entropy": 6.015877962112427,
      "epoch": 0.2229460324591453,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998612808464748,
      "loss": 6.1061,
      "mean_token_accuracy": 0.14999096170067788,
      "num_tokens": 4456592.0,
      "step": 1985
    },
    {
      "entropy": 6.005351543426514,
      "epoch": 0.22350760936710282,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998598676640283,
      "loss": 5.8623,
      "mean_token_accuracy": 0.16657372564077377,
      "num_tokens": 4467624.0,
      "step": 1990
    },
    {
      "entropy": 5.870333099365235,
      "epoch": 0.22406918627506037,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998584473219563,
      "loss": 6.0578,
      "mean_token_accuracy": 0.1522249922156334,
      "num_tokens": 4479766.0,
      "step": 1995
    },
    {
      "entropy": 6.041551637649536,
      "epoch": 0.22463076318301792,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998570198203043,
      "loss": 5.927,
      "mean_token_accuracy": 0.1541971117258072,
      "num_tokens": 4490023.0,
      "step": 2000
    },
    {
      "entropy": 5.939703559875488,
      "epoch": 0.22519234009097547,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998555851591175,
      "loss": 5.9578,
      "mean_token_accuracy": 0.15466774851083756,
      "num_tokens": 4500651.0,
      "step": 2005
    },
    {
      "entropy": 6.019627428054809,
      "epoch": 0.225753916998933,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998541433384416,
      "loss": 6.0605,
      "mean_token_accuracy": 0.1519089847803116,
      "num_tokens": 4512934.0,
      "step": 2010
    },
    {
      "entropy": 5.9258240222930905,
      "epoch": 0.22631549390689054,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998526943583226,
      "loss": 5.9379,
      "mean_token_accuracy": 0.16064678132534027,
      "num_tokens": 4524251.0,
      "step": 2015
    },
    {
      "entropy": 5.996880960464478,
      "epoch": 0.2268770708148481,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998512382188066,
      "loss": 6.022,
      "mean_token_accuracy": 0.14892027527093887,
      "num_tokens": 4536025.0,
      "step": 2020
    },
    {
      "entropy": 6.004436492919922,
      "epoch": 0.22743864772280564,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998497749199401,
      "loss": 5.9701,
      "mean_token_accuracy": 0.158987794816494,
      "num_tokens": 4546717.0,
      "step": 2025
    },
    {
      "entropy": 5.910756921768188,
      "epoch": 0.2280002246307632,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998483044617694,
      "loss": 5.8651,
      "mean_token_accuracy": 0.1554678872227669,
      "num_tokens": 4557206.0,
      "step": 2030
    },
    {
      "entropy": 5.822503995895386,
      "epoch": 0.22856180153872072,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998468268443416,
      "loss": 5.8574,
      "mean_token_accuracy": 0.16205435246229172,
      "num_tokens": 4568123.0,
      "step": 2035
    },
    {
      "entropy": 5.961733341217041,
      "epoch": 0.22912337844667827,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998453420677034,
      "loss": 5.9819,
      "mean_token_accuracy": 0.15619362890720367,
      "num_tokens": 4579385.0,
      "step": 2040
    },
    {
      "entropy": 5.9808855056762695,
      "epoch": 0.22968495535463582,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998438501319025,
      "loss": 6.0248,
      "mean_token_accuracy": 0.15128423795104026,
      "num_tokens": 4590514.0,
      "step": 2045
    },
    {
      "entropy": 6.025481367111206,
      "epoch": 0.23024653226259337,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998423510369862,
      "loss": 5.9706,
      "mean_token_accuracy": 0.15017401427030563,
      "num_tokens": 4601624.0,
      "step": 2050
    },
    {
      "entropy": 5.990740871429443,
      "epoch": 0.23080810917055092,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998408447830022,
      "loss": 5.9518,
      "mean_token_accuracy": 0.16115492284297944,
      "num_tokens": 4612822.0,
      "step": 2055
    },
    {
      "entropy": 5.939420604705811,
      "epoch": 0.23136968607850844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998393313699985,
      "loss": 5.9466,
      "mean_token_accuracy": 0.16056140810251235,
      "num_tokens": 4623443.0,
      "step": 2060
    },
    {
      "entropy": 5.930271577835083,
      "epoch": 0.231931262986466,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998378107980233,
      "loss": 5.8579,
      "mean_token_accuracy": 0.16032032370567323,
      "num_tokens": 4633600.0,
      "step": 2065
    },
    {
      "entropy": 5.932555675506592,
      "epoch": 0.23249283989442354,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998362830671249,
      "loss": 6.0102,
      "mean_token_accuracy": 0.15435325503349304,
      "num_tokens": 4644217.0,
      "step": 2070
    },
    {
      "entropy": 5.97379846572876,
      "epoch": 0.2330544168023811,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998347481773522,
      "loss": 6.0249,
      "mean_token_accuracy": 0.1527538165450096,
      "num_tokens": 4654671.0,
      "step": 2075
    },
    {
      "entropy": 5.9626545906066895,
      "epoch": 0.23361599371033864,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998332061287538,
      "loss": 6.0209,
      "mean_token_accuracy": 0.1507769837975502,
      "num_tokens": 4665920.0,
      "step": 2080
    },
    {
      "entropy": 5.939901208877563,
      "epoch": 0.23417757061829617,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000499831656921379,
      "loss": 5.8997,
      "mean_token_accuracy": 0.1583072543144226,
      "num_tokens": 4675406.0,
      "step": 2085
    },
    {
      "entropy": 6.018610954284668,
      "epoch": 0.23473914752625372,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499830100555277,
      "loss": 5.9357,
      "mean_token_accuracy": 0.15790688395500183,
      "num_tokens": 4686870.0,
      "step": 2090
    },
    {
      "entropy": 5.953982305526734,
      "epoch": 0.23530072443421127,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998285370304974,
      "loss": 6.0124,
      "mean_token_accuracy": 0.15949206352233886,
      "num_tokens": 4698926.0,
      "step": 2095
    },
    {
      "entropy": 6.0622725009918215,
      "epoch": 0.23586230134216882,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049982696634709,
      "loss": 6.0882,
      "mean_token_accuracy": 0.14462199285626412,
      "num_tokens": 4709318.0,
      "step": 2100
    },
    {
      "entropy": 5.849839591979981,
      "epoch": 0.23642387825012637,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998253885051048,
      "loss": 5.7796,
      "mean_token_accuracy": 0.16810209602117537,
      "num_tokens": 4718819.0,
      "step": 2105
    },
    {
      "entropy": 5.884680938720703,
      "epoch": 0.2369854551580839,
      "grad_norm": 1.0,
      "learning_rate": 0.000499823803504592,
      "loss": 5.8314,
      "mean_token_accuracy": 0.16093335300683975,
      "num_tokens": 4729513.0,
      "step": 2110
    },
    {
      "entropy": 5.859940433502198,
      "epoch": 0.23754703206604144,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998222113456022,
      "loss": 5.9287,
      "mean_token_accuracy": 0.15834257006645203,
      "num_tokens": 4739875.0,
      "step": 2115
    },
    {
      "entropy": 6.060333490371704,
      "epoch": 0.238108608973999,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998206120281858,
      "loss": 6.0609,
      "mean_token_accuracy": 0.14767249599099158,
      "num_tokens": 4752933.0,
      "step": 2120
    },
    {
      "entropy": 5.938970279693604,
      "epoch": 0.23867018588195654,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499819005552394,
      "loss": 5.8526,
      "mean_token_accuracy": 0.16048028618097304,
      "num_tokens": 4762969.0,
      "step": 2125
    },
    {
      "entropy": 5.958646631240844,
      "epoch": 0.23923176278991407,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998173919182779,
      "loss": 5.9439,
      "mean_token_accuracy": 0.15564859807491302,
      "num_tokens": 4773495.0,
      "step": 2130
    },
    {
      "entropy": 5.948372316360474,
      "epoch": 0.23979333969787162,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499815771125889,
      "loss": 5.9773,
      "mean_token_accuracy": 0.16150738894939423,
      "num_tokens": 4785335.0,
      "step": 2135
    },
    {
      "entropy": 6.081469774246216,
      "epoch": 0.24035491660582917,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998141431752785,
      "loss": 6.1175,
      "mean_token_accuracy": 0.14878264293074608,
      "num_tokens": 4798068.0,
      "step": 2140
    },
    {
      "entropy": 5.92440824508667,
      "epoch": 0.24091649351378672,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998125080664987,
      "loss": 5.9373,
      "mean_token_accuracy": 0.1596187397837639,
      "num_tokens": 4809338.0,
      "step": 2145
    },
    {
      "entropy": 5.900418853759765,
      "epoch": 0.24147807042174427,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998108657996012,
      "loss": 5.9278,
      "mean_token_accuracy": 0.1604372188448906,
      "num_tokens": 4819636.0,
      "step": 2150
    },
    {
      "entropy": 5.9336934089660645,
      "epoch": 0.2420396473297018,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998092163746387,
      "loss": 5.9066,
      "mean_token_accuracy": 0.16119810789823533,
      "num_tokens": 4830975.0,
      "step": 2155
    },
    {
      "entropy": 5.893611574172974,
      "epoch": 0.24260122423765934,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998075597916636,
      "loss": 5.9428,
      "mean_token_accuracy": 0.15798580050468444,
      "num_tokens": 4841951.0,
      "step": 2160
    },
    {
      "entropy": 6.011291074752807,
      "epoch": 0.2431628011456169,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998058960507286,
      "loss": 5.9185,
      "mean_token_accuracy": 0.15732098147273063,
      "num_tokens": 4852919.0,
      "step": 2165
    },
    {
      "entropy": 5.955287456512451,
      "epoch": 0.24372437805357444,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998042251518866,
      "loss": 5.9264,
      "mean_token_accuracy": 0.1588020294904709,
      "num_tokens": 4863433.0,
      "step": 2170
    },
    {
      "entropy": 6.019167184829712,
      "epoch": 0.244285954961532,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998025470951908,
      "loss": 6.0132,
      "mean_token_accuracy": 0.15414848923683167,
      "num_tokens": 4875685.0,
      "step": 2175
    },
    {
      "entropy": 5.848620128631592,
      "epoch": 0.24484753186948952,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998008618806949,
      "loss": 5.9081,
      "mean_token_accuracy": 0.15716440230607986,
      "num_tokens": 4886347.0,
      "step": 2180
    },
    {
      "entropy": 5.924430418014526,
      "epoch": 0.24540910877744707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997991695084523,
      "loss": 5.8998,
      "mean_token_accuracy": 0.16334613561630248,
      "num_tokens": 4896898.0,
      "step": 2185
    },
    {
      "entropy": 5.999474620819091,
      "epoch": 0.24597068568540462,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997974699785169,
      "loss": 6.0128,
      "mean_token_accuracy": 0.15362882167100905,
      "num_tokens": 4908892.0,
      "step": 2190
    },
    {
      "entropy": 5.907583093643188,
      "epoch": 0.24653226259336217,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997957632909429,
      "loss": 5.9471,
      "mean_token_accuracy": 0.15109152495861053,
      "num_tokens": 4920855.0,
      "step": 2195
    },
    {
      "entropy": 6.013603639602661,
      "epoch": 0.24709383950131972,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997940494457846,
      "loss": 5.9325,
      "mean_token_accuracy": 0.16000126004219056,
      "num_tokens": 4932857.0,
      "step": 2200
    },
    {
      "entropy": 5.8864202976226805,
      "epoch": 0.24765541640927724,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997923284430967,
      "loss": 5.9747,
      "mean_token_accuracy": 0.15626695305109023,
      "num_tokens": 4944063.0,
      "step": 2205
    },
    {
      "entropy": 5.871590852737427,
      "epoch": 0.2482169933172348,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997906002829338,
      "loss": 5.8432,
      "mean_token_accuracy": 0.16545321345329284,
      "num_tokens": 4955126.0,
      "step": 2210
    },
    {
      "entropy": 6.042226839065552,
      "epoch": 0.24877857022519234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499788864965351,
      "loss": 6.0353,
      "mean_token_accuracy": 0.15824319273233414,
      "num_tokens": 4965738.0,
      "step": 2215
    },
    {
      "entropy": 6.032025718688965,
      "epoch": 0.2493401471331499,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997871224904036,
      "loss": 6.0553,
      "mean_token_accuracy": 0.15088909566402436,
      "num_tokens": 4976963.0,
      "step": 2220
    },
    {
      "entropy": 5.968750953674316,
      "epoch": 0.24990172404110744,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997853728581468,
      "loss": 5.9356,
      "mean_token_accuracy": 0.15454284846782684,
      "num_tokens": 4987577.0,
      "step": 2225
    },
    {
      "entropy": 5.988882637023925,
      "epoch": 0.250463300949065,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997836160686368,
      "loss": 6.0657,
      "mean_token_accuracy": 0.15284020453691483,
      "num_tokens": 5000234.0,
      "step": 2230
    },
    {
      "entropy": 5.972144842147827,
      "epoch": 0.25102487785702254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997818521219291,
      "loss": 5.8895,
      "mean_token_accuracy": 0.15354024320840837,
      "num_tokens": 5011652.0,
      "step": 2235
    },
    {
      "entropy": 5.973802328109741,
      "epoch": 0.25158645476498004,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997800810180803,
      "loss": 5.9329,
      "mean_token_accuracy": 0.1593118280172348,
      "num_tokens": 5023603.0,
      "step": 2240
    },
    {
      "entropy": 5.892467927932739,
      "epoch": 0.2521480316729376,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997783027571464,
      "loss": 5.9091,
      "mean_token_accuracy": 0.15956079661846162,
      "num_tokens": 5034427.0,
      "step": 2245
    },
    {
      "entropy": 5.859362268447876,
      "epoch": 0.25270960858089514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997765173391841,
      "loss": 5.8988,
      "mean_token_accuracy": 0.15956594049930573,
      "num_tokens": 5045738.0,
      "step": 2250
    },
    {
      "entropy": 5.869083404541016,
      "epoch": 0.2532711854888527,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997747247642505,
      "loss": 5.8353,
      "mean_token_accuracy": 0.15759642720222472,
      "num_tokens": 5057021.0,
      "step": 2255
    },
    {
      "entropy": 5.92234172821045,
      "epoch": 0.25383276239681024,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997729250324023,
      "loss": 5.9244,
      "mean_token_accuracy": 0.15921089351177214,
      "num_tokens": 5067676.0,
      "step": 2260
    },
    {
      "entropy": 5.886710166931152,
      "epoch": 0.2543943393047678,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997711181436971,
      "loss": 5.8615,
      "mean_token_accuracy": 0.15837088972330093,
      "num_tokens": 5079422.0,
      "step": 2265
    },
    {
      "entropy": 5.822842216491699,
      "epoch": 0.25495591621272534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997693040981922,
      "loss": 5.8851,
      "mean_token_accuracy": 0.16085467636585235,
      "num_tokens": 5091242.0,
      "step": 2270
    },
    {
      "entropy": 6.006762552261352,
      "epoch": 0.2555174931206829,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997674828959456,
      "loss": 5.8909,
      "mean_token_accuracy": 0.15583527684211732,
      "num_tokens": 5102251.0,
      "step": 2275
    },
    {
      "entropy": 5.857793092727661,
      "epoch": 0.25607907002864044,
      "grad_norm": 1.5,
      "learning_rate": 0.0004997656545370151,
      "loss": 5.9104,
      "mean_token_accuracy": 0.1610156461596489,
      "num_tokens": 5113276.0,
      "step": 2280
    },
    {
      "entropy": 5.934245014190674,
      "epoch": 0.256640646936598,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499763819021459,
      "loss": 5.9191,
      "mean_token_accuracy": 0.15731906294822692,
      "num_tokens": 5124472.0,
      "step": 2285
    },
    {
      "entropy": 5.863279104232788,
      "epoch": 0.2572022238445555,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997619763493357,
      "loss": 5.8698,
      "mean_token_accuracy": 0.1613835647702217,
      "num_tokens": 5134896.0,
      "step": 2290
    },
    {
      "entropy": 5.962464809417725,
      "epoch": 0.25776380075251304,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499760126520704,
      "loss": 5.95,
      "mean_token_accuracy": 0.1634097322821617,
      "num_tokens": 5145115.0,
      "step": 2295
    },
    {
      "entropy": 5.927261829376221,
      "epoch": 0.2583253776604706,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997582695356225,
      "loss": 5.9289,
      "mean_token_accuracy": 0.15928202718496323,
      "num_tokens": 5155620.0,
      "step": 2300
    },
    {
      "entropy": 6.070884418487549,
      "epoch": 0.25888695456842814,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997564053941507,
      "loss": 6.0126,
      "mean_token_accuracy": 0.15451653599739074,
      "num_tokens": 5166260.0,
      "step": 2305
    },
    {
      "entropy": 5.954886245727539,
      "epoch": 0.2594485314763857,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997545340963478,
      "loss": 5.8733,
      "mean_token_accuracy": 0.15934662520885468,
      "num_tokens": 5177143.0,
      "step": 2310
    },
    {
      "entropy": 5.921825218200683,
      "epoch": 0.26001010838434324,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997526556422733,
      "loss": 6.0233,
      "mean_token_accuracy": 0.15293444991111754,
      "num_tokens": 5187795.0,
      "step": 2315
    },
    {
      "entropy": 5.922110557556152,
      "epoch": 0.2605716852923008,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997507700319872,
      "loss": 5.8811,
      "mean_token_accuracy": 0.16001545637845993,
      "num_tokens": 5199365.0,
      "step": 2320
    },
    {
      "entropy": 5.971580982208252,
      "epoch": 0.26113326220025834,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004997488772655492,
      "loss": 6.0976,
      "mean_token_accuracy": 0.15617025047540664,
      "num_tokens": 5212264.0,
      "step": 2325
    },
    {
      "entropy": 6.042904090881348,
      "epoch": 0.2616948391082159,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997469773430199,
      "loss": 5.9253,
      "mean_token_accuracy": 0.15450814068317414,
      "num_tokens": 5223544.0,
      "step": 2330
    },
    {
      "entropy": 5.868034029006958,
      "epoch": 0.2622564160161734,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997450702644596,
      "loss": 5.875,
      "mean_token_accuracy": 0.1604365214705467,
      "num_tokens": 5234713.0,
      "step": 2335
    },
    {
      "entropy": 5.9823534965515135,
      "epoch": 0.26281799292413094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997431560299292,
      "loss": 5.9591,
      "mean_token_accuracy": 0.15298095047473909,
      "num_tokens": 5246103.0,
      "step": 2340
    },
    {
      "entropy": 5.814671277999878,
      "epoch": 0.2633795698320885,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997412346394894,
      "loss": 5.7917,
      "mean_token_accuracy": 0.16624176353216172,
      "num_tokens": 5257372.0,
      "step": 2345
    },
    {
      "entropy": 5.792899942398071,
      "epoch": 0.26394114674004604,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997393060932017,
      "loss": 5.8609,
      "mean_token_accuracy": 0.16722605973482133,
      "num_tokens": 5268677.0,
      "step": 2350
    },
    {
      "entropy": 5.997194576263428,
      "epoch": 0.2645027236480036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997373703911273,
      "loss": 5.9247,
      "mean_token_accuracy": 0.16152105778455733,
      "num_tokens": 5279056.0,
      "step": 2355
    },
    {
      "entropy": 5.786316013336181,
      "epoch": 0.26506430055596114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997354275333277,
      "loss": 5.8069,
      "mean_token_accuracy": 0.16789699494838714,
      "num_tokens": 5289310.0,
      "step": 2360
    },
    {
      "entropy": 5.953582143783569,
      "epoch": 0.2656258774639187,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499733477519865,
      "loss": 5.8956,
      "mean_token_accuracy": 0.16771894991397857,
      "num_tokens": 5300619.0,
      "step": 2365
    },
    {
      "entropy": 6.053738594055176,
      "epoch": 0.26618745437187624,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997315203508013,
      "loss": 6.0608,
      "mean_token_accuracy": 0.1562006339430809,
      "num_tokens": 5311554.0,
      "step": 2370
    },
    {
      "entropy": 5.966435718536377,
      "epoch": 0.2667490312798338,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997295560261988,
      "loss": 5.883,
      "mean_token_accuracy": 0.1612299293279648,
      "num_tokens": 5323592.0,
      "step": 2375
    },
    {
      "entropy": 5.942288255691528,
      "epoch": 0.26731060818779134,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049972758454612,
      "loss": 5.9767,
      "mean_token_accuracy": 0.15383181422948838,
      "num_tokens": 5334978.0,
      "step": 2380
    },
    {
      "entropy": 5.898320913314819,
      "epoch": 0.26787218509574884,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997256059106279,
      "loss": 5.8667,
      "mean_token_accuracy": 0.15778945684432982,
      "num_tokens": 5345724.0,
      "step": 2385
    },
    {
      "entropy": 5.94132432937622,
      "epoch": 0.2684337620037064,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997236201197853,
      "loss": 5.9989,
      "mean_token_accuracy": 0.15381949245929719,
      "num_tokens": 5357141.0,
      "step": 2390
    },
    {
      "entropy": 5.9436756610870365,
      "epoch": 0.26899533891166394,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997216271736554,
      "loss": 5.995,
      "mean_token_accuracy": 0.15819256901741027,
      "num_tokens": 5367889.0,
      "step": 2395
    },
    {
      "entropy": 5.786690282821655,
      "epoch": 0.2695569158196215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997196270723018,
      "loss": 5.7261,
      "mean_token_accuracy": 0.16399870365858077,
      "num_tokens": 5379329.0,
      "step": 2400
    },
    {
      "entropy": 5.895531129837036,
      "epoch": 0.27011849272757904,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997176198157881,
      "loss": 5.8852,
      "mean_token_accuracy": 0.1617034539580345,
      "num_tokens": 5389525.0,
      "step": 2405
    },
    {
      "entropy": 5.895331811904907,
      "epoch": 0.2706800696355366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997156054041781,
      "loss": 5.7958,
      "mean_token_accuracy": 0.16883354038000106,
      "num_tokens": 5401061.0,
      "step": 2410
    },
    {
      "entropy": 5.873462343215943,
      "epoch": 0.27124164654349414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997135838375362,
      "loss": 5.8949,
      "mean_token_accuracy": 0.1616640195250511,
      "num_tokens": 5411791.0,
      "step": 2415
    },
    {
      "entropy": 5.908325338363648,
      "epoch": 0.2718032234514517,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997115551159267,
      "loss": 5.8467,
      "mean_token_accuracy": 0.16689947098493577,
      "num_tokens": 5422790.0,
      "step": 2420
    },
    {
      "entropy": 5.738713026046753,
      "epoch": 0.27236480035940924,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499709519239414,
      "loss": 5.7951,
      "mean_token_accuracy": 0.1616228774189949,
      "num_tokens": 5434170.0,
      "step": 2425
    },
    {
      "entropy": 5.972346496582031,
      "epoch": 0.2729263772673668,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004997074762080632,
      "loss": 5.922,
      "mean_token_accuracy": 0.15746588557958602,
      "num_tokens": 5445948.0,
      "step": 2430
    },
    {
      "entropy": 5.907391738891602,
      "epoch": 0.2734879541753243,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499705426021939,
      "loss": 5.8828,
      "mean_token_accuracy": 0.15904785543680192,
      "num_tokens": 5457786.0,
      "step": 2435
    },
    {
      "entropy": 5.784610605239868,
      "epoch": 0.27404953108328184,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499703368681107,
      "loss": 5.8589,
      "mean_token_accuracy": 0.16070218682289122,
      "num_tokens": 5469434.0,
      "step": 2440
    },
    {
      "entropy": 5.8818888664245605,
      "epoch": 0.2746111079912394,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997013041856326,
      "loss": 5.8549,
      "mean_token_accuracy": 0.1531570464372635,
      "num_tokens": 5480393.0,
      "step": 2445
    },
    {
      "entropy": 5.844133996963501,
      "epoch": 0.27517268489919694,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996992325355815,
      "loss": 5.8264,
      "mean_token_accuracy": 0.16376967430114747,
      "num_tokens": 5490452.0,
      "step": 2450
    },
    {
      "entropy": 5.843231868743897,
      "epoch": 0.2757342618071545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996971537310198,
      "loss": 5.7996,
      "mean_token_accuracy": 0.16604796648025513,
      "num_tokens": 5501404.0,
      "step": 2455
    },
    {
      "entropy": 5.97615795135498,
      "epoch": 0.27629583871511204,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996950677720133,
      "loss": 5.9411,
      "mean_token_accuracy": 0.15309768170118332,
      "num_tokens": 5512938.0,
      "step": 2460
    },
    {
      "entropy": 5.879659843444824,
      "epoch": 0.2768574156230696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499692974658629,
      "loss": 5.8641,
      "mean_token_accuracy": 0.15448336452245712,
      "num_tokens": 5524009.0,
      "step": 2465
    },
    {
      "entropy": 5.961763477325439,
      "epoch": 0.27741899253102714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996908743909331,
      "loss": 5.9491,
      "mean_token_accuracy": 0.15658549070358277,
      "num_tokens": 5535930.0,
      "step": 2470
    },
    {
      "entropy": 5.943891477584839,
      "epoch": 0.2779805694389847,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996887669689926,
      "loss": 5.9378,
      "mean_token_accuracy": 0.1560988336801529,
      "num_tokens": 5547295.0,
      "step": 2475
    },
    {
      "entropy": 5.934115314483643,
      "epoch": 0.2785421463469422,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996866523928747,
      "loss": 5.8393,
      "mean_token_accuracy": 0.1640307277441025,
      "num_tokens": 5558447.0,
      "step": 2480
    },
    {
      "entropy": 5.91972165107727,
      "epoch": 0.27910372325489974,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996845306626467,
      "loss": 5.9369,
      "mean_token_accuracy": 0.16300042271614074,
      "num_tokens": 5569822.0,
      "step": 2485
    },
    {
      "entropy": 5.9508904933929445,
      "epoch": 0.2796653001628573,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996824017783759,
      "loss": 5.9902,
      "mean_token_accuracy": 0.15508754998445512,
      "num_tokens": 5582284.0,
      "step": 2490
    },
    {
      "entropy": 5.9683637619018555,
      "epoch": 0.28022687707081484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996802657401304,
      "loss": 5.879,
      "mean_token_accuracy": 0.1611318975687027,
      "num_tokens": 5593765.0,
      "step": 2495
    },
    {
      "entropy": 5.801318073272705,
      "epoch": 0.2807884539787724,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996781225479781,
      "loss": 5.8487,
      "mean_token_accuracy": 0.15998225808143615,
      "num_tokens": 5605080.0,
      "step": 2500
    },
    {
      "entropy": 6.015531349182129,
      "epoch": 0.28135003088672994,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996759722019872,
      "loss": 6.011,
      "mean_token_accuracy": 0.15439892411231995,
      "num_tokens": 5616739.0,
      "step": 2505
    },
    {
      "entropy": 5.904989242553711,
      "epoch": 0.2819116077946875,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996738147022262,
      "loss": 5.8735,
      "mean_token_accuracy": 0.1576665908098221,
      "num_tokens": 5628779.0,
      "step": 2510
    },
    {
      "entropy": 5.889083909988403,
      "epoch": 0.28247318470264504,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996716500487638,
      "loss": 5.909,
      "mean_token_accuracy": 0.15748686641454696,
      "num_tokens": 5639168.0,
      "step": 2515
    },
    {
      "entropy": 5.893275928497315,
      "epoch": 0.2830347616106026,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499669478241669,
      "loss": 5.885,
      "mean_token_accuracy": 0.15599167197942734,
      "num_tokens": 5650739.0,
      "step": 2520
    },
    {
      "entropy": 5.933381366729736,
      "epoch": 0.28359633851856014,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996672992810108,
      "loss": 5.9054,
      "mean_token_accuracy": 0.15816623866558074,
      "num_tokens": 5662630.0,
      "step": 2525
    },
    {
      "entropy": 5.921469497680664,
      "epoch": 0.28415791542651764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996651131668587,
      "loss": 5.9083,
      "mean_token_accuracy": 0.1671989306807518,
      "num_tokens": 5673473.0,
      "step": 2530
    },
    {
      "entropy": 5.915191793441773,
      "epoch": 0.2847194923344752,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996629198992823,
      "loss": 6.0544,
      "mean_token_accuracy": 0.15136269852519035,
      "num_tokens": 5686577.0,
      "step": 2535
    },
    {
      "entropy": 5.963024854660034,
      "epoch": 0.28528106924243274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996607194783513,
      "loss": 5.8035,
      "mean_token_accuracy": 0.15977411419153215,
      "num_tokens": 5697570.0,
      "step": 2540
    },
    {
      "entropy": 5.849263906478882,
      "epoch": 0.2858426461503903,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996585119041359,
      "loss": 5.8717,
      "mean_token_accuracy": 0.15102790147066117,
      "num_tokens": 5708601.0,
      "step": 2545
    },
    {
      "entropy": 5.817156410217285,
      "epoch": 0.28640422305834784,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996562971767063,
      "loss": 5.8333,
      "mean_token_accuracy": 0.1580057695508003,
      "num_tokens": 5719619.0,
      "step": 2550
    },
    {
      "entropy": 5.881134510040283,
      "epoch": 0.2869657999663054,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499654075296133,
      "loss": 5.854,
      "mean_token_accuracy": 0.16304896473884584,
      "num_tokens": 5730061.0,
      "step": 2555
    },
    {
      "entropy": 5.887245082855225,
      "epoch": 0.28752737687426294,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499651846262487,
      "loss": 5.8781,
      "mean_token_accuracy": 0.16307636797428132,
      "num_tokens": 5741579.0,
      "step": 2560
    },
    {
      "entropy": 5.929854440689087,
      "epoch": 0.2880889537822205,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996496100758389,
      "loss": 5.8899,
      "mean_token_accuracy": 0.16015600562095642,
      "num_tokens": 5754001.0,
      "step": 2565
    },
    {
      "entropy": 5.956954383850098,
      "epoch": 0.28865053069017804,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996473667362604,
      "loss": 5.9762,
      "mean_token_accuracy": 0.15367593243718147,
      "num_tokens": 5766224.0,
      "step": 2570
    },
    {
      "entropy": 5.9173064708709715,
      "epoch": 0.28921210759813554,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996451162438224,
      "loss": 5.9514,
      "mean_token_accuracy": 0.1529170662164688,
      "num_tokens": 5777475.0,
      "step": 2575
    },
    {
      "entropy": 5.9922263622283936,
      "epoch": 0.2897736845060931,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996428585985967,
      "loss": 5.9243,
      "mean_token_accuracy": 0.15300409942865373,
      "num_tokens": 5788359.0,
      "step": 2580
    },
    {
      "entropy": 5.833764743804932,
      "epoch": 0.29033526141405064,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996405938006553,
      "loss": 5.88,
      "mean_token_accuracy": 0.1589822694659233,
      "num_tokens": 5799116.0,
      "step": 2585
    },
    {
      "entropy": 5.849058294296265,
      "epoch": 0.2908968383220082,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996383218500703,
      "loss": 5.8163,
      "mean_token_accuracy": 0.16485902518033982,
      "num_tokens": 5811221.0,
      "step": 2590
    },
    {
      "entropy": 5.903125619888305,
      "epoch": 0.29145841522996574,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499636042746914,
      "loss": 5.8986,
      "mean_token_accuracy": 0.1597510278224945,
      "num_tokens": 5822099.0,
      "step": 2595
    },
    {
      "entropy": 5.908592653274536,
      "epoch": 0.2920199921379233,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499633756491259,
      "loss": 5.8603,
      "mean_token_accuracy": 0.154727041721344,
      "num_tokens": 5832836.0,
      "step": 2600
    },
    {
      "entropy": 5.813444709777832,
      "epoch": 0.29258156904588084,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996314630831781,
      "loss": 5.8066,
      "mean_token_accuracy": 0.15686748921871185,
      "num_tokens": 5842892.0,
      "step": 2605
    },
    {
      "entropy": 5.8215662956237795,
      "epoch": 0.2931431459538384,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996291625227443,
      "loss": 5.7707,
      "mean_token_accuracy": 0.17027000188827515,
      "num_tokens": 5852909.0,
      "step": 2610
    },
    {
      "entropy": 5.975628280639649,
      "epoch": 0.29370472286179594,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996268548100307,
      "loss": 5.9509,
      "mean_token_accuracy": 0.15499749705195426,
      "num_tokens": 5864676.0,
      "step": 2615
    },
    {
      "entropy": 5.67653169631958,
      "epoch": 0.2942662997697535,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499624539945111,
      "loss": 5.6832,
      "mean_token_accuracy": 0.17367247641086578,
      "num_tokens": 5875258.0,
      "step": 2620
    },
    {
      "entropy": 5.918656921386718,
      "epoch": 0.294827876677711,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996222179280589,
      "loss": 5.8505,
      "mean_token_accuracy": 0.16116076856851577,
      "num_tokens": 5885611.0,
      "step": 2625
    },
    {
      "entropy": 5.846941518783569,
      "epoch": 0.29538945358566854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996198887589481,
      "loss": 5.8257,
      "mean_token_accuracy": 0.15930989608168603,
      "num_tokens": 5896726.0,
      "step": 2630
    },
    {
      "entropy": 5.917367076873779,
      "epoch": 0.2959510304936261,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996175524378531,
      "loss": 5.851,
      "mean_token_accuracy": 0.1528210312128067,
      "num_tokens": 5907860.0,
      "step": 2635
    },
    {
      "entropy": 5.806666278839112,
      "epoch": 0.29651260740158364,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499615208964848,
      "loss": 5.798,
      "mean_token_accuracy": 0.1638117402791977,
      "num_tokens": 5918676.0,
      "step": 2640
    },
    {
      "entropy": 5.9251220703125,
      "epoch": 0.2970741843095412,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996128583400077,
      "loss": 5.8618,
      "mean_token_accuracy": 0.16403974294662477,
      "num_tokens": 5928646.0,
      "step": 2645
    },
    {
      "entropy": 5.816641330718994,
      "epoch": 0.29763576121749874,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004996105005634066,
      "loss": 5.8735,
      "mean_token_accuracy": 0.1569816768169403,
      "num_tokens": 5938990.0,
      "step": 2650
    },
    {
      "entropy": 5.8617603302001955,
      "epoch": 0.2981973381254563,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996081356351202,
      "loss": 5.8194,
      "mean_token_accuracy": 0.16183129400014878,
      "num_tokens": 5948811.0,
      "step": 2655
    },
    {
      "entropy": 5.812265920639038,
      "epoch": 0.29875891503341384,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996057635552237,
      "loss": 5.8406,
      "mean_token_accuracy": 0.16514092236757277,
      "num_tokens": 5959227.0,
      "step": 2660
    },
    {
      "entropy": 5.8553258895874025,
      "epoch": 0.2993204919413714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996033843237925,
      "loss": 5.7535,
      "mean_token_accuracy": 0.16563905775547028,
      "num_tokens": 5970546.0,
      "step": 2665
    },
    {
      "entropy": 5.982068204879761,
      "epoch": 0.29988206884932894,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004996009979409025,
      "loss": 6.0502,
      "mean_token_accuracy": 0.147099506855011,
      "num_tokens": 5983936.0,
      "step": 2670
    },
    {
      "entropy": 5.802636957168579,
      "epoch": 0.30044364575728644,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995986044066295,
      "loss": 5.8494,
      "mean_token_accuracy": 0.15568828135728835,
      "num_tokens": 5994963.0,
      "step": 2675
    },
    {
      "entropy": 5.762459421157837,
      "epoch": 0.301005222665244,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049959620372105,
      "loss": 5.8079,
      "mean_token_accuracy": 0.16270437240600585,
      "num_tokens": 6005836.0,
      "step": 2680
    },
    {
      "entropy": 5.895258378982544,
      "epoch": 0.30156679957320154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995937958842401,
      "loss": 5.8057,
      "mean_token_accuracy": 0.16682981103658676,
      "num_tokens": 6017408.0,
      "step": 2685
    },
    {
      "entropy": 5.811723136901856,
      "epoch": 0.3021283764811591,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995913808962767,
      "loss": 5.8192,
      "mean_token_accuracy": 0.15878787338733674,
      "num_tokens": 6029281.0,
      "step": 2690
    },
    {
      "entropy": 5.850067806243897,
      "epoch": 0.30268995338911664,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995889587572366,
      "loss": 5.8088,
      "mean_token_accuracy": 0.16636761128902436,
      "num_tokens": 6039354.0,
      "step": 2695
    },
    {
      "entropy": 5.8515503883361815,
      "epoch": 0.3032515302970742,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499586529467197,
      "loss": 5.785,
      "mean_token_accuracy": 0.15962323993444444,
      "num_tokens": 6050053.0,
      "step": 2700
    },
    {
      "entropy": 5.90614161491394,
      "epoch": 0.30381310720503174,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995840930262351,
      "loss": 5.8451,
      "mean_token_accuracy": 0.15831548869609832,
      "num_tokens": 6060221.0,
      "step": 2705
    },
    {
      "entropy": 5.824668931961059,
      "epoch": 0.3043746841129893,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995816494344287,
      "loss": 5.859,
      "mean_token_accuracy": 0.1601285994052887,
      "num_tokens": 6070963.0,
      "step": 2710
    },
    {
      "entropy": 5.989436864852905,
      "epoch": 0.30493626102094684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995791986918554,
      "loss": 5.9386,
      "mean_token_accuracy": 0.15705745071172714,
      "num_tokens": 6082028.0,
      "step": 2715
    },
    {
      "entropy": 5.949339580535889,
      "epoch": 0.30549783792890434,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995767407985933,
      "loss": 5.8736,
      "mean_token_accuracy": 0.16265355795621872,
      "num_tokens": 6092778.0,
      "step": 2720
    },
    {
      "entropy": 5.677318954467774,
      "epoch": 0.3060594148368619,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995742757547207,
      "loss": 5.6421,
      "mean_token_accuracy": 0.16859253942966462,
      "num_tokens": 6103093.0,
      "step": 2725
    },
    {
      "entropy": 5.772259378433228,
      "epoch": 0.30662099174481944,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499571803560316,
      "loss": 5.7347,
      "mean_token_accuracy": 0.16355275958776475,
      "num_tokens": 6113516.0,
      "step": 2730
    },
    {
      "entropy": 5.810375738143921,
      "epoch": 0.307182568652777,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995693242154581,
      "loss": 5.8036,
      "mean_token_accuracy": 0.15890752524137497,
      "num_tokens": 6125723.0,
      "step": 2735
    },
    {
      "entropy": 5.847309589385986,
      "epoch": 0.30774414556073454,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004995668377202258,
      "loss": 5.8136,
      "mean_token_accuracy": 0.16112632900476456,
      "num_tokens": 6137091.0,
      "step": 2740
    },
    {
      "entropy": 5.77405891418457,
      "epoch": 0.3083057224686921,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995643440746982,
      "loss": 5.7528,
      "mean_token_accuracy": 0.16945552676916123,
      "num_tokens": 6147941.0,
      "step": 2745
    },
    {
      "entropy": 5.732852792739868,
      "epoch": 0.30886729937664964,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995618432789548,
      "loss": 5.7549,
      "mean_token_accuracy": 0.16648126393556595,
      "num_tokens": 6158493.0,
      "step": 2750
    },
    {
      "entropy": 5.8287242412567135,
      "epoch": 0.3094288762846072,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995593353330753,
      "loss": 5.8915,
      "mean_token_accuracy": 0.15722588896751405,
      "num_tokens": 6170759.0,
      "step": 2755
    },
    {
      "entropy": 5.898643875122071,
      "epoch": 0.30999045319256474,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995568202371393,
      "loss": 5.724,
      "mean_token_accuracy": 0.16514806002378463,
      "num_tokens": 6182488.0,
      "step": 2760
    },
    {
      "entropy": 5.841040992736817,
      "epoch": 0.3105520301005223,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995542979912272,
      "loss": 5.9165,
      "mean_token_accuracy": 0.15282527059316636,
      "num_tokens": 6193849.0,
      "step": 2765
    },
    {
      "entropy": 5.800279855728149,
      "epoch": 0.3111136070084798,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995517685954192,
      "loss": 5.7598,
      "mean_token_accuracy": 0.16608615219593048,
      "num_tokens": 6204132.0,
      "step": 2770
    },
    {
      "entropy": 5.946976375579834,
      "epoch": 0.31167518391643734,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995492320497958,
      "loss": 5.9018,
      "mean_token_accuracy": 0.1554347574710846,
      "num_tokens": 6215547.0,
      "step": 2775
    },
    {
      "entropy": 5.905535459518433,
      "epoch": 0.3122367608243949,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995466883544376,
      "loss": 5.9293,
      "mean_token_accuracy": 0.15042147785425186,
      "num_tokens": 6227150.0,
      "step": 2780
    },
    {
      "entropy": 5.8855243682861325,
      "epoch": 0.31279833773235244,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995441375094258,
      "loss": 5.9248,
      "mean_token_accuracy": 0.15840522199869156,
      "num_tokens": 6237387.0,
      "step": 2785
    },
    {
      "entropy": 5.872175312042236,
      "epoch": 0.31335991464031,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995415795148417,
      "loss": 5.812,
      "mean_token_accuracy": 0.16368252038955688,
      "num_tokens": 6248623.0,
      "step": 2790
    },
    {
      "entropy": 5.76340684890747,
      "epoch": 0.31392149154826754,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995390143707665,
      "loss": 5.7152,
      "mean_token_accuracy": 0.17597836554050444,
      "num_tokens": 6258596.0,
      "step": 2795
    },
    {
      "entropy": 5.822847366333008,
      "epoch": 0.3144830684562251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995364420772821,
      "loss": 5.9642,
      "mean_token_accuracy": 0.15163593217730523,
      "num_tokens": 6270529.0,
      "step": 2800
    },
    {
      "entropy": 6.0933667659759525,
      "epoch": 0.31504464536418264,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995338626344702,
      "loss": 5.8579,
      "mean_token_accuracy": 0.16046109050512314,
      "num_tokens": 6280891.0,
      "step": 2805
    },
    {
      "entropy": 5.7229210376739506,
      "epoch": 0.3156062222721402,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499531276042413,
      "loss": 5.7988,
      "mean_token_accuracy": 0.15810492336750032,
      "num_tokens": 6292247.0,
      "step": 2810
    },
    {
      "entropy": 5.79867639541626,
      "epoch": 0.31616779918009774,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995286823011929,
      "loss": 5.7847,
      "mean_token_accuracy": 0.16373041421175002,
      "num_tokens": 6302403.0,
      "step": 2815
    },
    {
      "entropy": 5.7128393173217775,
      "epoch": 0.31672937608805524,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995260814108926,
      "loss": 5.6696,
      "mean_token_accuracy": 0.1683839187026024,
      "num_tokens": 6313901.0,
      "step": 2820
    },
    {
      "entropy": 5.921040821075439,
      "epoch": 0.3172909529960128,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995234733715947,
      "loss": 5.8928,
      "mean_token_accuracy": 0.15783516615629195,
      "num_tokens": 6323711.0,
      "step": 2825
    },
    {
      "entropy": 5.817658710479736,
      "epoch": 0.31785252990397034,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995208581833824,
      "loss": 5.7877,
      "mean_token_accuracy": 0.1626381605863571,
      "num_tokens": 6334937.0,
      "step": 2830
    },
    {
      "entropy": 5.928051614761353,
      "epoch": 0.3184141068119279,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995182358463388,
      "loss": 5.8838,
      "mean_token_accuracy": 0.15879348814487457,
      "num_tokens": 6345723.0,
      "step": 2835
    },
    {
      "entropy": 5.844352579116821,
      "epoch": 0.31897568371988544,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995156063605475,
      "loss": 5.8219,
      "mean_token_accuracy": 0.15776338502764703,
      "num_tokens": 6357573.0,
      "step": 2840
    },
    {
      "entropy": 5.8675610542297365,
      "epoch": 0.319537260627843,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995129697260922,
      "loss": 5.8322,
      "mean_token_accuracy": 0.16539549678564072,
      "num_tokens": 6367417.0,
      "step": 2845
    },
    {
      "entropy": 5.872849082946777,
      "epoch": 0.32009883753580054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995103259430569,
      "loss": 5.841,
      "mean_token_accuracy": 0.16827336698770523,
      "num_tokens": 6377753.0,
      "step": 2850
    },
    {
      "entropy": 5.839739036560059,
      "epoch": 0.3206604144437581,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995076750115258,
      "loss": 5.8514,
      "mean_token_accuracy": 0.15937740802764894,
      "num_tokens": 6388613.0,
      "step": 2855
    },
    {
      "entropy": 5.815576601028442,
      "epoch": 0.32122199135171564,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995050169315832,
      "loss": 5.796,
      "mean_token_accuracy": 0.1593346953392029,
      "num_tokens": 6399415.0,
      "step": 2860
    },
    {
      "entropy": 5.867472791671753,
      "epoch": 0.32178356825967314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995023517033138,
      "loss": 5.8285,
      "mean_token_accuracy": 0.15754698514938353,
      "num_tokens": 6411435.0,
      "step": 2865
    },
    {
      "entropy": 5.926137781143188,
      "epoch": 0.3223451451676307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994996793268024,
      "loss": 5.976,
      "mean_token_accuracy": 0.15850219279527664,
      "num_tokens": 6423030.0,
      "step": 2870
    },
    {
      "entropy": 5.888271427154541,
      "epoch": 0.32290672207558824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994969998021342,
      "loss": 5.7765,
      "mean_token_accuracy": 0.16209373772144317,
      "num_tokens": 6434330.0,
      "step": 2875
    },
    {
      "entropy": 5.859491014480591,
      "epoch": 0.3234682989835458,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994943131293944,
      "loss": 5.8646,
      "mean_token_accuracy": 0.16070864647626876,
      "num_tokens": 6446452.0,
      "step": 2880
    },
    {
      "entropy": 5.8355710983276365,
      "epoch": 0.32402987589150334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994916193086687,
      "loss": 5.8139,
      "mean_token_accuracy": 0.1690973237156868,
      "num_tokens": 6456723.0,
      "step": 2885
    },
    {
      "entropy": 5.899684286117553,
      "epoch": 0.3245914527994609,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994889183400428,
      "loss": 5.9365,
      "mean_token_accuracy": 0.15811522603034972,
      "num_tokens": 6468117.0,
      "step": 2890
    },
    {
      "entropy": 5.826454973220825,
      "epoch": 0.32515302970741844,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994862102236025,
      "loss": 5.7204,
      "mean_token_accuracy": 0.16832807213068007,
      "num_tokens": 6479465.0,
      "step": 2895
    },
    {
      "entropy": 5.89757194519043,
      "epoch": 0.325714606615376,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004994834949594342,
      "loss": 5.9526,
      "mean_token_accuracy": 0.1629941999912262,
      "num_tokens": 6490868.0,
      "step": 2900
    },
    {
      "entropy": 5.972553682327271,
      "epoch": 0.32627618352333354,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994807725476246,
      "loss": 5.8245,
      "mean_token_accuracy": 0.16008032113313675,
      "num_tokens": 6502126.0,
      "step": 2905
    },
    {
      "entropy": 5.72076644897461,
      "epoch": 0.3268377604312911,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994780429882599,
      "loss": 5.9091,
      "mean_token_accuracy": 0.15670095086097718,
      "num_tokens": 6514382.0,
      "step": 2910
    },
    {
      "entropy": 5.882012605667114,
      "epoch": 0.3273993373392486,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994753062814274,
      "loss": 5.7914,
      "mean_token_accuracy": 0.15964423269033431,
      "num_tokens": 6526272.0,
      "step": 2915
    },
    {
      "entropy": 5.866258335113526,
      "epoch": 0.32796091424720614,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994725624272139,
      "loss": 5.8134,
      "mean_token_accuracy": 0.16369341611862182,
      "num_tokens": 6537511.0,
      "step": 2920
    },
    {
      "entropy": 5.805434799194336,
      "epoch": 0.3285224911551637,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994698114257071,
      "loss": 5.7523,
      "mean_token_accuracy": 0.16356185525655748,
      "num_tokens": 6549737.0,
      "step": 2925
    },
    {
      "entropy": 5.735032749176026,
      "epoch": 0.32908406806312124,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004994670532769944,
      "loss": 5.842,
      "mean_token_accuracy": 0.16296304613351822,
      "num_tokens": 6560318.0,
      "step": 2930
    },
    {
      "entropy": 5.818574094772339,
      "epoch": 0.3296456449710788,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994642879811637,
      "loss": 5.7497,
      "mean_token_accuracy": 0.16560823023319243,
      "num_tokens": 6570575.0,
      "step": 2935
    },
    {
      "entropy": 5.807329988479614,
      "epoch": 0.33020722187903634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499461515538303,
      "loss": 5.8172,
      "mean_token_accuracy": 0.1601566642522812,
      "num_tokens": 6582083.0,
      "step": 2940
    },
    {
      "entropy": 6.00074577331543,
      "epoch": 0.3307687987869939,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994587359485006,
      "loss": 5.9414,
      "mean_token_accuracy": 0.16097475886344909,
      "num_tokens": 6592240.0,
      "step": 2945
    },
    {
      "entropy": 5.806456136703491,
      "epoch": 0.33133037569495144,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499455949211845,
      "loss": 5.7185,
      "mean_token_accuracy": 0.1677255555987358,
      "num_tokens": 6602680.0,
      "step": 2950
    },
    {
      "entropy": 5.796722078323365,
      "epoch": 0.331891952602909,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994531553284248,
      "loss": 5.8381,
      "mean_token_accuracy": 0.16280671954154968,
      "num_tokens": 6613399.0,
      "step": 2955
    },
    {
      "entropy": 5.918023681640625,
      "epoch": 0.3324535295108665,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994503542983293,
      "loss": 5.91,
      "mean_token_accuracy": 0.15692875981330873,
      "num_tokens": 6624410.0,
      "step": 2960
    },
    {
      "entropy": 5.893260526657104,
      "epoch": 0.33301510641882404,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994475461216472,
      "loss": 5.7623,
      "mean_token_accuracy": 0.1689491778612137,
      "num_tokens": 6635523.0,
      "step": 2965
    },
    {
      "entropy": 5.8270707607269285,
      "epoch": 0.3335766833267816,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004994447307984684,
      "loss": 5.9249,
      "mean_token_accuracy": 0.1531895712018013,
      "num_tokens": 6648416.0,
      "step": 2970
    },
    {
      "entropy": 5.764549970626831,
      "epoch": 0.33413826023473914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994419083288823,
      "loss": 5.7112,
      "mean_token_accuracy": 0.16065409630537034,
      "num_tokens": 6658907.0,
      "step": 2975
    },
    {
      "entropy": 5.840141725540161,
      "epoch": 0.3346998371426967,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994390787129787,
      "loss": 5.8754,
      "mean_token_accuracy": 0.1610553666949272,
      "num_tokens": 6669682.0,
      "step": 2980
    },
    {
      "entropy": 5.9260087490081785,
      "epoch": 0.33526141405065424,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994362419508478,
      "loss": 5.9334,
      "mean_token_accuracy": 0.1567075252532959,
      "num_tokens": 6681919.0,
      "step": 2985
    },
    {
      "entropy": 5.9462815284729,
      "epoch": 0.3358229909586118,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049943339804258,
      "loss": 5.8707,
      "mean_token_accuracy": 0.1613776594400406,
      "num_tokens": 6693988.0,
      "step": 2990
    },
    {
      "entropy": 5.7616931915283205,
      "epoch": 0.33638456786656934,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004994305469882657,
      "loss": 5.8059,
      "mean_token_accuracy": 0.15830975472927095,
      "num_tokens": 6705413.0,
      "step": 2995
    },
    {
      "entropy": 5.859702157974243,
      "epoch": 0.3369461447745269,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994276887879957,
      "loss": 5.763,
      "mean_token_accuracy": 0.16223098933696747,
      "num_tokens": 6716217.0,
      "step": 3000
    },
    {
      "epoch": 0.3369461447745269,
      "eval_entropy": 5.720634406333059,
      "eval_loss": 5.844816207885742,
      "eval_mean_token_accuracy": 0.1682441681402459,
      "eval_num_tokens": 6716217.0,
      "eval_runtime": 23.7412,
      "eval_samples_per_second": 1306.296,
      "eval_steps_per_second": 163.303,
      "step": 3000
    },
    {
      "entropy": 5.799144506454468,
      "epoch": 0.33750772168248444,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994248234418611,
      "loss": 5.8366,
      "mean_token_accuracy": 0.162978595495224,
      "num_tokens": 6728059.0,
      "step": 3005
    },
    {
      "entropy": 5.893440103530883,
      "epoch": 0.33806929859044194,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499421950949953,
      "loss": 5.7807,
      "mean_token_accuracy": 0.16390309780836104,
      "num_tokens": 6740650.0,
      "step": 3010
    },
    {
      "entropy": 5.698744678497315,
      "epoch": 0.3386308754983995,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499419071312363,
      "loss": 5.6836,
      "mean_token_accuracy": 0.16469298750162126,
      "num_tokens": 6751467.0,
      "step": 3015
    },
    {
      "entropy": 5.7961132526397705,
      "epoch": 0.33919245240635704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994161845291828,
      "loss": 5.7861,
      "mean_token_accuracy": 0.16712161004543305,
      "num_tokens": 6761377.0,
      "step": 3020
    },
    {
      "entropy": 5.79654335975647,
      "epoch": 0.3397540293143146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994132906005041,
      "loss": 5.7393,
      "mean_token_accuracy": 0.16078124195337296,
      "num_tokens": 6773296.0,
      "step": 3025
    },
    {
      "entropy": 5.845974779129028,
      "epoch": 0.34031560622227214,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994103895264193,
      "loss": 5.7954,
      "mean_token_accuracy": 0.15900143682956697,
      "num_tokens": 6784122.0,
      "step": 3030
    },
    {
      "entropy": 5.704025363922119,
      "epoch": 0.3408771831302297,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994074813070207,
      "loss": 5.6383,
      "mean_token_accuracy": 0.16705414056777954,
      "num_tokens": 6795604.0,
      "step": 3035
    },
    {
      "entropy": 5.769487285614014,
      "epoch": 0.34143876003818724,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994045659424007,
      "loss": 5.7981,
      "mean_token_accuracy": 0.16261281818151474,
      "num_tokens": 6806671.0,
      "step": 3040
    },
    {
      "entropy": 5.9057597637176515,
      "epoch": 0.3420003369461448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994016434326524,
      "loss": 5.8216,
      "mean_token_accuracy": 0.16332198679447174,
      "num_tokens": 6818271.0,
      "step": 3045
    },
    {
      "entropy": 5.7448649406433105,
      "epoch": 0.34256191385410234,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993987137778687,
      "loss": 5.8169,
      "mean_token_accuracy": 0.1635512739419937,
      "num_tokens": 6830049.0,
      "step": 3050
    },
    {
      "entropy": 5.942811059951782,
      "epoch": 0.3431234907620599,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499395776978143,
      "loss": 5.9403,
      "mean_token_accuracy": 0.15486131906509398,
      "num_tokens": 6842275.0,
      "step": 3055
    },
    {
      "entropy": 5.956660890579224,
      "epoch": 0.3436850676700174,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993928330335686,
      "loss": 5.8766,
      "mean_token_accuracy": 0.1637545719742775,
      "num_tokens": 6854910.0,
      "step": 3060
    },
    {
      "entropy": 5.791312170028687,
      "epoch": 0.34424664457797494,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993898819442395,
      "loss": 5.8284,
      "mean_token_accuracy": 0.16965375244617462,
      "num_tokens": 6866031.0,
      "step": 3065
    },
    {
      "entropy": 5.93616361618042,
      "epoch": 0.3448082214859325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993869237102494,
      "loss": 5.8471,
      "mean_token_accuracy": 0.15883387625217438,
      "num_tokens": 6877672.0,
      "step": 3070
    },
    {
      "entropy": 5.765334987640381,
      "epoch": 0.34536979839389004,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993839583316927,
      "loss": 5.7543,
      "mean_token_accuracy": 0.16836509853601456,
      "num_tokens": 6889089.0,
      "step": 3075
    },
    {
      "entropy": 5.673844814300537,
      "epoch": 0.3459313753018476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993809858086638,
      "loss": 5.6994,
      "mean_token_accuracy": 0.1676324874162674,
      "num_tokens": 6899877.0,
      "step": 3080
    },
    {
      "entropy": 5.793374395370483,
      "epoch": 0.34649295220980514,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993780061412571,
      "loss": 5.6829,
      "mean_token_accuracy": 0.1715866893529892,
      "num_tokens": 6910686.0,
      "step": 3085
    },
    {
      "entropy": 5.734429454803466,
      "epoch": 0.3470545291177627,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993750193295678,
      "loss": 5.7143,
      "mean_token_accuracy": 0.17135517448186874,
      "num_tokens": 6920867.0,
      "step": 3090
    },
    {
      "entropy": 5.899741554260254,
      "epoch": 0.34761610602572024,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993720253736909,
      "loss": 5.9044,
      "mean_token_accuracy": 0.16466446667909623,
      "num_tokens": 6932779.0,
      "step": 3095
    },
    {
      "entropy": 5.86834020614624,
      "epoch": 0.3481776829336778,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004993690242737216,
      "loss": 5.8055,
      "mean_token_accuracy": 0.16673210114240647,
      "num_tokens": 6943378.0,
      "step": 3100
    },
    {
      "entropy": 5.760788822174073,
      "epoch": 0.3487392598416353,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993660160297555,
      "loss": 5.7867,
      "mean_token_accuracy": 0.16799272298812867,
      "num_tokens": 6955469.0,
      "step": 3105
    },
    {
      "entropy": 5.971051931381226,
      "epoch": 0.34930083674959284,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004993630006418885,
      "loss": 5.8788,
      "mean_token_accuracy": 0.16197884529829026,
      "num_tokens": 6966750.0,
      "step": 3110
    },
    {
      "entropy": 5.811239528656006,
      "epoch": 0.3498624136575504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993599781102166,
      "loss": 5.7912,
      "mean_token_accuracy": 0.16256844401359558,
      "num_tokens": 6979010.0,
      "step": 3115
    },
    {
      "entropy": 5.775843524932862,
      "epoch": 0.35042399056550794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993569484348359,
      "loss": 5.7764,
      "mean_token_accuracy": 0.16187913715839386,
      "num_tokens": 6990566.0,
      "step": 3120
    },
    {
      "entropy": 5.8379576206207275,
      "epoch": 0.3509855674734655,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004993539116158429,
      "loss": 5.7852,
      "mean_token_accuracy": 0.1654660627245903,
      "num_tokens": 7000288.0,
      "step": 3125
    },
    {
      "entropy": 5.689918518066406,
      "epoch": 0.35154714438142304,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993508676533346,
      "loss": 5.7248,
      "mean_token_accuracy": 0.16818051636219025,
      "num_tokens": 7011143.0,
      "step": 3130
    },
    {
      "entropy": 5.967591333389282,
      "epoch": 0.3521087212893806,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993478165474074,
      "loss": 5.8802,
      "mean_token_accuracy": 0.15500763505697251,
      "num_tokens": 7022797.0,
      "step": 3135
    },
    {
      "entropy": 5.829250907897949,
      "epoch": 0.35267029819733814,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993447582981588,
      "loss": 5.8523,
      "mean_token_accuracy": 0.15978631526231765,
      "num_tokens": 7033477.0,
      "step": 3140
    },
    {
      "entropy": 5.825192642211914,
      "epoch": 0.3532318751052957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499341692905686,
      "loss": 5.8713,
      "mean_token_accuracy": 0.16687492281198502,
      "num_tokens": 7045249.0,
      "step": 3145
    },
    {
      "entropy": 5.924324893951416,
      "epoch": 0.35379345201325324,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993386203700868,
      "loss": 5.8924,
      "mean_token_accuracy": 0.16145936846733094,
      "num_tokens": 7056890.0,
      "step": 3150
    },
    {
      "entropy": 5.861443424224854,
      "epoch": 0.35435502892121074,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993355406914589,
      "loss": 5.7678,
      "mean_token_accuracy": 0.1675198808312416,
      "num_tokens": 7067181.0,
      "step": 3155
    },
    {
      "entropy": 5.785042476654053,
      "epoch": 0.3549166058291683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993324538699003,
      "loss": 5.7324,
      "mean_token_accuracy": 0.16697331219911576,
      "num_tokens": 7078220.0,
      "step": 3160
    },
    {
      "entropy": 5.833939266204834,
      "epoch": 0.35547818273712584,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993293599055094,
      "loss": 5.8066,
      "mean_token_accuracy": 0.16511004269123078,
      "num_tokens": 7089536.0,
      "step": 3165
    },
    {
      "entropy": 5.8496153354644775,
      "epoch": 0.3560397596450834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993262587983848,
      "loss": 5.8848,
      "mean_token_accuracy": 0.16400568187236786,
      "num_tokens": 7101535.0,
      "step": 3170
    },
    {
      "entropy": 5.862979078292847,
      "epoch": 0.35660133655304094,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499323150548625,
      "loss": 5.7782,
      "mean_token_accuracy": 0.1608920007944107,
      "num_tokens": 7112468.0,
      "step": 3175
    },
    {
      "entropy": 5.801843690872192,
      "epoch": 0.3571629134609985,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004993200351563291,
      "loss": 5.7695,
      "mean_token_accuracy": 0.16856954991817474,
      "num_tokens": 7123192.0,
      "step": 3180
    },
    {
      "entropy": 5.776076078414917,
      "epoch": 0.35772449036895604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993169126215962,
      "loss": 5.6971,
      "mean_token_accuracy": 0.17152681350708007,
      "num_tokens": 7133063.0,
      "step": 3185
    },
    {
      "entropy": 5.832470464706421,
      "epoch": 0.3582860672769136,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993137829445259,
      "loss": 5.7899,
      "mean_token_accuracy": 0.16362055987119675,
      "num_tokens": 7144190.0,
      "step": 3190
    },
    {
      "entropy": 5.7901458740234375,
      "epoch": 0.35884764418487114,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993106461252177,
      "loss": 5.779,
      "mean_token_accuracy": 0.1660066932439804,
      "num_tokens": 7155433.0,
      "step": 3195
    },
    {
      "entropy": 5.8804432392120365,
      "epoch": 0.3594092210928287,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993075021637716,
      "loss": 5.8442,
      "mean_token_accuracy": 0.16579061150550842,
      "num_tokens": 7166084.0,
      "step": 3200
    },
    {
      "entropy": 5.849256277084351,
      "epoch": 0.3599707980007862,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993043510602875,
      "loss": 5.8714,
      "mean_token_accuracy": 0.1630292773246765,
      "num_tokens": 7176948.0,
      "step": 3205
    },
    {
      "entropy": 5.846883821487427,
      "epoch": 0.36053237490874374,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499301192814866,
      "loss": 5.7289,
      "mean_token_accuracy": 0.16961728185415267,
      "num_tokens": 7188601.0,
      "step": 3210
    },
    {
      "entropy": 5.7951243877410885,
      "epoch": 0.3610939518167013,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004992980274276074,
      "loss": 5.7635,
      "mean_token_accuracy": 0.16846298575401306,
      "num_tokens": 7198495.0,
      "step": 3215
    },
    {
      "entropy": 5.826298189163208,
      "epoch": 0.36165552872465884,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004992948548986128,
      "loss": 5.753,
      "mean_token_accuracy": 0.16387602537870408,
      "num_tokens": 7209848.0,
      "step": 3220
    },
    {
      "entropy": 5.814381456375122,
      "epoch": 0.3622171056326164,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004992916752279829,
      "loss": 5.8787,
      "mean_token_accuracy": 0.16065715700387956,
      "num_tokens": 7220366.0,
      "step": 3225
    },
    {
      "entropy": 5.806219387054443,
      "epoch": 0.36277868254057394,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499288488415819,
      "loss": 5.7833,
      "mean_token_accuracy": 0.16056074649095536,
      "num_tokens": 7231442.0,
      "step": 3230
    },
    {
      "entropy": 5.850152111053466,
      "epoch": 0.3633402594485315,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004992852944622229,
      "loss": 5.825,
      "mean_token_accuracy": 0.1608233243227005,
      "num_tokens": 7242391.0,
      "step": 3235
    },
    {
      "entropy": 5.733070659637451,
      "epoch": 0.36390183635648904,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004992820933672958,
      "loss": 5.7124,
      "mean_token_accuracy": 0.16292389035224913,
      "num_tokens": 7252061.0,
      "step": 3240
    },
    {
      "entropy": 5.832701826095581,
      "epoch": 0.3644634132644466,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992788851311399,
      "loss": 5.7926,
      "mean_token_accuracy": 0.1660913810133934,
      "num_tokens": 7264035.0,
      "step": 3245
    },
    {
      "entropy": 5.919405937194824,
      "epoch": 0.3650249901724041,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992756697538574,
      "loss": 5.8652,
      "mean_token_accuracy": 0.16157617121934892,
      "num_tokens": 7275988.0,
      "step": 3250
    },
    {
      "entropy": 5.77119779586792,
      "epoch": 0.36558656708036164,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992724472355505,
      "loss": 5.7206,
      "mean_token_accuracy": 0.16446294635534286,
      "num_tokens": 7286509.0,
      "step": 3255
    },
    {
      "entropy": 5.865907192230225,
      "epoch": 0.3661481439883192,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004992692175763219,
      "loss": 5.8134,
      "mean_token_accuracy": 0.16690248399972915,
      "num_tokens": 7296916.0,
      "step": 3260
    },
    {
      "entropy": 5.820225429534912,
      "epoch": 0.36670972089627674,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992659807762745,
      "loss": 5.8982,
      "mean_token_accuracy": 0.15828224942088126,
      "num_tokens": 7309563.0,
      "step": 3265
    },
    {
      "entropy": 5.88502197265625,
      "epoch": 0.3672712978042343,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992627368355112,
      "loss": 5.7505,
      "mean_token_accuracy": 0.16168039441108703,
      "num_tokens": 7320264.0,
      "step": 3270
    },
    {
      "entropy": 5.719604015350342,
      "epoch": 0.36783287471219184,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004992594857541355,
      "loss": 5.7186,
      "mean_token_accuracy": 0.16655558794736863,
      "num_tokens": 7330442.0,
      "step": 3275
    },
    {
      "entropy": 5.853730297088623,
      "epoch": 0.3683944516201494,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004992562275322506,
      "loss": 5.9159,
      "mean_token_accuracy": 0.1572217419743538,
      "num_tokens": 7342292.0,
      "step": 3280
    },
    {
      "entropy": 5.909031820297241,
      "epoch": 0.36895602852810694,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004992529621699606,
      "loss": 5.8141,
      "mean_token_accuracy": 0.16075354367494582,
      "num_tokens": 7353857.0,
      "step": 3285
    },
    {
      "entropy": 5.804742479324341,
      "epoch": 0.3695176054360645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992496896673692,
      "loss": 5.7377,
      "mean_token_accuracy": 0.16423878073692322,
      "num_tokens": 7365653.0,
      "step": 3290
    },
    {
      "entropy": 5.790262269973755,
      "epoch": 0.37007918234402204,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992464100245809,
      "loss": 5.7455,
      "mean_token_accuracy": 0.15999912321567536,
      "num_tokens": 7376522.0,
      "step": 3295
    },
    {
      "entropy": 5.796634531021118,
      "epoch": 0.37064075925197953,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004992431232416997,
      "loss": 5.7407,
      "mean_token_accuracy": 0.1694473460316658,
      "num_tokens": 7388562.0,
      "step": 3300
    },
    {
      "entropy": 5.8152648448944095,
      "epoch": 0.3712023361599371,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992398293188307,
      "loss": 5.794,
      "mean_token_accuracy": 0.16164605543017388,
      "num_tokens": 7398597.0,
      "step": 3305
    },
    {
      "entropy": 5.75906810760498,
      "epoch": 0.37176391306789464,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992365282560784,
      "loss": 5.7432,
      "mean_token_accuracy": 0.16698759496212007,
      "num_tokens": 7409656.0,
      "step": 3310
    },
    {
      "entropy": 5.769649410247803,
      "epoch": 0.3723254899758522,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004992332200535483,
      "loss": 5.7906,
      "mean_token_accuracy": 0.1635170191526413,
      "num_tokens": 7420621.0,
      "step": 3315
    },
    {
      "entropy": 5.914590549468994,
      "epoch": 0.37288706688380974,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992299047113452,
      "loss": 5.8258,
      "mean_token_accuracy": 0.1644177705049515,
      "num_tokens": 7432951.0,
      "step": 3320
    },
    {
      "entropy": 5.746472978591919,
      "epoch": 0.3734486437917673,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992265822295752,
      "loss": 5.8086,
      "mean_token_accuracy": 0.1657330572605133,
      "num_tokens": 7443630.0,
      "step": 3325
    },
    {
      "entropy": 5.878232765197754,
      "epoch": 0.37401022069972484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992232526083438,
      "loss": 5.7904,
      "mean_token_accuracy": 0.1630190759897232,
      "num_tokens": 7454535.0,
      "step": 3330
    },
    {
      "entropy": 5.851468801498413,
      "epoch": 0.3745717976076824,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499219915847757,
      "loss": 5.8617,
      "mean_token_accuracy": 0.16609936952590942,
      "num_tokens": 7467751.0,
      "step": 3335
    },
    {
      "entropy": 5.7327392578125,
      "epoch": 0.37513337451563994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992165719479211,
      "loss": 5.7229,
      "mean_token_accuracy": 0.1723914325237274,
      "num_tokens": 7479347.0,
      "step": 3340
    },
    {
      "entropy": 5.792557239532471,
      "epoch": 0.37569495142359743,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992132209089427,
      "loss": 5.7559,
      "mean_token_accuracy": 0.16422603875398636,
      "num_tokens": 7490916.0,
      "step": 3345
    },
    {
      "entropy": 5.839074325561524,
      "epoch": 0.376256528331555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992098627309283,
      "loss": 5.7285,
      "mean_token_accuracy": 0.16606951653957366,
      "num_tokens": 7500824.0,
      "step": 3350
    },
    {
      "entropy": 5.8109547138214115,
      "epoch": 0.37681810523951254,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499206497413985,
      "loss": 5.8818,
      "mean_token_accuracy": 0.15842843651771546,
      "num_tokens": 7511778.0,
      "step": 3355
    },
    {
      "entropy": 5.980834197998047,
      "epoch": 0.3773796821474701,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992031249582197,
      "loss": 5.9424,
      "mean_token_accuracy": 0.15256146043539048,
      "num_tokens": 7522616.0,
      "step": 3360
    },
    {
      "entropy": 5.881027460098267,
      "epoch": 0.37794125905542764,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004991997453637401,
      "loss": 5.8116,
      "mean_token_accuracy": 0.1650218188762665,
      "num_tokens": 7533019.0,
      "step": 3365
    },
    {
      "entropy": 5.749409818649292,
      "epoch": 0.3785028359633852,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991963586306536,
      "loss": 5.7544,
      "mean_token_accuracy": 0.16154202967882156,
      "num_tokens": 7544139.0,
      "step": 3370
    },
    {
      "entropy": 5.80066704750061,
      "epoch": 0.37906441287134274,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004991929647590682,
      "loss": 5.7377,
      "mean_token_accuracy": 0.16260901242494583,
      "num_tokens": 7554638.0,
      "step": 3375
    },
    {
      "entropy": 5.953235530853272,
      "epoch": 0.3796259897793003,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991895637490917,
      "loss": 5.8049,
      "mean_token_accuracy": 0.16714229732751845,
      "num_tokens": 7565158.0,
      "step": 3380
    },
    {
      "entropy": 5.809656095504761,
      "epoch": 0.38018756668725784,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991861556008326,
      "loss": 5.811,
      "mean_token_accuracy": 0.16301512867212295,
      "num_tokens": 7576381.0,
      "step": 3385
    },
    {
      "entropy": 5.718403387069702,
      "epoch": 0.3807491435952154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991827403143992,
      "loss": 5.6948,
      "mean_token_accuracy": 0.1689286082983017,
      "num_tokens": 7586631.0,
      "step": 3390
    },
    {
      "entropy": 5.807317209243775,
      "epoch": 0.3813107205031729,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004991793178899006,
      "loss": 5.7382,
      "mean_token_accuracy": 0.165502168238163,
      "num_tokens": 7598321.0,
      "step": 3395
    },
    {
      "entropy": 5.771745109558106,
      "epoch": 0.38187229741113043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991758883274455,
      "loss": 5.7698,
      "mean_token_accuracy": 0.1654894694685936,
      "num_tokens": 7609336.0,
      "step": 3400
    },
    {
      "entropy": 5.721436452865601,
      "epoch": 0.382433874319088,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004991724516271432,
      "loss": 5.6716,
      "mean_token_accuracy": 0.16680669486522676,
      "num_tokens": 7619116.0,
      "step": 3405
    },
    {
      "entropy": 5.82664680480957,
      "epoch": 0.38299545122704554,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991690077891032,
      "loss": 5.8092,
      "mean_token_accuracy": 0.15846319049596785,
      "num_tokens": 7630567.0,
      "step": 3410
    },
    {
      "entropy": 5.8768939018249515,
      "epoch": 0.3835570281350031,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991655568134348,
      "loss": 5.8718,
      "mean_token_accuracy": 0.16050247251987457,
      "num_tokens": 7642564.0,
      "step": 3415
    },
    {
      "entropy": 5.909596014022827,
      "epoch": 0.38411860504296064,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004991620987002483,
      "loss": 5.8983,
      "mean_token_accuracy": 0.15932083278894424,
      "num_tokens": 7654588.0,
      "step": 3420
    },
    {
      "entropy": 5.780777931213379,
      "epoch": 0.3846801819509182,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004991586334496536,
      "loss": 5.6903,
      "mean_token_accuracy": 0.1646028310060501,
      "num_tokens": 7666120.0,
      "step": 3425
    },
    {
      "entropy": 5.796514987945557,
      "epoch": 0.38524175885887574,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004991551610617611,
      "loss": 5.8408,
      "mean_token_accuracy": 0.16049385517835618,
      "num_tokens": 7677132.0,
      "step": 3430
    },
    {
      "entropy": 5.968781232833862,
      "epoch": 0.3858033357668333,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991516815366812,
      "loss": 5.9041,
      "mean_token_accuracy": 0.1601366326212883,
      "num_tokens": 7687515.0,
      "step": 3435
    },
    {
      "entropy": 5.767029762268066,
      "epoch": 0.38636491267479084,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499148194874525,
      "loss": 5.7118,
      "mean_token_accuracy": 0.16504708528518677,
      "num_tokens": 7698454.0,
      "step": 3440
    },
    {
      "entropy": 5.698157548904419,
      "epoch": 0.38692648958274833,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004991447010754032,
      "loss": 5.6899,
      "mean_token_accuracy": 0.1678033024072647,
      "num_tokens": 7708940.0,
      "step": 3445
    },
    {
      "entropy": 5.8262519359588625,
      "epoch": 0.3874880664907059,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004991412001394273,
      "loss": 5.7239,
      "mean_token_accuracy": 0.16594443321228028,
      "num_tokens": 7719705.0,
      "step": 3450
    },
    {
      "entropy": 5.863242149353027,
      "epoch": 0.38804964339866344,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004991376920667086,
      "loss": 5.8811,
      "mean_token_accuracy": 0.15880899280309677,
      "num_tokens": 7730649.0,
      "step": 3455
    },
    {
      "entropy": 5.779720497131348,
      "epoch": 0.388611220306621,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991341768573589,
      "loss": 5.8953,
      "mean_token_accuracy": 0.1590091198682785,
      "num_tokens": 7742014.0,
      "step": 3460
    },
    {
      "entropy": 5.901068115234375,
      "epoch": 0.38917279721457854,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049913065451149,
      "loss": 5.799,
      "mean_token_accuracy": 0.1558951511979103,
      "num_tokens": 7752892.0,
      "step": 3465
    },
    {
      "entropy": 5.8446044921875,
      "epoch": 0.3897343741225361,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991271250292141,
      "loss": 5.81,
      "mean_token_accuracy": 0.157416270673275,
      "num_tokens": 7762959.0,
      "step": 3470
    },
    {
      "entropy": 5.81669659614563,
      "epoch": 0.39029595103049364,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004991235884106437,
      "loss": 5.7144,
      "mean_token_accuracy": 0.16509674340486527,
      "num_tokens": 7773566.0,
      "step": 3475
    },
    {
      "entropy": 5.7814857959747314,
      "epoch": 0.3908575279384512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991200446558913,
      "loss": 5.748,
      "mean_token_accuracy": 0.15518904775381087,
      "num_tokens": 7785782.0,
      "step": 3480
    },
    {
      "entropy": 5.841710948944092,
      "epoch": 0.39141910484640874,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991164937650698,
      "loss": 5.9032,
      "mean_token_accuracy": 0.16198572665452957,
      "num_tokens": 7797551.0,
      "step": 3485
    },
    {
      "entropy": 5.751931428909302,
      "epoch": 0.39198068175436623,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991129357382922,
      "loss": 5.6772,
      "mean_token_accuracy": 0.17679855972528458,
      "num_tokens": 7808819.0,
      "step": 3490
    },
    {
      "entropy": 5.647474765777588,
      "epoch": 0.3925422586623238,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004991093705756718,
      "loss": 5.6283,
      "mean_token_accuracy": 0.16952918767929076,
      "num_tokens": 7818940.0,
      "step": 3495
    },
    {
      "entropy": 5.671689987182617,
      "epoch": 0.39310383557028133,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004991057982773221,
      "loss": 5.6507,
      "mean_token_accuracy": 0.17054159790277482,
      "num_tokens": 7829733.0,
      "step": 3500
    },
    {
      "entropy": 5.847164916992187,
      "epoch": 0.3936654124782389,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991022188433569,
      "loss": 5.7133,
      "mean_token_accuracy": 0.17065165787935258,
      "num_tokens": 7840427.0,
      "step": 3505
    },
    {
      "entropy": 5.859947633743286,
      "epoch": 0.39422698938619644,
      "grad_norm": 1.125,
      "learning_rate": 0.00049909863227389,
      "loss": 5.775,
      "mean_token_accuracy": 0.16124107241630553,
      "num_tokens": 7852834.0,
      "step": 3510
    },
    {
      "entropy": 5.7385765552520756,
      "epoch": 0.394788566294154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990950385690358,
      "loss": 5.8178,
      "mean_token_accuracy": 0.1656463325023651,
      "num_tokens": 7864993.0,
      "step": 3515
    },
    {
      "entropy": 5.750460052490235,
      "epoch": 0.39535014320211154,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004990914377289088,
      "loss": 5.793,
      "mean_token_accuracy": 0.16076247096061708,
      "num_tokens": 7876578.0,
      "step": 3520
    },
    {
      "entropy": 5.729760694503784,
      "epoch": 0.3959117201100691,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990878297536233,
      "loss": 5.6272,
      "mean_token_accuracy": 0.17054133117198944,
      "num_tokens": 7886770.0,
      "step": 3525
    },
    {
      "entropy": 5.800265884399414,
      "epoch": 0.39647329701802664,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004990842146432946,
      "loss": 5.7904,
      "mean_token_accuracy": 0.1667746052145958,
      "num_tokens": 7897106.0,
      "step": 3530
    },
    {
      "entropy": 5.749977684020996,
      "epoch": 0.3970348739259842,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990805923980375,
      "loss": 5.6444,
      "mean_token_accuracy": 0.17652151733636856,
      "num_tokens": 7907321.0,
      "step": 3535
    },
    {
      "entropy": 5.651272344589233,
      "epoch": 0.3975964508339417,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990769630179673,
      "loss": 5.6569,
      "mean_token_accuracy": 0.170053867995739,
      "num_tokens": 7918628.0,
      "step": 3540
    },
    {
      "entropy": 5.872209167480468,
      "epoch": 0.39815802774189923,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004990733265032,
      "loss": 5.8387,
      "mean_token_accuracy": 0.15291183888912202,
      "num_tokens": 7929387.0,
      "step": 3545
    },
    {
      "entropy": 5.92167649269104,
      "epoch": 0.3987196046498568,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990696828538508,
      "loss": 5.8595,
      "mean_token_accuracy": 0.16960857212543487,
      "num_tokens": 7940840.0,
      "step": 3550
    },
    {
      "entropy": 5.664200639724731,
      "epoch": 0.39928118155781434,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499066032070036,
      "loss": 5.7269,
      "mean_token_accuracy": 0.1674627184867859,
      "num_tokens": 7952243.0,
      "step": 3555
    },
    {
      "entropy": 5.863605737686157,
      "epoch": 0.3998427584657719,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004990623741518719,
      "loss": 5.7724,
      "mean_token_accuracy": 0.16179074645042418,
      "num_tokens": 7963150.0,
      "step": 3560
    },
    {
      "entropy": 5.818461227416992,
      "epoch": 0.40040433537372944,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499058709099475,
      "loss": 5.8741,
      "mean_token_accuracy": 0.15937056988477707,
      "num_tokens": 7974862.0,
      "step": 3565
    },
    {
      "entropy": 5.877570915222168,
      "epoch": 0.400965912281687,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004990550369129618,
      "loss": 5.8819,
      "mean_token_accuracy": 0.16194065660238266,
      "num_tokens": 7986438.0,
      "step": 3570
    },
    {
      "entropy": 5.8217613697052,
      "epoch": 0.40152748918964454,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004990513575924494,
      "loss": 5.7173,
      "mean_token_accuracy": 0.16897953897714615,
      "num_tokens": 7996958.0,
      "step": 3575
    },
    {
      "entropy": 5.835204267501831,
      "epoch": 0.4020890660976021,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990476711380548,
      "loss": 5.8001,
      "mean_token_accuracy": 0.1581471100449562,
      "num_tokens": 8008862.0,
      "step": 3580
    },
    {
      "entropy": 5.774871110916138,
      "epoch": 0.40265064300555964,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990439775498954,
      "loss": 5.7798,
      "mean_token_accuracy": 0.1676494598388672,
      "num_tokens": 8018940.0,
      "step": 3585
    },
    {
      "entropy": 5.914262056350708,
      "epoch": 0.40321221991351713,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499040276828089,
      "loss": 5.7946,
      "mean_token_accuracy": 0.15799015909433364,
      "num_tokens": 8029148.0,
      "step": 3590
    },
    {
      "entropy": 5.742766189575195,
      "epoch": 0.4037737968214747,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990365689727533,
      "loss": 5.7104,
      "mean_token_accuracy": 0.1686326339840889,
      "num_tokens": 8039618.0,
      "step": 3595
    },
    {
      "entropy": 5.744286060333252,
      "epoch": 0.40433537372943223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990328539840062,
      "loss": 5.6934,
      "mean_token_accuracy": 0.16474545300006865,
      "num_tokens": 8051145.0,
      "step": 3600
    },
    {
      "entropy": 5.7836991310119625,
      "epoch": 0.4048969506373898,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004990291318619663,
      "loss": 5.7609,
      "mean_token_accuracy": 0.1606362998485565,
      "num_tokens": 8061520.0,
      "step": 3605
    },
    {
      "entropy": 5.726670598983764,
      "epoch": 0.40545852754534734,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990254026067518,
      "loss": 5.7099,
      "mean_token_accuracy": 0.16285099536180497,
      "num_tokens": 8072876.0,
      "step": 3610
    },
    {
      "entropy": 5.934379577636719,
      "epoch": 0.4060201044533049,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990216662184817,
      "loss": 5.8154,
      "mean_token_accuracy": 0.16516112834215163,
      "num_tokens": 8084240.0,
      "step": 3615
    },
    {
      "entropy": 5.753713512420655,
      "epoch": 0.40658168136126244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990179226972749,
      "loss": 5.6932,
      "mean_token_accuracy": 0.16864500045776368,
      "num_tokens": 8095773.0,
      "step": 3620
    },
    {
      "entropy": 5.629334354400635,
      "epoch": 0.40714325826922,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004990141720432504,
      "loss": 5.6819,
      "mean_token_accuracy": 0.167998369038105,
      "num_tokens": 8107003.0,
      "step": 3625
    },
    {
      "entropy": 5.72817096710205,
      "epoch": 0.40770483517717754,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004990104142565279,
      "loss": 5.6439,
      "mean_token_accuracy": 0.17401827424764632,
      "num_tokens": 8117864.0,
      "step": 3630
    },
    {
      "entropy": 5.835469007492065,
      "epoch": 0.40826641208513503,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499006649337227,
      "loss": 5.86,
      "mean_token_accuracy": 0.16062401831150055,
      "num_tokens": 8129363.0,
      "step": 3635
    },
    {
      "entropy": 5.805969905853272,
      "epoch": 0.4088279889930926,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990028772854674,
      "loss": 5.7791,
      "mean_token_accuracy": 0.1627799391746521,
      "num_tokens": 8141319.0,
      "step": 3640
    },
    {
      "entropy": 5.784128999710083,
      "epoch": 0.40938956590105013,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004989990981013693,
      "loss": 5.6802,
      "mean_token_accuracy": 0.1702434554696083,
      "num_tokens": 8153344.0,
      "step": 3645
    },
    {
      "entropy": 5.774145793914795,
      "epoch": 0.4099511428090077,
      "grad_norm": 1.15625,
      "learning_rate": 0.000498995311785053,
      "loss": 5.8402,
      "mean_token_accuracy": 0.162425434589386,
      "num_tokens": 8164196.0,
      "step": 3650
    },
    {
      "entropy": 5.816594219207763,
      "epoch": 0.41051271971696524,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004989915183366392,
      "loss": 5.7159,
      "mean_token_accuracy": 0.16790103316307067,
      "num_tokens": 8174663.0,
      "step": 3655
    },
    {
      "entropy": 5.713854694366455,
      "epoch": 0.4110742966249228,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989877177562485,
      "loss": 5.683,
      "mean_token_accuracy": 0.1685110315680504,
      "num_tokens": 8185215.0,
      "step": 3660
    },
    {
      "entropy": 5.739247465133667,
      "epoch": 0.41163587353288034,
      "grad_norm": 1.234375,
      "learning_rate": 0.000498983910044002,
      "loss": 5.7765,
      "mean_token_accuracy": 0.17141249477863313,
      "num_tokens": 8195694.0,
      "step": 3665
    },
    {
      "entropy": 5.681199550628662,
      "epoch": 0.4121974504408379,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989800952000209,
      "loss": 5.6616,
      "mean_token_accuracy": 0.17553078830242158,
      "num_tokens": 8206112.0,
      "step": 3670
    },
    {
      "entropy": 5.731632661819458,
      "epoch": 0.41275902734879544,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004989762732244268,
      "loss": 5.7515,
      "mean_token_accuracy": 0.1661844179034233,
      "num_tokens": 8216548.0,
      "step": 3675
    },
    {
      "entropy": 5.843146276473999,
      "epoch": 0.413320604256753,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004989724441173413,
      "loss": 5.7867,
      "mean_token_accuracy": 0.16781337559223175,
      "num_tokens": 8226643.0,
      "step": 3680
    },
    {
      "entropy": 5.857947587966919,
      "epoch": 0.4138821811647105,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989686078788862,
      "loss": 5.8183,
      "mean_token_accuracy": 0.16166319102048873,
      "num_tokens": 8236755.0,
      "step": 3685
    },
    {
      "entropy": 5.782751274108887,
      "epoch": 0.41444375807266803,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004989647645091839,
      "loss": 5.6963,
      "mean_token_accuracy": 0.16801404505968093,
      "num_tokens": 8247526.0,
      "step": 3690
    },
    {
      "entropy": 5.734130477905273,
      "epoch": 0.4150053349806256,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004989609140083565,
      "loss": 5.794,
      "mean_token_accuracy": 0.15583045184612274,
      "num_tokens": 8258649.0,
      "step": 3695
    },
    {
      "entropy": 5.765203428268433,
      "epoch": 0.41556691188858313,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004989570563765269,
      "loss": 5.676,
      "mean_token_accuracy": 0.1731864795088768,
      "num_tokens": 8268260.0,
      "step": 3700
    },
    {
      "entropy": 5.9308096408844,
      "epoch": 0.4161284887965407,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004989531916138176,
      "loss": 5.8785,
      "mean_token_accuracy": 0.1627708837389946,
      "num_tokens": 8280090.0,
      "step": 3705
    },
    {
      "entropy": 5.841350793838501,
      "epoch": 0.41669006570449824,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989493197203519,
      "loss": 5.7489,
      "mean_token_accuracy": 0.1697134330868721,
      "num_tokens": 8291217.0,
      "step": 3710
    },
    {
      "entropy": 5.754220199584961,
      "epoch": 0.4172516426124558,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000498945440696253,
      "loss": 5.6683,
      "mean_token_accuracy": 0.17081218361854553,
      "num_tokens": 8303541.0,
      "step": 3715
    },
    {
      "entropy": 5.719450426101685,
      "epoch": 0.41781321952041334,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989415545416444,
      "loss": 5.7355,
      "mean_token_accuracy": 0.1609584018588066,
      "num_tokens": 8316104.0,
      "step": 3720
    },
    {
      "entropy": 5.714957666397095,
      "epoch": 0.4183747964283709,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989376612566498,
      "loss": 5.6832,
      "mean_token_accuracy": 0.16480520740151405,
      "num_tokens": 8327158.0,
      "step": 3725
    },
    {
      "entropy": 5.846664381027222,
      "epoch": 0.4189363733363284,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989337608413933,
      "loss": 5.8338,
      "mean_token_accuracy": 0.16076735854148866,
      "num_tokens": 8337685.0,
      "step": 3730
    },
    {
      "entropy": 5.870333480834961,
      "epoch": 0.41949795024428593,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004989298532959989,
      "loss": 5.8683,
      "mean_token_accuracy": 0.16113072335720063,
      "num_tokens": 8349778.0,
      "step": 3735
    },
    {
      "entropy": 5.7798456192016605,
      "epoch": 0.4200595271522435,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004989259386205912,
      "loss": 5.7038,
      "mean_token_accuracy": 0.176594515144825,
      "num_tokens": 8361041.0,
      "step": 3740
    },
    {
      "entropy": 5.8628821849823,
      "epoch": 0.42062110406020103,
      "grad_norm": 1.25,
      "learning_rate": 0.0004989220168152948,
      "loss": 5.8965,
      "mean_token_accuracy": 0.15907129645347595,
      "num_tokens": 8373227.0,
      "step": 3745
    },
    {
      "entropy": 5.7868818759918215,
      "epoch": 0.4211826809681586,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004989180878802346,
      "loss": 5.7209,
      "mean_token_accuracy": 0.16800411641597748,
      "num_tokens": 8384223.0,
      "step": 3750
    },
    {
      "entropy": 5.731780624389648,
      "epoch": 0.42174425787611614,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004989141518155354,
      "loss": 5.7506,
      "mean_token_accuracy": 0.1691988691687584,
      "num_tokens": 8395014.0,
      "step": 3755
    },
    {
      "entropy": 5.823819208145141,
      "epoch": 0.4223058347840737,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004989102086213229,
      "loss": 5.7283,
      "mean_token_accuracy": 0.16270644217729568,
      "num_tokens": 8406445.0,
      "step": 3760
    },
    {
      "entropy": 5.693364000320434,
      "epoch": 0.42286741169203124,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989062582977225,
      "loss": 5.665,
      "mean_token_accuracy": 0.1706314980983734,
      "num_tokens": 8417488.0,
      "step": 3765
    },
    {
      "entropy": 5.793571662902832,
      "epoch": 0.4234289885999888,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989023008448599,
      "loss": 5.8689,
      "mean_token_accuracy": 0.15837376266717912,
      "num_tokens": 8430582.0,
      "step": 3770
    },
    {
      "entropy": 5.831238794326782,
      "epoch": 0.42399056550794634,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004988983362628614,
      "loss": 5.6906,
      "mean_token_accuracy": 0.164572674036026,
      "num_tokens": 8442033.0,
      "step": 3775
    },
    {
      "entropy": 5.690927267074585,
      "epoch": 0.42455214241590383,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004988943645518528,
      "loss": 5.6636,
      "mean_token_accuracy": 0.16775935739278794,
      "num_tokens": 8451722.0,
      "step": 3780
    },
    {
      "entropy": 5.668238401412964,
      "epoch": 0.4251137193238614,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004988903857119611,
      "loss": 5.75,
      "mean_token_accuracy": 0.1658486858010292,
      "num_tokens": 8462753.0,
      "step": 3785
    },
    {
      "entropy": 5.861290073394775,
      "epoch": 0.42567529623181893,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004988863997433125,
      "loss": 5.7587,
      "mean_token_accuracy": 0.16550534069538117,
      "num_tokens": 8472755.0,
      "step": 3790
    },
    {
      "entropy": 5.766337490081787,
      "epoch": 0.4262368731397765,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004988824066460341,
      "loss": 5.7399,
      "mean_token_accuracy": 0.16958845853805543,
      "num_tokens": 8483296.0,
      "step": 3795
    },
    {
      "entropy": 5.7786054611206055,
      "epoch": 0.42679845004773403,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498878406420253,
      "loss": 5.6058,
      "mean_token_accuracy": 0.17350622415542602,
      "num_tokens": 8493739.0,
      "step": 3800
    },
    {
      "entropy": 5.682381916046142,
      "epoch": 0.4273600269556916,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004988743990660967,
      "loss": 5.7767,
      "mean_token_accuracy": 0.162274731695652,
      "num_tokens": 8504788.0,
      "step": 3805
    },
    {
      "entropy": 5.797802543640136,
      "epoch": 0.42792160386364914,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004988703845836927,
      "loss": 5.7306,
      "mean_token_accuracy": 0.1645368903875351,
      "num_tokens": 8514708.0,
      "step": 3810
    },
    {
      "entropy": 5.790168046951294,
      "epoch": 0.4284831807716067,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004988663629731687,
      "loss": 5.6135,
      "mean_token_accuracy": 0.17496834099292755,
      "num_tokens": 8525227.0,
      "step": 3815
    },
    {
      "entropy": 5.622102165222168,
      "epoch": 0.42904475767956424,
      "grad_norm": 1.171875,
      "learning_rate": 0.000498862334234653,
      "loss": 5.6619,
      "mean_token_accuracy": 0.17011212110519408,
      "num_tokens": 8536502.0,
      "step": 3820
    },
    {
      "entropy": 5.7700787544250485,
      "epoch": 0.4296063345875218,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004988582983682737,
      "loss": 5.8033,
      "mean_token_accuracy": 0.16546136736869813,
      "num_tokens": 8548087.0,
      "step": 3825
    },
    {
      "entropy": 5.85570330619812,
      "epoch": 0.4301679114954793,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004988542553741594,
      "loss": 5.809,
      "mean_token_accuracy": 0.1667978622019291,
      "num_tokens": 8560856.0,
      "step": 3830
    },
    {
      "entropy": 5.931719970703125,
      "epoch": 0.43072948840343683,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004988502052524388,
      "loss": 5.9145,
      "mean_token_accuracy": 0.16501590609550476,
      "num_tokens": 8572521.0,
      "step": 3835
    },
    {
      "entropy": 5.84990963935852,
      "epoch": 0.4312910653113944,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004988461480032408,
      "loss": 5.7442,
      "mean_token_accuracy": 0.1674473449587822,
      "num_tokens": 8583977.0,
      "step": 3840
    },
    {
      "entropy": 5.704129409790039,
      "epoch": 0.43185264221935193,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988420836266946,
      "loss": 5.6924,
      "mean_token_accuracy": 0.16754663586616517,
      "num_tokens": 8594441.0,
      "step": 3845
    },
    {
      "entropy": 5.84066071510315,
      "epoch": 0.4324142191273095,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004988380121229298,
      "loss": 5.7392,
      "mean_token_accuracy": 0.16255563497543335,
      "num_tokens": 8604843.0,
      "step": 3850
    },
    {
      "entropy": 5.746822309494019,
      "epoch": 0.43297579603526704,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004988339334920757,
      "loss": 5.7298,
      "mean_token_accuracy": 0.16818083822727203,
      "num_tokens": 8616059.0,
      "step": 3855
    },
    {
      "entropy": 5.74391360282898,
      "epoch": 0.4335373729432246,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004988298477342625,
      "loss": 5.6845,
      "mean_token_accuracy": 0.17156679034233094,
      "num_tokens": 8626203.0,
      "step": 3860
    },
    {
      "entropy": 5.739895057678223,
      "epoch": 0.43409894985118214,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00049882575484962,
      "loss": 5.684,
      "mean_token_accuracy": 0.16509181261062622,
      "num_tokens": 8636921.0,
      "step": 3865
    },
    {
      "entropy": 5.6815019130706785,
      "epoch": 0.4346605267591397,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004988216548382787,
      "loss": 5.6579,
      "mean_token_accuracy": 0.17124742418527603,
      "num_tokens": 8648526.0,
      "step": 3870
    },
    {
      "entropy": 5.796351671218872,
      "epoch": 0.4352221036670972,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004988175477003691,
      "loss": 5.8049,
      "mean_token_accuracy": 0.15377746522426605,
      "num_tokens": 8660339.0,
      "step": 3875
    },
    {
      "entropy": 5.9060732364654545,
      "epoch": 0.43578368057505473,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498813433436022,
      "loss": 5.8632,
      "mean_token_accuracy": 0.1561919257044792,
      "num_tokens": 8671835.0,
      "step": 3880
    },
    {
      "entropy": 5.678639030456543,
      "epoch": 0.4363452574830123,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004988093120453684,
      "loss": 5.5652,
      "mean_token_accuracy": 0.17134900540113449,
      "num_tokens": 8681146.0,
      "step": 3885
    },
    {
      "entropy": 5.650815439224243,
      "epoch": 0.43690683439096983,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004988051835285394,
      "loss": 5.6231,
      "mean_token_accuracy": 0.17194372713565825,
      "num_tokens": 8691623.0,
      "step": 3890
    },
    {
      "entropy": 5.765362548828125,
      "epoch": 0.4374684112989274,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004988010478856667,
      "loss": 5.7243,
      "mean_token_accuracy": 0.16024973541498183,
      "num_tokens": 8703702.0,
      "step": 3895
    },
    {
      "entropy": 5.937811851501465,
      "epoch": 0.43802998820688493,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004987969051168816,
      "loss": 5.8895,
      "mean_token_accuracy": 0.16733449399471284,
      "num_tokens": 8715434.0,
      "step": 3900
    },
    {
      "entropy": 5.7049939155578615,
      "epoch": 0.4385915651148425,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004987927552223165,
      "loss": 5.6636,
      "mean_token_accuracy": 0.17958703339099885,
      "num_tokens": 8726135.0,
      "step": 3905
    },
    {
      "entropy": 5.635489463806152,
      "epoch": 0.43915314202280004,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004987885982021032,
      "loss": 5.761,
      "mean_token_accuracy": 0.16973788887262345,
      "num_tokens": 8736538.0,
      "step": 3910
    },
    {
      "entropy": 5.879006290435791,
      "epoch": 0.4397147189307576,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004987844340563741,
      "loss": 5.6806,
      "mean_token_accuracy": 0.17044662386178971,
      "num_tokens": 8746456.0,
      "step": 3915
    },
    {
      "entropy": 5.7940206050872805,
      "epoch": 0.44027629583871514,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004987802627852619,
      "loss": 5.7909,
      "mean_token_accuracy": 0.16806718856096267,
      "num_tokens": 8757143.0,
      "step": 3920
    },
    {
      "entropy": 5.737888336181641,
      "epoch": 0.44083787274667263,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004987760843888992,
      "loss": 5.7794,
      "mean_token_accuracy": 0.17163707613945006,
      "num_tokens": 8769084.0,
      "step": 3925
    },
    {
      "entropy": 5.792534828186035,
      "epoch": 0.4413994496546302,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004987718988674193,
      "loss": 5.7753,
      "mean_token_accuracy": 0.16346744000911712,
      "num_tokens": 8780707.0,
      "step": 3930
    },
    {
      "entropy": 5.78266863822937,
      "epoch": 0.44196102656258773,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004987677062209552,
      "loss": 5.7305,
      "mean_token_accuracy": 0.16145867258310317,
      "num_tokens": 8792157.0,
      "step": 3935
    },
    {
      "entropy": 5.824311637878418,
      "epoch": 0.4425226034705453,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004987635064496407,
      "loss": 5.7576,
      "mean_token_accuracy": 0.16702643185853958,
      "num_tokens": 8802691.0,
      "step": 3940
    },
    {
      "entropy": 5.8741062641143795,
      "epoch": 0.44308418037850283,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004987592995536093,
      "loss": 5.7672,
      "mean_token_accuracy": 0.1658184953033924,
      "num_tokens": 8814869.0,
      "step": 3945
    },
    {
      "entropy": 5.73998441696167,
      "epoch": 0.4436457572864604,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004987550855329951,
      "loss": 5.7099,
      "mean_token_accuracy": 0.17231065332889556,
      "num_tokens": 8825264.0,
      "step": 3950
    },
    {
      "entropy": 5.8049551486969,
      "epoch": 0.44420733419441794,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004987508643879322,
      "loss": 5.787,
      "mean_token_accuracy": 0.16581701785326003,
      "num_tokens": 8837128.0,
      "step": 3955
    },
    {
      "entropy": 5.801353025436401,
      "epoch": 0.4447689111023755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004987466361185549,
      "loss": 5.7861,
      "mean_token_accuracy": 0.15883654057979585,
      "num_tokens": 8848587.0,
      "step": 3960
    },
    {
      "entropy": 5.848768043518066,
      "epoch": 0.44533048801033304,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004987424007249979,
      "loss": 5.8546,
      "mean_token_accuracy": 0.16280081868171692,
      "num_tokens": 8859219.0,
      "step": 3965
    },
    {
      "entropy": 5.788423871994018,
      "epoch": 0.4458920649182906,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004987381582073962,
      "loss": 5.7406,
      "mean_token_accuracy": 0.17112360298633575,
      "num_tokens": 8868780.0,
      "step": 3970
    },
    {
      "entropy": 5.834964323043823,
      "epoch": 0.4464536418262481,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004987339085658848,
      "loss": 5.7172,
      "mean_token_accuracy": 0.1637463465332985,
      "num_tokens": 8880903.0,
      "step": 3975
    },
    {
      "entropy": 5.660829162597656,
      "epoch": 0.44701521873420563,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004987296518005988,
      "loss": 5.5606,
      "mean_token_accuracy": 0.17245692163705825,
      "num_tokens": 8890335.0,
      "step": 3980
    },
    {
      "entropy": 5.665368318557739,
      "epoch": 0.4475767956421632,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498725387911674,
      "loss": 5.7033,
      "mean_token_accuracy": 0.16330161690711975,
      "num_tokens": 8904104.0,
      "step": 3985
    },
    {
      "entropy": 5.9002443790435795,
      "epoch": 0.44813837255012073,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004987211168992461,
      "loss": 5.7955,
      "mean_token_accuracy": 0.1588960111141205,
      "num_tokens": 8915118.0,
      "step": 3990
    },
    {
      "entropy": 5.748874807357788,
      "epoch": 0.4486999494580783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498716838763451,
      "loss": 5.7518,
      "mean_token_accuracy": 0.16183157116174698,
      "num_tokens": 8927450.0,
      "step": 3995
    },
    {
      "entropy": 5.855266857147217,
      "epoch": 0.44926152636603583,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000498712553504425,
      "loss": 5.8327,
      "mean_token_accuracy": 0.16530437469482423,
      "num_tokens": 8939902.0,
      "step": 4000
    },
    {
      "entropy": 5.803925609588623,
      "epoch": 0.4498231032739934,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004987082611223046,
      "loss": 5.6966,
      "mean_token_accuracy": 0.16004012376070023,
      "num_tokens": 8951736.0,
      "step": 4005
    },
    {
      "entropy": 5.824748182296753,
      "epoch": 0.45038468018195094,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004987039616172265,
      "loss": 5.7918,
      "mean_token_accuracy": 0.16543588787317276,
      "num_tokens": 8963283.0,
      "step": 4010
    },
    {
      "entropy": 5.746975517272949,
      "epoch": 0.4509462570899085,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986996549893273,
      "loss": 5.6875,
      "mean_token_accuracy": 0.1728332132101059,
      "num_tokens": 8974323.0,
      "step": 4015
    },
    {
      "entropy": 5.742187404632569,
      "epoch": 0.451507833997866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004986953412387445,
      "loss": 5.7779,
      "mean_token_accuracy": 0.1621204733848572,
      "num_tokens": 8986687.0,
      "step": 4020
    },
    {
      "entropy": 5.705374097824096,
      "epoch": 0.45206941090582353,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004986910203656151,
      "loss": 5.6662,
      "mean_token_accuracy": 0.17241707146167756,
      "num_tokens": 8997734.0,
      "step": 4025
    },
    {
      "entropy": 5.743109798431396,
      "epoch": 0.4526309878137811,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004986866923700769,
      "loss": 5.79,
      "mean_token_accuracy": 0.15861670374870301,
      "num_tokens": 9008743.0,
      "step": 4030
    },
    {
      "entropy": 5.864490699768067,
      "epoch": 0.45319256472173863,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004986823572522677,
      "loss": 5.649,
      "mean_token_accuracy": 0.167484150826931,
      "num_tokens": 9019495.0,
      "step": 4035
    },
    {
      "entropy": 5.750733947753906,
      "epoch": 0.4537541416296962,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004986780150123254,
      "loss": 5.7626,
      "mean_token_accuracy": 0.16347892284393312,
      "num_tokens": 9029546.0,
      "step": 4040
    },
    {
      "entropy": 5.62654070854187,
      "epoch": 0.45431571853765373,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004986736656503885,
      "loss": 5.5203,
      "mean_token_accuracy": 0.17364439517259597,
      "num_tokens": 9040393.0,
      "step": 4045
    },
    {
      "entropy": 5.732595634460449,
      "epoch": 0.4548772954456113,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004986693091665952,
      "loss": 5.7232,
      "mean_token_accuracy": 0.16511399149894715,
      "num_tokens": 9050840.0,
      "step": 4050
    },
    {
      "entropy": 5.76080846786499,
      "epoch": 0.45543887235356884,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004986649455610845,
      "loss": 5.6551,
      "mean_token_accuracy": 0.16657507717609404,
      "num_tokens": 9061615.0,
      "step": 4055
    },
    {
      "entropy": 5.688541269302368,
      "epoch": 0.4560004492615264,
      "grad_norm": 1.328125,
      "learning_rate": 0.000498660574833995,
      "loss": 5.6307,
      "mean_token_accuracy": 0.17303157597780228,
      "num_tokens": 9071902.0,
      "step": 4060
    },
    {
      "entropy": 5.747698211669922,
      "epoch": 0.45656202616948394,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004986561969854661,
      "loss": 5.7872,
      "mean_token_accuracy": 0.16532022655010223,
      "num_tokens": 9083613.0,
      "step": 4065
    },
    {
      "entropy": 5.754197359085083,
      "epoch": 0.45712360307744143,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004986518120156373,
      "loss": 5.6888,
      "mean_token_accuracy": 0.16673017293214798,
      "num_tokens": 9095678.0,
      "step": 4070
    },
    {
      "entropy": 5.761954212188721,
      "epoch": 0.457685179985399,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000498647419924648,
      "loss": 5.696,
      "mean_token_accuracy": 0.17291476875543593,
      "num_tokens": 9106637.0,
      "step": 4075
    },
    {
      "entropy": 5.667988586425781,
      "epoch": 0.45824675689335653,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000498643020712638,
      "loss": 5.6597,
      "mean_token_accuracy": 0.17288853377103805,
      "num_tokens": 9117343.0,
      "step": 4080
    },
    {
      "entropy": 5.798785257339477,
      "epoch": 0.4588083338013141,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986386143797475,
      "loss": 5.7887,
      "mean_token_accuracy": 0.15990065932273864,
      "num_tokens": 9129315.0,
      "step": 4085
    },
    {
      "entropy": 5.7697852611541744,
      "epoch": 0.45936991070927163,
      "grad_norm": 1.25,
      "learning_rate": 0.0004986342009261169,
      "loss": 5.6784,
      "mean_token_accuracy": 0.16411612257361413,
      "num_tokens": 9139701.0,
      "step": 4090
    },
    {
      "entropy": 5.79812912940979,
      "epoch": 0.4599314876172292,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004986297803518865,
      "loss": 5.8133,
      "mean_token_accuracy": 0.16509659737348556,
      "num_tokens": 9152334.0,
      "step": 4095
    },
    {
      "entropy": 5.762947416305542,
      "epoch": 0.46049306452518673,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004986253526571971,
      "loss": 5.6195,
      "mean_token_accuracy": 0.17192814201116563,
      "num_tokens": 9163277.0,
      "step": 4100
    },
    {
      "entropy": 5.772461366653443,
      "epoch": 0.4610546414331443,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986209178421899,
      "loss": 5.802,
      "mean_token_accuracy": 0.1619513064622879,
      "num_tokens": 9174909.0,
      "step": 4105
    },
    {
      "entropy": 5.647283840179443,
      "epoch": 0.46161621834110184,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000498616475907006,
      "loss": 5.5928,
      "mean_token_accuracy": 0.16549195647239684,
      "num_tokens": 9186058.0,
      "step": 4110
    },
    {
      "entropy": 5.73241286277771,
      "epoch": 0.46217779524905933,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004986120268517865,
      "loss": 5.7547,
      "mean_token_accuracy": 0.16822669506072999,
      "num_tokens": 9196919.0,
      "step": 4115
    },
    {
      "entropy": 5.772916507720947,
      "epoch": 0.4627393721570169,
      "grad_norm": 1.25,
      "learning_rate": 0.0004986075706766736,
      "loss": 5.7294,
      "mean_token_accuracy": 0.16655745655298232,
      "num_tokens": 9208452.0,
      "step": 4120
    },
    {
      "entropy": 5.783675956726074,
      "epoch": 0.46330094906497443,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004986031073818088,
      "loss": 5.7842,
      "mean_token_accuracy": 0.1649884581565857,
      "num_tokens": 9219605.0,
      "step": 4125
    },
    {
      "entropy": 5.74765625,
      "epoch": 0.463862525972932,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004985986369673343,
      "loss": 5.6159,
      "mean_token_accuracy": 0.17641097754240037,
      "num_tokens": 9230491.0,
      "step": 4130
    },
    {
      "entropy": 5.717433166503906,
      "epoch": 0.46442410288088953,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004985941594333926,
      "loss": 5.7567,
      "mean_token_accuracy": 0.16423618644475937,
      "num_tokens": 9241768.0,
      "step": 4135
    },
    {
      "entropy": 5.800970554351807,
      "epoch": 0.4649856797888471,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498589674780126,
      "loss": 5.7934,
      "mean_token_accuracy": 0.16235399544239043,
      "num_tokens": 9253779.0,
      "step": 4140
    },
    {
      "entropy": 5.94872088432312,
      "epoch": 0.46554725669680463,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004985851830076775,
      "loss": 5.6916,
      "mean_token_accuracy": 0.17534759044647216,
      "num_tokens": 9264567.0,
      "step": 4145
    },
    {
      "entropy": 5.591338491439819,
      "epoch": 0.4661088336047622,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004985806841161901,
      "loss": 5.6529,
      "mean_token_accuracy": 0.17427073419094086,
      "num_tokens": 9275960.0,
      "step": 4150
    },
    {
      "entropy": 5.708569431304932,
      "epoch": 0.46667041051271974,
      "grad_norm": 1.390625,
      "learning_rate": 0.000498576178105807,
      "loss": 5.7891,
      "mean_token_accuracy": 0.17374880164861678,
      "num_tokens": 9288305.0,
      "step": 4155
    },
    {
      "entropy": 5.802067136764526,
      "epoch": 0.4672319874206773,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004985716649766717,
      "loss": 5.6165,
      "mean_token_accuracy": 0.17416664063930512,
      "num_tokens": 9299579.0,
      "step": 4160
    },
    {
      "entropy": 5.667328643798828,
      "epoch": 0.4677935643286348,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004985671447289277,
      "loss": 5.6468,
      "mean_token_accuracy": 0.16909621357917787,
      "num_tokens": 9311221.0,
      "step": 4165
    },
    {
      "entropy": 5.732243394851684,
      "epoch": 0.46835514123659233,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004985626173627194,
      "loss": 5.7524,
      "mean_token_accuracy": 0.16208277642726898,
      "num_tokens": 9322488.0,
      "step": 4170
    },
    {
      "entropy": 5.74122986793518,
      "epoch": 0.4689167181445499,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004985580828781904,
      "loss": 5.7099,
      "mean_token_accuracy": 0.16819628775119783,
      "num_tokens": 9333672.0,
      "step": 4175
    },
    {
      "entropy": 5.8608152866363525,
      "epoch": 0.46947829505250743,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004985535412754855,
      "loss": 5.7418,
      "mean_token_accuracy": 0.16545120030641555,
      "num_tokens": 9344449.0,
      "step": 4180
    },
    {
      "entropy": 5.826900577545166,
      "epoch": 0.470039871960465,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000498548992554749,
      "loss": 5.8074,
      "mean_token_accuracy": 0.1609753578901291,
      "num_tokens": 9355279.0,
      "step": 4185
    },
    {
      "entropy": 5.751828336715699,
      "epoch": 0.47060144886842253,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985444367161259,
      "loss": 5.7103,
      "mean_token_accuracy": 0.16859401911497116,
      "num_tokens": 9366942.0,
      "step": 4190
    },
    {
      "entropy": 5.685658740997314,
      "epoch": 0.4711630257763801,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985398737597614,
      "loss": 5.6355,
      "mean_token_accuracy": 0.1736157312989235,
      "num_tokens": 9378129.0,
      "step": 4195
    },
    {
      "entropy": 5.801950359344483,
      "epoch": 0.47172460268433763,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004985353036858004,
      "loss": 5.7287,
      "mean_token_accuracy": 0.16660770624876023,
      "num_tokens": 9388757.0,
      "step": 4200
    },
    {
      "entropy": 5.744151067733765,
      "epoch": 0.4722861795922952,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004985307264943887,
      "loss": 5.6745,
      "mean_token_accuracy": 0.16941459774971007,
      "num_tokens": 9399112.0,
      "step": 4205
    },
    {
      "entropy": 5.7844664573669435,
      "epoch": 0.47284775650025274,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985261421856721,
      "loss": 5.7691,
      "mean_token_accuracy": 0.16700249612331391,
      "num_tokens": 9410156.0,
      "step": 4210
    },
    {
      "entropy": 5.833684635162354,
      "epoch": 0.47340933340821023,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004985215507597963,
      "loss": 5.7515,
      "mean_token_accuracy": 0.17010293602943422,
      "num_tokens": 9420402.0,
      "step": 4215
    },
    {
      "entropy": 5.729236125946045,
      "epoch": 0.4739709103161678,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004985169522169076,
      "loss": 5.7532,
      "mean_token_accuracy": 0.1657623678445816,
      "num_tokens": 9431159.0,
      "step": 4220
    },
    {
      "entropy": 5.701308298110962,
      "epoch": 0.47453248722412533,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985123465571526,
      "loss": 5.6592,
      "mean_token_accuracy": 0.17336241155862808,
      "num_tokens": 9441737.0,
      "step": 4225
    },
    {
      "entropy": 5.6865349292755125,
      "epoch": 0.4750940641320829,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985077337806777,
      "loss": 5.6196,
      "mean_token_accuracy": 0.1779865562915802,
      "num_tokens": 9452756.0,
      "step": 4230
    },
    {
      "entropy": 5.739285612106324,
      "epoch": 0.47565564104004043,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985031138876299,
      "loss": 5.6878,
      "mean_token_accuracy": 0.1680801346898079,
      "num_tokens": 9464231.0,
      "step": 4235
    },
    {
      "entropy": 5.764004802703857,
      "epoch": 0.476217217947998,
      "grad_norm": 1.125,
      "learning_rate": 0.0004984984868781563,
      "loss": 5.7067,
      "mean_token_accuracy": 0.1760090947151184,
      "num_tokens": 9475648.0,
      "step": 4240
    },
    {
      "entropy": 5.744698429107666,
      "epoch": 0.47677879485595553,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984938527524041,
      "loss": 5.751,
      "mean_token_accuracy": 0.16820831894874572,
      "num_tokens": 9486594.0,
      "step": 4245
    },
    {
      "entropy": 5.672558355331421,
      "epoch": 0.4773403717639131,
      "grad_norm": 1.1875,
      "learning_rate": 0.000498489211510521,
      "loss": 5.6621,
      "mean_token_accuracy": 0.17208858728408813,
      "num_tokens": 9497085.0,
      "step": 4250
    },
    {
      "entropy": 5.784761142730713,
      "epoch": 0.47790194867187064,
      "grad_norm": 1.125,
      "learning_rate": 0.0004984845631526547,
      "loss": 5.7477,
      "mean_token_accuracy": 0.16722307056188584,
      "num_tokens": 9509765.0,
      "step": 4255
    },
    {
      "entropy": 5.725100564956665,
      "epoch": 0.47846352557982813,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004984799076789534,
      "loss": 5.5731,
      "mean_token_accuracy": 0.16857376992702483,
      "num_tokens": 9521341.0,
      "step": 4260
    },
    {
      "entropy": 5.630601024627685,
      "epoch": 0.4790251024877857,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004984752450895649,
      "loss": 5.6929,
      "mean_token_accuracy": 0.1659396380186081,
      "num_tokens": 9533196.0,
      "step": 4265
    },
    {
      "entropy": 5.8102864742279055,
      "epoch": 0.47958667939574323,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498470575384638,
      "loss": 5.8286,
      "mean_token_accuracy": 0.15910447090864183,
      "num_tokens": 9543687.0,
      "step": 4270
    },
    {
      "entropy": 5.903976631164551,
      "epoch": 0.4801482563037008,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004984658985643214,
      "loss": 5.7156,
      "mean_token_accuracy": 0.16543746441602708,
      "num_tokens": 9555320.0,
      "step": 4275
    },
    {
      "entropy": 5.636949110031128,
      "epoch": 0.48070983321165833,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004984612146287638,
      "loss": 5.6659,
      "mean_token_accuracy": 0.16941267549991607,
      "num_tokens": 9566347.0,
      "step": 4280
    },
    {
      "entropy": 5.636159801483155,
      "epoch": 0.4812714101196159,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984565235781147,
      "loss": 5.6669,
      "mean_token_accuracy": 0.16980277746915817,
      "num_tokens": 9577111.0,
      "step": 4285
    },
    {
      "entropy": 5.761268234252929,
      "epoch": 0.48183298702757343,
      "grad_norm": 1.5,
      "learning_rate": 0.000498451825412523,
      "loss": 5.671,
      "mean_token_accuracy": 0.16593389809131623,
      "num_tokens": 9586612.0,
      "step": 4290
    },
    {
      "entropy": 5.774903297424316,
      "epoch": 0.482394563935531,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984471201321386,
      "loss": 5.6786,
      "mean_token_accuracy": 0.17062007635831833,
      "num_tokens": 9597347.0,
      "step": 4295
    },
    {
      "entropy": 5.674345445632935,
      "epoch": 0.48295614084348854,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004984424077371112,
      "loss": 5.6524,
      "mean_token_accuracy": 0.17706192433834075,
      "num_tokens": 9608507.0,
      "step": 4300
    },
    {
      "entropy": 5.708226156234741,
      "epoch": 0.4835177177514461,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004984376882275909,
      "loss": 5.7082,
      "mean_token_accuracy": 0.17436117678880692,
      "num_tokens": 9619449.0,
      "step": 4305
    },
    {
      "entropy": 5.680789613723755,
      "epoch": 0.4840792946594036,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004984329616037279,
      "loss": 5.6346,
      "mean_token_accuracy": 0.16903165727853775,
      "num_tokens": 9630613.0,
      "step": 4310
    },
    {
      "entropy": 5.761481714248657,
      "epoch": 0.48464087156736113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984282278656727,
      "loss": 5.7086,
      "mean_token_accuracy": 0.16750625967979432,
      "num_tokens": 9642149.0,
      "step": 4315
    },
    {
      "entropy": 5.755221033096314,
      "epoch": 0.4852024484753187,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984234870135762,
      "loss": 5.6496,
      "mean_token_accuracy": 0.17484654635190963,
      "num_tokens": 9652960.0,
      "step": 4320
    },
    {
      "entropy": 5.68124532699585,
      "epoch": 0.48576402538327623,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004984187390475891,
      "loss": 5.6417,
      "mean_token_accuracy": 0.17295996993780136,
      "num_tokens": 9662676.0,
      "step": 4325
    },
    {
      "entropy": 5.838757276535034,
      "epoch": 0.4863256022912338,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004984139839678628,
      "loss": 5.7493,
      "mean_token_accuracy": 0.16618652939796447,
      "num_tokens": 9674138.0,
      "step": 4330
    },
    {
      "entropy": 5.714073657989502,
      "epoch": 0.48688717919919133,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004984092217745486,
      "loss": 5.7113,
      "mean_token_accuracy": 0.1709674835205078,
      "num_tokens": 9685397.0,
      "step": 4335
    },
    {
      "entropy": 5.790579319000244,
      "epoch": 0.4874487561071489,
      "grad_norm": 1.125,
      "learning_rate": 0.0004984044524677981,
      "loss": 5.8231,
      "mean_token_accuracy": 0.1605924040079117,
      "num_tokens": 9697266.0,
      "step": 4340
    },
    {
      "entropy": 5.828333997726441,
      "epoch": 0.48801033301510643,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004983996760477632,
      "loss": 5.6552,
      "mean_token_accuracy": 0.16677314341068267,
      "num_tokens": 9707967.0,
      "step": 4345
    },
    {
      "entropy": 5.62044587135315,
      "epoch": 0.488571909923064,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498394892514596,
      "loss": 5.6341,
      "mean_token_accuracy": 0.17084338515996933,
      "num_tokens": 9718162.0,
      "step": 4350
    },
    {
      "entropy": 5.717535734176636,
      "epoch": 0.48913348683102154,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004983901018684486,
      "loss": 5.6886,
      "mean_token_accuracy": 0.16895658969879152,
      "num_tokens": 9729439.0,
      "step": 4355
    },
    {
      "entropy": 5.775072956085205,
      "epoch": 0.48969506373897903,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983853041094739,
      "loss": 5.7112,
      "mean_token_accuracy": 0.16845377534627914,
      "num_tokens": 9741383.0,
      "step": 4360
    },
    {
      "entropy": 5.712487268447876,
      "epoch": 0.4902566406469366,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004983804992378243,
      "loss": 5.7325,
      "mean_token_accuracy": 0.17263987958431243,
      "num_tokens": 9753050.0,
      "step": 4365
    },
    {
      "entropy": 5.681834554672241,
      "epoch": 0.49081821755489413,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004983756872536532,
      "loss": 5.6141,
      "mean_token_accuracy": 0.1744478538632393,
      "num_tokens": 9763715.0,
      "step": 4370
    },
    {
      "entropy": 5.699193477630615,
      "epoch": 0.4913797944628517,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004983708681571134,
      "loss": 5.5872,
      "mean_token_accuracy": 0.1801524966955185,
      "num_tokens": 9774145.0,
      "step": 4375
    },
    {
      "entropy": 5.711805200576782,
      "epoch": 0.49194137137080923,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004983660419483586,
      "loss": 5.6484,
      "mean_token_accuracy": 0.1740181788802147,
      "num_tokens": 9784703.0,
      "step": 4380
    },
    {
      "entropy": 5.7284644603729244,
      "epoch": 0.4925029482787668,
      "grad_norm": 1.125,
      "learning_rate": 0.0004983612086275424,
      "loss": 5.6657,
      "mean_token_accuracy": 0.16666370034217834,
      "num_tokens": 9795426.0,
      "step": 4385
    },
    {
      "entropy": 5.744356107711792,
      "epoch": 0.49306452518672433,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004983563681948187,
      "loss": 5.7488,
      "mean_token_accuracy": 0.16563860327005386,
      "num_tokens": 9806539.0,
      "step": 4390
    },
    {
      "entropy": 5.838592100143432,
      "epoch": 0.4936261020946819,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004983515206503416,
      "loss": 5.7354,
      "mean_token_accuracy": 0.16389266550540924,
      "num_tokens": 9816806.0,
      "step": 4395
    },
    {
      "entropy": 5.741393136978149,
      "epoch": 0.49418767900263944,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004983466659942655,
      "loss": 5.6239,
      "mean_token_accuracy": 0.1763622909784317,
      "num_tokens": 9828001.0,
      "step": 4400
    },
    {
      "entropy": 5.568348503112793,
      "epoch": 0.49474925591059693,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004983418042267449,
      "loss": 5.6675,
      "mean_token_accuracy": 0.17209520637989045,
      "num_tokens": 9838773.0,
      "step": 4405
    },
    {
      "entropy": 5.723768329620361,
      "epoch": 0.4953108328185545,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983369353479345,
      "loss": 5.664,
      "mean_token_accuracy": 0.17434099316596985,
      "num_tokens": 9848680.0,
      "step": 4410
    },
    {
      "entropy": 5.700279903411865,
      "epoch": 0.49587240972651203,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004983320593579895,
      "loss": 5.6349,
      "mean_token_accuracy": 0.16858525723218917,
      "num_tokens": 9859168.0,
      "step": 4415
    },
    {
      "entropy": 5.66675705909729,
      "epoch": 0.4964339866344696,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004983271762570653,
      "loss": 5.671,
      "mean_token_accuracy": 0.17040987610816954,
      "num_tokens": 9869071.0,
      "step": 4420
    },
    {
      "entropy": 5.750021266937256,
      "epoch": 0.49699556354242713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498322286045317,
      "loss": 5.6767,
      "mean_token_accuracy": 0.17082203477621077,
      "num_tokens": 9880208.0,
      "step": 4425
    },
    {
      "entropy": 5.7331212043762205,
      "epoch": 0.4975571404503847,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004983173887229006,
      "loss": 5.7172,
      "mean_token_accuracy": 0.17034917324781418,
      "num_tokens": 9891192.0,
      "step": 4430
    },
    {
      "entropy": 5.790711212158203,
      "epoch": 0.49811871735834223,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004983124842899719,
      "loss": 5.7379,
      "mean_token_accuracy": 0.16513569355010987,
      "num_tokens": 9903038.0,
      "step": 4435
    },
    {
      "entropy": 5.830787420272827,
      "epoch": 0.4986802942662998,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004983075727466871,
      "loss": 5.7921,
      "mean_token_accuracy": 0.1677357569336891,
      "num_tokens": 9916104.0,
      "step": 4440
    },
    {
      "entropy": 5.6895956039428714,
      "epoch": 0.49924187117425733,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004983026540932026,
      "loss": 5.6103,
      "mean_token_accuracy": 0.171831975877285,
      "num_tokens": 9927094.0,
      "step": 4445
    },
    {
      "entropy": 5.784496068954468,
      "epoch": 0.4998034480822149,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004982977283296749,
      "loss": 5.8228,
      "mean_token_accuracy": 0.1600668027997017,
      "num_tokens": 9937964.0,
      "step": 4450
    },
    {
      "entropy": 5.8589826107025145,
      "epoch": 0.5003650249901724,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000498292795456261,
      "loss": 5.7841,
      "mean_token_accuracy": 0.16068200916051864,
      "num_tokens": 9949238.0,
      "step": 4455
    },
    {
      "entropy": 5.740886306762695,
      "epoch": 0.50092660189813,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004982878554731178,
      "loss": 5.6144,
      "mean_token_accuracy": 0.17186491191387177,
      "num_tokens": 9960125.0,
      "step": 4460
    },
    {
      "entropy": 5.579822969436646,
      "epoch": 0.5014881788060875,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004982829083804028,
      "loss": 5.6379,
      "mean_token_accuracy": 0.1658797100186348,
      "num_tokens": 9971130.0,
      "step": 4465
    },
    {
      "entropy": 5.683795070648193,
      "epoch": 0.5020497557140451,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004982779541782733,
      "loss": 5.5301,
      "mean_token_accuracy": 0.17982134222984314,
      "num_tokens": 9981516.0,
      "step": 4470
    },
    {
      "entropy": 5.691908359527588,
      "epoch": 0.5026113326220026,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004982729928668872,
      "loss": 5.765,
      "mean_token_accuracy": 0.16426052749156952,
      "num_tokens": 9992360.0,
      "step": 4475
    },
    {
      "entropy": 5.844257497787476,
      "epoch": 0.5031729095299601,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004982680244464024,
      "loss": 5.865,
      "mean_token_accuracy": 0.16424636989831926,
      "num_tokens": 10003410.0,
      "step": 4480
    },
    {
      "entropy": 5.831128740310669,
      "epoch": 0.5037344864379176,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004982630489169771,
      "loss": 5.6665,
      "mean_token_accuracy": 0.1697389677166939,
      "num_tokens": 10014974.0,
      "step": 4485
    },
    {
      "entropy": 5.730645370483399,
      "epoch": 0.5042960633458752,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004982580662787697,
      "loss": 5.7355,
      "mean_token_accuracy": 0.1644341304898262,
      "num_tokens": 10026702.0,
      "step": 4490
    },
    {
      "entropy": 5.7113964557647705,
      "epoch": 0.5048576402538327,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004982530765319389,
      "loss": 5.7192,
      "mean_token_accuracy": 0.1672709807753563,
      "num_tokens": 10039318.0,
      "step": 4495
    },
    {
      "entropy": 5.741805553436279,
      "epoch": 0.5054192171617903,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004982480796766436,
      "loss": 5.6632,
      "mean_token_accuracy": 0.1635326474905014,
      "num_tokens": 10049726.0,
      "step": 4500
    },
    {
      "entropy": 5.691263341903687,
      "epoch": 0.5059807940697478,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004982430757130429,
      "loss": 5.5585,
      "mean_token_accuracy": 0.1718662425875664,
      "num_tokens": 10059663.0,
      "step": 4505
    },
    {
      "entropy": 5.583151340484619,
      "epoch": 0.5065423709777054,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004982380646412959,
      "loss": 5.6013,
      "mean_token_accuracy": 0.16774161756038666,
      "num_tokens": 10071195.0,
      "step": 4510
    },
    {
      "entropy": 5.827329444885254,
      "epoch": 0.5071039478856629,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004982330464615625,
      "loss": 5.8071,
      "mean_token_accuracy": 0.16533977538347244,
      "num_tokens": 10082014.0,
      "step": 4515
    },
    {
      "entropy": 5.811189365386963,
      "epoch": 0.5076655247936205,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004982280211740023,
      "loss": 5.6143,
      "mean_token_accuracy": 0.17575001269578933,
      "num_tokens": 10093156.0,
      "step": 4520
    },
    {
      "entropy": 5.631749963760376,
      "epoch": 0.508227101701578,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004982229887787753,
      "loss": 5.7599,
      "mean_token_accuracy": 0.1561910793185234,
      "num_tokens": 10104993.0,
      "step": 4525
    },
    {
      "entropy": 5.7125547409057615,
      "epoch": 0.5087886786095356,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004982179492760418,
      "loss": 5.5668,
      "mean_token_accuracy": 0.18113714456558228,
      "num_tokens": 10114590.0,
      "step": 4530
    },
    {
      "entropy": 5.67285680770874,
      "epoch": 0.5093502555174931,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004982129026659621,
      "loss": 5.5738,
      "mean_token_accuracy": 0.17499979734420776,
      "num_tokens": 10125849.0,
      "step": 4535
    },
    {
      "entropy": 5.626177024841309,
      "epoch": 0.5099118324254507,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004982078489486972,
      "loss": 5.5941,
      "mean_token_accuracy": 0.17779029458761214,
      "num_tokens": 10135493.0,
      "step": 4540
    },
    {
      "entropy": 5.7630730152130125,
      "epoch": 0.5104734093334082,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004982027881244076,
      "loss": 5.7775,
      "mean_token_accuracy": 0.1641564130783081,
      "num_tokens": 10146987.0,
      "step": 4545
    },
    {
      "entropy": 5.765534067153931,
      "epoch": 0.5110349862413658,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004981977201932549,
      "loss": 5.7086,
      "mean_token_accuracy": 0.17529913783073425,
      "num_tokens": 10157823.0,
      "step": 4550
    },
    {
      "entropy": 5.639018440246582,
      "epoch": 0.5115965631493233,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004981926451554001,
      "loss": 5.6758,
      "mean_token_accuracy": 0.17088306099176406,
      "num_tokens": 10169305.0,
      "step": 4555
    },
    {
      "entropy": 5.824262237548828,
      "epoch": 0.5121581400572809,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004981875630110048,
      "loss": 5.7553,
      "mean_token_accuracy": 0.1605099081993103,
      "num_tokens": 10180346.0,
      "step": 4560
    },
    {
      "entropy": 5.753881549835205,
      "epoch": 0.5127197169652384,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004981824737602311,
      "loss": 5.6425,
      "mean_token_accuracy": 0.17663276344537734,
      "num_tokens": 10190870.0,
      "step": 4565
    },
    {
      "entropy": 5.745822525024414,
      "epoch": 0.513281293873196,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004981773774032409,
      "loss": 5.7113,
      "mean_token_accuracy": 0.15780248045921325,
      "num_tokens": 10202408.0,
      "step": 4570
    },
    {
      "entropy": 5.637575244903564,
      "epoch": 0.5138428707811534,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004981722739401962,
      "loss": 5.5405,
      "mean_token_accuracy": 0.17590135782957078,
      "num_tokens": 10213367.0,
      "step": 4575
    },
    {
      "entropy": 5.701434135437012,
      "epoch": 0.514404447689111,
      "grad_norm": 1.3125,
      "learning_rate": 0.00049816716337126,
      "loss": 5.6765,
      "mean_token_accuracy": 0.16563864946365356,
      "num_tokens": 10223865.0,
      "step": 4580
    },
    {
      "entropy": 5.599912309646607,
      "epoch": 0.5149660245970685,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004981620456965946,
      "loss": 5.4468,
      "mean_token_accuracy": 0.17510706782341004,
      "num_tokens": 10235268.0,
      "step": 4585
    },
    {
      "entropy": 5.6726030826568605,
      "epoch": 0.5155276015050261,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004981569209163633,
      "loss": 5.7124,
      "mean_token_accuracy": 0.16576266586780547,
      "num_tokens": 10247700.0,
      "step": 4590
    },
    {
      "entropy": 5.761053800582886,
      "epoch": 0.5160891784129836,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004981517890307288,
      "loss": 5.76,
      "mean_token_accuracy": 0.16685858815908433,
      "num_tokens": 10258553.0,
      "step": 4595
    },
    {
      "entropy": 5.758320093154907,
      "epoch": 0.5166507553209412,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000498146650039855,
      "loss": 5.6481,
      "mean_token_accuracy": 0.17269993275403978,
      "num_tokens": 10268496.0,
      "step": 4600
    },
    {
      "entropy": 5.774955701828003,
      "epoch": 0.5172123322288987,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004981415039439054,
      "loss": 5.7253,
      "mean_token_accuracy": 0.16718546748161317,
      "num_tokens": 10278803.0,
      "step": 4605
    },
    {
      "entropy": 5.711667108535766,
      "epoch": 0.5177739091368563,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004981363507430436,
      "loss": 5.5835,
      "mean_token_accuracy": 0.1724689558148384,
      "num_tokens": 10289805.0,
      "step": 4610
    },
    {
      "entropy": 5.688625192642212,
      "epoch": 0.5183354860448138,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004981311904374339,
      "loss": 5.7046,
      "mean_token_accuracy": 0.174521566927433,
      "num_tokens": 10300295.0,
      "step": 4615
    },
    {
      "entropy": 5.841267871856689,
      "epoch": 0.5188970629527714,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981260230272405,
      "loss": 5.7889,
      "mean_token_accuracy": 0.16799403876066207,
      "num_tokens": 10312175.0,
      "step": 4620
    },
    {
      "entropy": 5.63864483833313,
      "epoch": 0.5194586398607289,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004981208485126281,
      "loss": 5.6779,
      "mean_token_accuracy": 0.1660699799656868,
      "num_tokens": 10322888.0,
      "step": 4625
    },
    {
      "entropy": 5.693124485015869,
      "epoch": 0.5200202167686865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004981156668937613,
      "loss": 5.6516,
      "mean_token_accuracy": 0.17714009433984756,
      "num_tokens": 10333305.0,
      "step": 4630
    },
    {
      "entropy": 5.757165384292603,
      "epoch": 0.520581793676644,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004981104781708051,
      "loss": 5.6988,
      "mean_token_accuracy": 0.16862452477216722,
      "num_tokens": 10344190.0,
      "step": 4635
    },
    {
      "entropy": 5.797582530975342,
      "epoch": 0.5211433705846016,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981052823439247,
      "loss": 5.7742,
      "mean_token_accuracy": 0.16717695221304893,
      "num_tokens": 10356271.0,
      "step": 4640
    },
    {
      "entropy": 5.836018705368042,
      "epoch": 0.5217049474925591,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004981000794132858,
      "loss": 5.8366,
      "mean_token_accuracy": 0.15832895487546922,
      "num_tokens": 10368756.0,
      "step": 4645
    },
    {
      "entropy": 5.814976835250855,
      "epoch": 0.5222665244005167,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004980948693790536,
      "loss": 5.6962,
      "mean_token_accuracy": 0.1687784492969513,
      "num_tokens": 10380029.0,
      "step": 4650
    },
    {
      "entropy": 5.691782474517822,
      "epoch": 0.5228281013084742,
      "grad_norm": 1.25,
      "learning_rate": 0.0004980896522413943,
      "loss": 5.6803,
      "mean_token_accuracy": 0.16841885596513748,
      "num_tokens": 10390358.0,
      "step": 4655
    },
    {
      "entropy": 5.686275625228882,
      "epoch": 0.5233896782164318,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004980844280004739,
      "loss": 5.6551,
      "mean_token_accuracy": 0.1711376890540123,
      "num_tokens": 10401331.0,
      "step": 4660
    },
    {
      "entropy": 5.753196573257446,
      "epoch": 0.5239512551243893,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004980791966564589,
      "loss": 5.6783,
      "mean_token_accuracy": 0.16949976235628128,
      "num_tokens": 10412755.0,
      "step": 4665
    },
    {
      "entropy": 5.692148733139038,
      "epoch": 0.5245128320323468,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004980739582095158,
      "loss": 5.6194,
      "mean_token_accuracy": 0.17993575930595399,
      "num_tokens": 10422698.0,
      "step": 4670
    },
    {
      "entropy": 5.739432430267334,
      "epoch": 0.5250744089403043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004980687126598111,
      "loss": 5.753,
      "mean_token_accuracy": 0.16443512141704558,
      "num_tokens": 10434467.0,
      "step": 4675
    },
    {
      "entropy": 5.691123342514038,
      "epoch": 0.5256359858482619,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004980634600075123,
      "loss": 5.6647,
      "mean_token_accuracy": 0.17190593779087066,
      "num_tokens": 10446007.0,
      "step": 4680
    },
    {
      "entropy": 5.758034658432007,
      "epoch": 0.5261975627562194,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004980582002527863,
      "loss": 5.6857,
      "mean_token_accuracy": 0.17210673540830612,
      "num_tokens": 10457538.0,
      "step": 4685
    },
    {
      "entropy": 5.737921142578125,
      "epoch": 0.526759139664177,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004980529333958008,
      "loss": 5.6438,
      "mean_token_accuracy": 0.17155391573905945,
      "num_tokens": 10468389.0,
      "step": 4690
    },
    {
      "entropy": 5.64755654335022,
      "epoch": 0.5273207165721345,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004980476594367233,
      "loss": 5.6167,
      "mean_token_accuracy": 0.166074176132679,
      "num_tokens": 10479032.0,
      "step": 4695
    },
    {
      "entropy": 5.7380961894989015,
      "epoch": 0.5278822934800921,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004980423783757219,
      "loss": 5.6203,
      "mean_token_accuracy": 0.1761274218559265,
      "num_tokens": 10490505.0,
      "step": 4700
    },
    {
      "entropy": 5.777146339416504,
      "epoch": 0.5284438703880496,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004980370902129646,
      "loss": 5.6841,
      "mean_token_accuracy": 0.17220447063446045,
      "num_tokens": 10501382.0,
      "step": 4705
    },
    {
      "entropy": 5.7597027778625485,
      "epoch": 0.5290054472960072,
      "grad_norm": 1.171875,
      "learning_rate": 0.00049803179494862,
      "loss": 5.6519,
      "mean_token_accuracy": 0.17078416049480438,
      "num_tokens": 10512727.0,
      "step": 4710
    },
    {
      "entropy": 5.627700519561768,
      "epoch": 0.5295670242039647,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004980264925828563,
      "loss": 5.6379,
      "mean_token_accuracy": 0.17503894567489625,
      "num_tokens": 10523383.0,
      "step": 4715
    },
    {
      "entropy": 5.723991012573242,
      "epoch": 0.5301286011119223,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004980211831158427,
      "loss": 5.6443,
      "mean_token_accuracy": 0.17036737948656083,
      "num_tokens": 10534416.0,
      "step": 4720
    },
    {
      "entropy": 5.713537502288818,
      "epoch": 0.5306901780198798,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004980158665477481,
      "loss": 5.6192,
      "mean_token_accuracy": 0.1723712608218193,
      "num_tokens": 10544291.0,
      "step": 4725
    },
    {
      "entropy": 5.703956174850464,
      "epoch": 0.5312517549278374,
      "grad_norm": 1.25,
      "learning_rate": 0.0004980105428787419,
      "loss": 5.585,
      "mean_token_accuracy": 0.1744757741689682,
      "num_tokens": 10555291.0,
      "step": 4730
    },
    {
      "entropy": 5.739058351516723,
      "epoch": 0.5318133318357949,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004980052121089934,
      "loss": 5.6839,
      "mean_token_accuracy": 0.1688266322016716,
      "num_tokens": 10565812.0,
      "step": 4735
    },
    {
      "entropy": 5.644025564193726,
      "epoch": 0.5323749087437525,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004979998742386725,
      "loss": 5.6093,
      "mean_token_accuracy": 0.17441638559103012,
      "num_tokens": 10577351.0,
      "step": 4740
    },
    {
      "entropy": 5.761340379714966,
      "epoch": 0.53293648565171,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004979945292679492,
      "loss": 5.6919,
      "mean_token_accuracy": 0.1682047963142395,
      "num_tokens": 10589236.0,
      "step": 4745
    },
    {
      "entropy": 5.650798368453979,
      "epoch": 0.5334980625596676,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004979891771969935,
      "loss": 5.6363,
      "mean_token_accuracy": 0.17235926836729049,
      "num_tokens": 10599926.0,
      "step": 4750
    },
    {
      "entropy": 5.7528852939605715,
      "epoch": 0.5340596394676251,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000497983818025976,
      "loss": 5.7496,
      "mean_token_accuracy": 0.16808155328035354,
      "num_tokens": 10611495.0,
      "step": 4755
    },
    {
      "entropy": 5.749897480010986,
      "epoch": 0.5346212163755827,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004979784517550671,
      "loss": 5.7608,
      "mean_token_accuracy": 0.1683972179889679,
      "num_tokens": 10622455.0,
      "step": 4760
    },
    {
      "entropy": 5.70879487991333,
      "epoch": 0.5351827932835401,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004979730783844379,
      "loss": 5.6031,
      "mean_token_accuracy": 0.17482960671186448,
      "num_tokens": 10634255.0,
      "step": 4765
    },
    {
      "entropy": 5.737195301055908,
      "epoch": 0.5357443701914977,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004979676979142595,
      "loss": 5.7169,
      "mean_token_accuracy": 0.1722305104136467,
      "num_tokens": 10644816.0,
      "step": 4770
    },
    {
      "entropy": 5.732215309143067,
      "epoch": 0.5363059470994552,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000497962310344703,
      "loss": 5.6459,
      "mean_token_accuracy": 0.17166779786348343,
      "num_tokens": 10655449.0,
      "step": 4775
    },
    {
      "entropy": 5.718864822387696,
      "epoch": 0.5368675240074128,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00049795691567594,
      "loss": 5.6553,
      "mean_token_accuracy": 0.16806258261203766,
      "num_tokens": 10666343.0,
      "step": 4780
    },
    {
      "entropy": 5.72328372001648,
      "epoch": 0.5374291009153703,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004979515139081426,
      "loss": 5.7064,
      "mean_token_accuracy": 0.16548898369073867,
      "num_tokens": 10678511.0,
      "step": 4785
    },
    {
      "entropy": 5.665549087524414,
      "epoch": 0.5379906778233279,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004979461050414822,
      "loss": 5.6909,
      "mean_token_accuracy": 0.17560625523328782,
      "num_tokens": 10690005.0,
      "step": 4790
    },
    {
      "entropy": 5.679067897796631,
      "epoch": 0.5385522547312854,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004979406890761315,
      "loss": 5.681,
      "mean_token_accuracy": 0.1686423435807228,
      "num_tokens": 10701062.0,
      "step": 4795
    },
    {
      "entropy": 5.821629285812378,
      "epoch": 0.539113831639243,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004979352660122629,
      "loss": 5.5993,
      "mean_token_accuracy": 0.1772200345993042,
      "num_tokens": 10711986.0,
      "step": 4800
    },
    {
      "entropy": 5.744456338882446,
      "epoch": 0.5396754085472005,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004979298358500488,
      "loss": 5.7093,
      "mean_token_accuracy": 0.16843186989426612,
      "num_tokens": 10722509.0,
      "step": 4805
    },
    {
      "entropy": 5.742955589294434,
      "epoch": 0.5402369854551581,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004979243985896624,
      "loss": 5.6577,
      "mean_token_accuracy": 0.16745564192533494,
      "num_tokens": 10733755.0,
      "step": 4810
    },
    {
      "entropy": 5.69589319229126,
      "epoch": 0.5407985623631156,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004979189542312765,
      "loss": 5.688,
      "mean_token_accuracy": 0.16879853308200837,
      "num_tokens": 10744356.0,
      "step": 4815
    },
    {
      "entropy": 5.710256290435791,
      "epoch": 0.5413601392710732,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004979135027750649,
      "loss": 5.5928,
      "mean_token_accuracy": 0.17196511775255202,
      "num_tokens": 10756474.0,
      "step": 4820
    },
    {
      "entropy": 5.702636575698852,
      "epoch": 0.5419217161790307,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004979080442212007,
      "loss": 5.6718,
      "mean_token_accuracy": 0.17200553864240647,
      "num_tokens": 10767551.0,
      "step": 4825
    },
    {
      "entropy": 5.7792284965515135,
      "epoch": 0.5424832930869883,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004979025785698581,
      "loss": 5.7624,
      "mean_token_accuracy": 0.16749392002820968,
      "num_tokens": 10778524.0,
      "step": 4830
    },
    {
      "entropy": 5.679055833816529,
      "epoch": 0.5430448699949458,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004978971058212108,
      "loss": 5.6711,
      "mean_token_accuracy": 0.17368635684251785,
      "num_tokens": 10789798.0,
      "step": 4835
    },
    {
      "entropy": 5.730114889144898,
      "epoch": 0.5436064469029034,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004978916259754331,
      "loss": 5.6202,
      "mean_token_accuracy": 0.1732513874769211,
      "num_tokens": 10800500.0,
      "step": 4840
    },
    {
      "entropy": 5.761911201477051,
      "epoch": 0.5441680238108609,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004978861390326998,
      "loss": 5.6492,
      "mean_token_accuracy": 0.17786261439323425,
      "num_tokens": 10812047.0,
      "step": 4845
    },
    {
      "entropy": 5.655855131149292,
      "epoch": 0.5447296007188185,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004978806449931854,
      "loss": 5.6656,
      "mean_token_accuracy": 0.1673392355442047,
      "num_tokens": 10823272.0,
      "step": 4850
    },
    {
      "entropy": 5.585034036636353,
      "epoch": 0.545291177626776,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004978751438570647,
      "loss": 5.5368,
      "mean_token_accuracy": 0.17088105529546738,
      "num_tokens": 10834485.0,
      "step": 4855
    },
    {
      "entropy": 5.683445978164673,
      "epoch": 0.5458527545347336,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004978696356245129,
      "loss": 5.614,
      "mean_token_accuracy": 0.16837672442197799,
      "num_tokens": 10843547.0,
      "step": 4860
    },
    {
      "entropy": 5.719631719589233,
      "epoch": 0.546414331442691,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004978641202957057,
      "loss": 5.7497,
      "mean_token_accuracy": 0.16445043236017226,
      "num_tokens": 10855890.0,
      "step": 4865
    },
    {
      "entropy": 5.91353702545166,
      "epoch": 0.5469759083506486,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004978585978708183,
      "loss": 5.7949,
      "mean_token_accuracy": 0.16377429962158202,
      "num_tokens": 10868060.0,
      "step": 4870
    },
    {
      "entropy": 5.783869218826294,
      "epoch": 0.5475374852586061,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004978530683500267,
      "loss": 5.7163,
      "mean_token_accuracy": 0.1670852780342102,
      "num_tokens": 10879323.0,
      "step": 4875
    },
    {
      "entropy": 5.661490631103516,
      "epoch": 0.5480990621665637,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004978475317335071,
      "loss": 5.5598,
      "mean_token_accuracy": 0.1758361741900444,
      "num_tokens": 10890545.0,
      "step": 4880
    },
    {
      "entropy": 5.658501005172729,
      "epoch": 0.5486606390745212,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004978419880214356,
      "loss": 5.6612,
      "mean_token_accuracy": 0.16851151436567308,
      "num_tokens": 10900761.0,
      "step": 4885
    },
    {
      "entropy": 5.717454147338867,
      "epoch": 0.5492222159824788,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004978364372139887,
      "loss": 5.6471,
      "mean_token_accuracy": 0.17221338003873826,
      "num_tokens": 10911874.0,
      "step": 4890
    },
    {
      "entropy": 5.68519868850708,
      "epoch": 0.5497837928904363,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004978308793113434,
      "loss": 5.5666,
      "mean_token_accuracy": 0.1797461986541748,
      "num_tokens": 10922949.0,
      "step": 4895
    },
    {
      "entropy": 5.734178066253662,
      "epoch": 0.5503453697983939,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004978253143136762,
      "loss": 5.7637,
      "mean_token_accuracy": 0.16003435403108596,
      "num_tokens": 10935415.0,
      "step": 4900
    },
    {
      "entropy": 5.712212610244751,
      "epoch": 0.5509069467063514,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004978197422211647,
      "loss": 5.5989,
      "mean_token_accuracy": 0.1794376477599144,
      "num_tokens": 10946983.0,
      "step": 4905
    },
    {
      "entropy": 5.645990180969238,
      "epoch": 0.551468523614309,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004978141630339863,
      "loss": 5.6755,
      "mean_token_accuracy": 0.16025410294532777,
      "num_tokens": 10957884.0,
      "step": 4910
    },
    {
      "entropy": 5.732716608047485,
      "epoch": 0.5520301005222665,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004978085767523184,
      "loss": 5.6752,
      "mean_token_accuracy": 0.17449066042900085,
      "num_tokens": 10968518.0,
      "step": 4915
    },
    {
      "entropy": 5.708969449996948,
      "epoch": 0.5525916774302241,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000497802983376339,
      "loss": 5.6864,
      "mean_token_accuracy": 0.16695939302444457,
      "num_tokens": 10979621.0,
      "step": 4920
    },
    {
      "entropy": 5.764673709869385,
      "epoch": 0.5531532543381816,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004977973829062263,
      "loss": 5.619,
      "mean_token_accuracy": 0.1710394114255905,
      "num_tokens": 10990158.0,
      "step": 4925
    },
    {
      "entropy": 5.673544979095459,
      "epoch": 0.5537148312461392,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004977917753421585,
      "loss": 5.7305,
      "mean_token_accuracy": 0.16447483003139496,
      "num_tokens": 11001732.0,
      "step": 4930
    },
    {
      "entropy": 5.843268728256225,
      "epoch": 0.5542764081540967,
      "grad_norm": 1.390625,
      "learning_rate": 0.000497786160684314,
      "loss": 5.7428,
      "mean_token_accuracy": 0.1634986564517021,
      "num_tokens": 11013483.0,
      "step": 4935
    },
    {
      "entropy": 5.796664619445801,
      "epoch": 0.5548379850620543,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004977805389328719,
      "loss": 5.7391,
      "mean_token_accuracy": 0.16766392290592194,
      "num_tokens": 11025423.0,
      "step": 4940
    },
    {
      "entropy": 5.703360080718994,
      "epoch": 0.5553995619700118,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000497774910088011,
      "loss": 5.5716,
      "mean_token_accuracy": 0.17025255560874938,
      "num_tokens": 11037041.0,
      "step": 4945
    },
    {
      "entropy": 5.723888826370239,
      "epoch": 0.5559611388779694,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004977692741499106,
      "loss": 5.7208,
      "mean_token_accuracy": 0.16566304564476014,
      "num_tokens": 11049206.0,
      "step": 4950
    },
    {
      "entropy": 5.74290828704834,
      "epoch": 0.5565227157859269,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00049776363111875,
      "loss": 5.6884,
      "mean_token_accuracy": 0.17315239310264588,
      "num_tokens": 11060407.0,
      "step": 4955
    },
    {
      "entropy": 5.688162040710449,
      "epoch": 0.5570842926938844,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004977579809947091,
      "loss": 5.6215,
      "mean_token_accuracy": 0.1681760311126709,
      "num_tokens": 11071269.0,
      "step": 4960
    },
    {
      "entropy": 5.701694965362549,
      "epoch": 0.5576458696018419,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004977523237779676,
      "loss": 5.6215,
      "mean_token_accuracy": 0.17521134316921233,
      "num_tokens": 11082078.0,
      "step": 4965
    },
    {
      "entropy": 5.906933164596557,
      "epoch": 0.5582074465097995,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004977466594687057,
      "loss": 5.8526,
      "mean_token_accuracy": 0.1614489182829857,
      "num_tokens": 11094596.0,
      "step": 4970
    },
    {
      "entropy": 5.737157964706421,
      "epoch": 0.558769023417757,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004977409880671039,
      "loss": 5.7163,
      "mean_token_accuracy": 0.16220240145921708,
      "num_tokens": 11107705.0,
      "step": 4975
    },
    {
      "entropy": 5.703377056121826,
      "epoch": 0.5593306003257146,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004977353095733427,
      "loss": 5.622,
      "mean_token_accuracy": 0.17367908358573914,
      "num_tokens": 11118152.0,
      "step": 4980
    },
    {
      "entropy": 5.711778259277343,
      "epoch": 0.5598921772336721,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004977296239876026,
      "loss": 5.6611,
      "mean_token_accuracy": 0.17005926817655564,
      "num_tokens": 11129261.0,
      "step": 4985
    },
    {
      "entropy": 5.815555191040039,
      "epoch": 0.5604537541416297,
      "grad_norm": 1.203125,
      "learning_rate": 0.000497723931310065,
      "loss": 5.7661,
      "mean_token_accuracy": 0.1613462135195732,
      "num_tokens": 11142227.0,
      "step": 4990
    },
    {
      "entropy": 5.781078052520752,
      "epoch": 0.5610153310495872,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004977182315409111,
      "loss": 5.6717,
      "mean_token_accuracy": 0.1686568960547447,
      "num_tokens": 11154277.0,
      "step": 4995
    },
    {
      "entropy": 5.681065464019776,
      "epoch": 0.5615769079575448,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004977125246803224,
      "loss": 5.6813,
      "mean_token_accuracy": 0.1690026491880417,
      "num_tokens": 11164902.0,
      "step": 5000
    },
    {
      "entropy": 5.634833431243896,
      "epoch": 0.5621384848655023,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004977068107284803,
      "loss": 5.5775,
      "mean_token_accuracy": 0.16621919125318527,
      "num_tokens": 11176010.0,
      "step": 5005
    },
    {
      "entropy": 5.798631000518799,
      "epoch": 0.5627000617734599,
      "grad_norm": 1.375,
      "learning_rate": 0.0004977010896855672,
      "loss": 5.6064,
      "mean_token_accuracy": 0.17368722558021546,
      "num_tokens": 11187773.0,
      "step": 5010
    },
    {
      "entropy": 5.680552387237549,
      "epoch": 0.5632616386814174,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004976953615517649,
      "loss": 5.6802,
      "mean_token_accuracy": 0.17477964162826537,
      "num_tokens": 11198420.0,
      "step": 5015
    },
    {
      "entropy": 5.5892163753509525,
      "epoch": 0.563823215589375,
      "grad_norm": 2.109375,
      "learning_rate": 0.000497689626327256,
      "loss": 5.4906,
      "mean_token_accuracy": 0.17867352068424225,
      "num_tokens": 11208553.0,
      "step": 5020
    },
    {
      "entropy": 5.70281949043274,
      "epoch": 0.5643847924973325,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000497683884012223,
      "loss": 5.6524,
      "mean_token_accuracy": 0.1714022785425186,
      "num_tokens": 11219737.0,
      "step": 5025
    },
    {
      "entropy": 5.650163793563843,
      "epoch": 0.5649463694052901,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004976781346068487,
      "loss": 5.6602,
      "mean_token_accuracy": 0.16621274948120118,
      "num_tokens": 11230997.0,
      "step": 5030
    },
    {
      "entropy": 5.775289058685303,
      "epoch": 0.5655079463132476,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004976723781113163,
      "loss": 5.658,
      "mean_token_accuracy": 0.17257270216941833,
      "num_tokens": 11242338.0,
      "step": 5035
    },
    {
      "entropy": 5.775376224517823,
      "epoch": 0.5660695232212052,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004976666145258089,
      "loss": 5.6957,
      "mean_token_accuracy": 0.17020781189203263,
      "num_tokens": 11252671.0,
      "step": 5040
    },
    {
      "entropy": 5.555762767791748,
      "epoch": 0.5666311001291627,
      "grad_norm": 1.25,
      "learning_rate": 0.0004976608438505103,
      "loss": 5.5207,
      "mean_token_accuracy": 0.17949869185686113,
      "num_tokens": 11262721.0,
      "step": 5045
    },
    {
      "entropy": 5.700231218338013,
      "epoch": 0.5671926770371203,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004976550660856041,
      "loss": 5.7102,
      "mean_token_accuracy": 0.16806024610996245,
      "num_tokens": 11274074.0,
      "step": 5050
    },
    {
      "entropy": 5.789868927001953,
      "epoch": 0.5677542539450777,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000497649281231274,
      "loss": 5.6467,
      "mean_token_accuracy": 0.1738978683948517,
      "num_tokens": 11286051.0,
      "step": 5055
    },
    {
      "entropy": 5.801263046264649,
      "epoch": 0.5683158308530353,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004976434892877046,
      "loss": 5.7707,
      "mean_token_accuracy": 0.16562236547470094,
      "num_tokens": 11298841.0,
      "step": 5060
    },
    {
      "entropy": 5.760428190231323,
      "epoch": 0.5688774077609928,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004976376902550802,
      "loss": 5.7601,
      "mean_token_accuracy": 0.1714622288942337,
      "num_tokens": 11309898.0,
      "step": 5065
    },
    {
      "entropy": 5.770657253265381,
      "epoch": 0.5694389846689504,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004976318841335854,
      "loss": 5.7862,
      "mean_token_accuracy": 0.16856923401355745,
      "num_tokens": 11322163.0,
      "step": 5070
    },
    {
      "entropy": 5.779578781127929,
      "epoch": 0.5700005615769079,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004976260709234049,
      "loss": 5.7287,
      "mean_token_accuracy": 0.165048947930336,
      "num_tokens": 11333405.0,
      "step": 5075
    },
    {
      "entropy": 5.838383293151855,
      "epoch": 0.5705621384848655,
      "grad_norm": 1.1875,
      "learning_rate": 0.000497620250624724,
      "loss": 5.7896,
      "mean_token_accuracy": 0.16699133217334747,
      "num_tokens": 11345037.0,
      "step": 5080
    },
    {
      "entropy": 5.788431072235108,
      "epoch": 0.571123715392823,
      "grad_norm": 1.40625,
      "learning_rate": 0.000497614423237728,
      "loss": 5.7095,
      "mean_token_accuracy": 0.16881436556577684,
      "num_tokens": 11356403.0,
      "step": 5085
    },
    {
      "entropy": 5.6463860988616945,
      "epoch": 0.5716852923007806,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004976085887626026,
      "loss": 5.6361,
      "mean_token_accuracy": 0.1750362992286682,
      "num_tokens": 11367913.0,
      "step": 5090
    },
    {
      "entropy": 5.6775815963745115,
      "epoch": 0.5722468692087381,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004976027471995331,
      "loss": 5.6461,
      "mean_token_accuracy": 0.17520785629749297,
      "num_tokens": 11378519.0,
      "step": 5095
    },
    {
      "entropy": 5.759826040267944,
      "epoch": 0.5728084461166957,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000497596898548706,
      "loss": 5.6073,
      "mean_token_accuracy": 0.17434988915920258,
      "num_tokens": 11389530.0,
      "step": 5100
    },
    {
      "entropy": 5.59257664680481,
      "epoch": 0.5733700230246532,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004975910428103072,
      "loss": 5.5348,
      "mean_token_accuracy": 0.17843396216630936,
      "num_tokens": 11400468.0,
      "step": 5105
    },
    {
      "entropy": 5.500880479812622,
      "epoch": 0.5739315999326108,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004975851799845233,
      "loss": 5.544,
      "mean_token_accuracy": 0.1740495353937149,
      "num_tokens": 11411242.0,
      "step": 5110
    },
    {
      "entropy": 5.757795429229736,
      "epoch": 0.5744931768405683,
      "grad_norm": 1.375,
      "learning_rate": 0.000497579310071541,
      "loss": 5.6211,
      "mean_token_accuracy": 0.17112922221422194,
      "num_tokens": 11420909.0,
      "step": 5115
    },
    {
      "entropy": 5.599933767318726,
      "epoch": 0.5750547537485259,
      "grad_norm": 1.375,
      "learning_rate": 0.0004975734330715471,
      "loss": 5.5078,
      "mean_token_accuracy": 0.17252909243106843,
      "num_tokens": 11431791.0,
      "step": 5120
    },
    {
      "entropy": 5.802357625961304,
      "epoch": 0.5756163306564834,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004975675489847287,
      "loss": 5.7397,
      "mean_token_accuracy": 0.16458898186683654,
      "num_tokens": 11443056.0,
      "step": 5125
    },
    {
      "entropy": 5.745362329483032,
      "epoch": 0.576177907564441,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004975616578112734,
      "loss": 5.7048,
      "mean_token_accuracy": 0.17184705585241317,
      "num_tokens": 11453413.0,
      "step": 5130
    },
    {
      "entropy": 5.848228311538696,
      "epoch": 0.5767394844723985,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004975557595513684,
      "loss": 5.9266,
      "mean_token_accuracy": 0.15690554976463317,
      "num_tokens": 11465457.0,
      "step": 5135
    },
    {
      "entropy": 5.914571619033813,
      "epoch": 0.5773010613803561,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004975498542052018,
      "loss": 5.7023,
      "mean_token_accuracy": 0.1729892149567604,
      "num_tokens": 11476899.0,
      "step": 5140
    },
    {
      "entropy": 5.682943296432495,
      "epoch": 0.5778626382883136,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004975439417729615,
      "loss": 5.6611,
      "mean_token_accuracy": 0.17298050820827485,
      "num_tokens": 11488652.0,
      "step": 5145
    },
    {
      "entropy": 5.711069393157959,
      "epoch": 0.5784242151962711,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004975380222548358,
      "loss": 5.7013,
      "mean_token_accuracy": 0.16747490912675858,
      "num_tokens": 11499540.0,
      "step": 5150
    },
    {
      "entropy": 5.864771842956543,
      "epoch": 0.5789857921042286,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004975320956510131,
      "loss": 5.7277,
      "mean_token_accuracy": 0.17383481860160827,
      "num_tokens": 11511156.0,
      "step": 5155
    },
    {
      "entropy": 5.793862915039062,
      "epoch": 0.5795473690121862,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004975261619616822,
      "loss": 5.7969,
      "mean_token_accuracy": 0.16044799983501434,
      "num_tokens": 11523489.0,
      "step": 5160
    },
    {
      "entropy": 5.701024532318115,
      "epoch": 0.5801089459201437,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004975202211870321,
      "loss": 5.62,
      "mean_token_accuracy": 0.16963277608156205,
      "num_tokens": 11534497.0,
      "step": 5165
    },
    {
      "entropy": 5.849947166442871,
      "epoch": 0.5806705228281013,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004975142733272518,
      "loss": 5.7043,
      "mean_token_accuracy": 0.16341470927000046,
      "num_tokens": 11545585.0,
      "step": 5170
    },
    {
      "entropy": 5.594762372970581,
      "epoch": 0.5812320997360588,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004975083183825306,
      "loss": 5.553,
      "mean_token_accuracy": 0.17195095419883727,
      "num_tokens": 11556278.0,
      "step": 5175
    },
    {
      "entropy": 5.634889745712281,
      "epoch": 0.5817936766440164,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004975023563530584,
      "loss": 5.6343,
      "mean_token_accuracy": 0.17483613938093184,
      "num_tokens": 11568358.0,
      "step": 5180
    },
    {
      "entropy": 5.6792529106140135,
      "epoch": 0.5823552535519739,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004974963872390249,
      "loss": 5.6092,
      "mean_token_accuracy": 0.17366641908884048,
      "num_tokens": 11579777.0,
      "step": 5185
    },
    {
      "entropy": 5.678589296340943,
      "epoch": 0.5829168304599315,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00049749041104062,
      "loss": 5.6641,
      "mean_token_accuracy": 0.1732245460152626,
      "num_tokens": 11591479.0,
      "step": 5190
    },
    {
      "entropy": 5.741658306121826,
      "epoch": 0.583478407367889,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004974844277580343,
      "loss": 5.6151,
      "mean_token_accuracy": 0.1689417690038681,
      "num_tokens": 11601452.0,
      "step": 5195
    },
    {
      "entropy": 5.639203977584839,
      "epoch": 0.5840399842758466,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004974784373914579,
      "loss": 5.479,
      "mean_token_accuracy": 0.1778860330581665,
      "num_tokens": 11611661.0,
      "step": 5200
    },
    {
      "entropy": 5.637208557128906,
      "epoch": 0.5846015611838041,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004974724399410818,
      "loss": 5.662,
      "mean_token_accuracy": 0.16758041232824325,
      "num_tokens": 11622940.0,
      "step": 5205
    },
    {
      "entropy": 5.700867700576782,
      "epoch": 0.5851631380917617,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004974664354070971,
      "loss": 5.6744,
      "mean_token_accuracy": 0.17014639526605607,
      "num_tokens": 11634290.0,
      "step": 5210
    },
    {
      "entropy": 5.802171611785889,
      "epoch": 0.5857247149997192,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004974604237896947,
      "loss": 5.7173,
      "mean_token_accuracy": 0.16413109749555588,
      "num_tokens": 11645812.0,
      "step": 5215
    },
    {
      "entropy": 5.730967092514038,
      "epoch": 0.5862862919076768,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004974544050890662,
      "loss": 5.711,
      "mean_token_accuracy": 0.16343657672405243,
      "num_tokens": 11657520.0,
      "step": 5220
    },
    {
      "entropy": 5.6769603252410885,
      "epoch": 0.5868478688156343,
      "grad_norm": 1.328125,
      "learning_rate": 0.000497448379305403,
      "loss": 5.567,
      "mean_token_accuracy": 0.17653037905693053,
      "num_tokens": 11666887.0,
      "step": 5225
    },
    {
      "entropy": 5.796042203903198,
      "epoch": 0.5874094457235919,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004974423464388973,
      "loss": 5.7423,
      "mean_token_accuracy": 0.16953345388174057,
      "num_tokens": 11677337.0,
      "step": 5230
    },
    {
      "entropy": 5.756515645980835,
      "epoch": 0.5879710226315494,
      "grad_norm": 1.453125,
      "learning_rate": 0.000497436306489741,
      "loss": 5.6623,
      "mean_token_accuracy": 0.17309503853321076,
      "num_tokens": 11688203.0,
      "step": 5235
    },
    {
      "entropy": 5.702898836135864,
      "epoch": 0.588532599539507,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004974302594581264,
      "loss": 5.702,
      "mean_token_accuracy": 0.16658136248588562,
      "num_tokens": 11699727.0,
      "step": 5240
    },
    {
      "entropy": 5.732393646240235,
      "epoch": 0.5890941764474645,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004974242053442461,
      "loss": 5.7085,
      "mean_token_accuracy": 0.16990188509225845,
      "num_tokens": 11710956.0,
      "step": 5245
    },
    {
      "entropy": 5.743574380874634,
      "epoch": 0.589655753355422,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004974181441482927,
      "loss": 5.5437,
      "mean_token_accuracy": 0.17698764950037002,
      "num_tokens": 11721681.0,
      "step": 5250
    },
    {
      "entropy": 5.713166904449463,
      "epoch": 0.5902173302633795,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004974120758704595,
      "loss": 5.6719,
      "mean_token_accuracy": 0.16816408634185792,
      "num_tokens": 11731679.0,
      "step": 5255
    },
    {
      "entropy": 5.8048162937164305,
      "epoch": 0.5907789071713371,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004974060005109394,
      "loss": 5.8048,
      "mean_token_accuracy": 0.1657121554017067,
      "num_tokens": 11743953.0,
      "step": 5260
    },
    {
      "entropy": 5.6638165473937985,
      "epoch": 0.5913404840792946,
      "grad_norm": 1.3125,
      "learning_rate": 0.000497399918069926,
      "loss": 5.5795,
      "mean_token_accuracy": 0.17744518518447877,
      "num_tokens": 11754091.0,
      "step": 5265
    },
    {
      "entropy": 5.639192199707031,
      "epoch": 0.5919020609872522,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004973938285476131,
      "loss": 5.6066,
      "mean_token_accuracy": 0.1745060995221138,
      "num_tokens": 11764771.0,
      "step": 5270
    },
    {
      "entropy": 5.734188079833984,
      "epoch": 0.5924636378952097,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004973877319441943,
      "loss": 5.6757,
      "mean_token_accuracy": 0.16847294121980666,
      "num_tokens": 11776287.0,
      "step": 5275
    },
    {
      "entropy": 5.747639417648315,
      "epoch": 0.5930252148031673,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004973816282598639,
      "loss": 5.6861,
      "mean_token_accuracy": 0.16214133352041243,
      "num_tokens": 11787142.0,
      "step": 5280
    },
    {
      "entropy": 5.72074031829834,
      "epoch": 0.5935867917111248,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004973755174948162,
      "loss": 5.6593,
      "mean_token_accuracy": 0.16755084693431854,
      "num_tokens": 11798966.0,
      "step": 5285
    },
    {
      "entropy": 5.771004104614258,
      "epoch": 0.5941483686190824,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004973693996492458,
      "loss": 5.7396,
      "mean_token_accuracy": 0.16607423722743989,
      "num_tokens": 11810529.0,
      "step": 5290
    },
    {
      "entropy": 5.65753321647644,
      "epoch": 0.5947099455270399,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004973632747233475,
      "loss": 5.584,
      "mean_token_accuracy": 0.1776696190237999,
      "num_tokens": 11821965.0,
      "step": 5295
    },
    {
      "entropy": 5.7572873592376705,
      "epoch": 0.5952715224349975,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004973571427173163,
      "loss": 5.7412,
      "mean_token_accuracy": 0.1662243828177452,
      "num_tokens": 11833936.0,
      "step": 5300
    },
    {
      "entropy": 5.793422603607178,
      "epoch": 0.595833099342955,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004973510036313473,
      "loss": 5.6572,
      "mean_token_accuracy": 0.17151601612567902,
      "num_tokens": 11845245.0,
      "step": 5305
    },
    {
      "entropy": 5.701361179351807,
      "epoch": 0.5963946762509126,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004973448574656362,
      "loss": 5.6248,
      "mean_token_accuracy": 0.17118371576070784,
      "num_tokens": 11856010.0,
      "step": 5310
    },
    {
      "entropy": 5.748444557189941,
      "epoch": 0.5969562531588701,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004973387042203786,
      "loss": 5.7801,
      "mean_token_accuracy": 0.16333963572978974,
      "num_tokens": 11866194.0,
      "step": 5315
    },
    {
      "entropy": 5.785281944274902,
      "epoch": 0.5975178300668277,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004973325438957704,
      "loss": 5.6244,
      "mean_token_accuracy": 0.170792356133461,
      "num_tokens": 11877224.0,
      "step": 5320
    },
    {
      "entropy": 5.662514305114746,
      "epoch": 0.5980794069747852,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004973263764920079,
      "loss": 5.5478,
      "mean_token_accuracy": 0.17639639377593994,
      "num_tokens": 11887410.0,
      "step": 5325
    },
    {
      "entropy": 5.627243566513061,
      "epoch": 0.5986409838827428,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004973202020092871,
      "loss": 5.7129,
      "mean_token_accuracy": 0.1726754367351532,
      "num_tokens": 11898816.0,
      "step": 5330
    },
    {
      "entropy": 5.750458002090454,
      "epoch": 0.5992025607907003,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000497314020447805,
      "loss": 5.6623,
      "mean_token_accuracy": 0.16915512531995774,
      "num_tokens": 11909220.0,
      "step": 5335
    },
    {
      "entropy": 5.7871287822723385,
      "epoch": 0.5997641376986579,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004973078318077583,
      "loss": 5.6694,
      "mean_token_accuracy": 0.16608534008264542,
      "num_tokens": 11919584.0,
      "step": 5340
    },
    {
      "entropy": 5.664968776702881,
      "epoch": 0.6003257146066153,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000497301636089344,
      "loss": 5.6782,
      "mean_token_accuracy": 0.16855976581573487,
      "num_tokens": 11929298.0,
      "step": 5345
    },
    {
      "entropy": 5.701796770095825,
      "epoch": 0.6008872915145729,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004972954332927592,
      "loss": 5.6465,
      "mean_token_accuracy": 0.17208643406629562,
      "num_tokens": 11939809.0,
      "step": 5350
    },
    {
      "entropy": 5.727794981002807,
      "epoch": 0.6014488684225304,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004972892234182018,
      "loss": 5.541,
      "mean_token_accuracy": 0.1775716796517372,
      "num_tokens": 11950292.0,
      "step": 5355
    },
    {
      "entropy": 5.610636901855469,
      "epoch": 0.602010445330488,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004972830064658692,
      "loss": 5.568,
      "mean_token_accuracy": 0.17599292993545532,
      "num_tokens": 11961774.0,
      "step": 5360
    },
    {
      "entropy": 5.764124965667724,
      "epoch": 0.6025720222384455,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004972767824359596,
      "loss": 5.6824,
      "mean_token_accuracy": 0.16461727768182755,
      "num_tokens": 11973351.0,
      "step": 5365
    },
    {
      "entropy": 5.664096021652222,
      "epoch": 0.6031335991464031,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004972705513286709,
      "loss": 5.5476,
      "mean_token_accuracy": 0.178704696893692,
      "num_tokens": 11984472.0,
      "step": 5370
    },
    {
      "entropy": 5.647040939331054,
      "epoch": 0.6036951760543606,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004972643131442017,
      "loss": 5.6518,
      "mean_token_accuracy": 0.17527108937501906,
      "num_tokens": 11994271.0,
      "step": 5375
    },
    {
      "entropy": 5.860633420944214,
      "epoch": 0.6042567529623182,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004972580678827506,
      "loss": 5.7576,
      "mean_token_accuracy": 0.16238637864589692,
      "num_tokens": 12007647.0,
      "step": 5380
    },
    {
      "entropy": 5.786860418319702,
      "epoch": 0.6048183298702757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004972518155445163,
      "loss": 5.7092,
      "mean_token_accuracy": 0.16535102427005768,
      "num_tokens": 12020331.0,
      "step": 5385
    },
    {
      "entropy": 5.657439231872559,
      "epoch": 0.6053799067782333,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004972455561296982,
      "loss": 5.7088,
      "mean_token_accuracy": 0.16050323843955994,
      "num_tokens": 12031616.0,
      "step": 5390
    },
    {
      "entropy": 5.811110305786133,
      "epoch": 0.6059414836861908,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004972392896384952,
      "loss": 5.6699,
      "mean_token_accuracy": 0.1691645324230194,
      "num_tokens": 12044107.0,
      "step": 5395
    },
    {
      "entropy": 5.748564100265503,
      "epoch": 0.6065030605941484,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004972330160711071,
      "loss": 5.6608,
      "mean_token_accuracy": 0.17236095368862153,
      "num_tokens": 12055747.0,
      "step": 5400
    },
    {
      "entropy": 5.757844591140747,
      "epoch": 0.6070646375021059,
      "grad_norm": 1.25,
      "learning_rate": 0.0004972267354277336,
      "loss": 5.6581,
      "mean_token_accuracy": 0.17029646039009094,
      "num_tokens": 12067849.0,
      "step": 5405
    },
    {
      "entropy": 5.722942781448364,
      "epoch": 0.6076262144100635,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004972204477085746,
      "loss": 5.6759,
      "mean_token_accuracy": 0.16894526779651642,
      "num_tokens": 12078534.0,
      "step": 5410
    },
    {
      "entropy": 5.657567405700684,
      "epoch": 0.608187791318021,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004972141529138303,
      "loss": 5.6041,
      "mean_token_accuracy": 0.17073164135217667,
      "num_tokens": 12089534.0,
      "step": 5415
    },
    {
      "entropy": 5.70810079574585,
      "epoch": 0.6087493682259786,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004972078510437013,
      "loss": 5.5066,
      "mean_token_accuracy": 0.18743069767951964,
      "num_tokens": 12099513.0,
      "step": 5420
    },
    {
      "entropy": 5.70379114151001,
      "epoch": 0.6093109451339361,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004972015420983882,
      "loss": 5.712,
      "mean_token_accuracy": 0.17635903805494307,
      "num_tokens": 12110106.0,
      "step": 5425
    },
    {
      "entropy": 5.705994033813477,
      "epoch": 0.6098725220418937,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004971952260780915,
      "loss": 5.604,
      "mean_token_accuracy": 0.1767367586493492,
      "num_tokens": 12121302.0,
      "step": 5430
    },
    {
      "entropy": 5.692618274688721,
      "epoch": 0.6104340989498512,
      "grad_norm": 1.328125,
      "learning_rate": 0.000497188902983013,
      "loss": 5.6502,
      "mean_token_accuracy": 0.17055889666080476,
      "num_tokens": 12132879.0,
      "step": 5435
    },
    {
      "entropy": 5.642355585098267,
      "epoch": 0.6109956758578087,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004971825728133535,
      "loss": 5.5734,
      "mean_token_accuracy": 0.17844846993684768,
      "num_tokens": 12143597.0,
      "step": 5440
    },
    {
      "entropy": 5.729561996459961,
      "epoch": 0.6115572527657662,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004971762355693146,
      "loss": 5.6705,
      "mean_token_accuracy": 0.17112083733081818,
      "num_tokens": 12156036.0,
      "step": 5445
    },
    {
      "entropy": 5.774557399749756,
      "epoch": 0.6121188296737238,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004971698912510983,
      "loss": 5.6849,
      "mean_token_accuracy": 0.1741700991988182,
      "num_tokens": 12167305.0,
      "step": 5450
    },
    {
      "entropy": 5.718290376663208,
      "epoch": 0.6126804065816813,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004971635398589066,
      "loss": 5.6458,
      "mean_token_accuracy": 0.17421892285346985,
      "num_tokens": 12176772.0,
      "step": 5455
    },
    {
      "entropy": 5.706922245025635,
      "epoch": 0.6132419834896389,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004971571813929415,
      "loss": 5.6452,
      "mean_token_accuracy": 0.17209067046642304,
      "num_tokens": 12187852.0,
      "step": 5460
    },
    {
      "entropy": 5.790262699127197,
      "epoch": 0.6138035603975964,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004971508158534056,
      "loss": 5.7057,
      "mean_token_accuracy": 0.17010461091995238,
      "num_tokens": 12198574.0,
      "step": 5465
    },
    {
      "entropy": 5.622075748443604,
      "epoch": 0.614365137305554,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004971444432405015,
      "loss": 5.4607,
      "mean_token_accuracy": 0.18198293149471284,
      "num_tokens": 12209463.0,
      "step": 5470
    },
    {
      "entropy": 5.731533527374268,
      "epoch": 0.6149267142135115,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004971380635544323,
      "loss": 5.7366,
      "mean_token_accuracy": 0.16922779083251954,
      "num_tokens": 12220993.0,
      "step": 5475
    },
    {
      "entropy": 5.730971240997315,
      "epoch": 0.6154882911214691,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004971316767954009,
      "loss": 5.6102,
      "mean_token_accuracy": 0.17507601529359818,
      "num_tokens": 12232648.0,
      "step": 5480
    },
    {
      "entropy": 5.736530876159668,
      "epoch": 0.6160498680294266,
      "grad_norm": 1.625,
      "learning_rate": 0.0004971252829636108,
      "loss": 5.65,
      "mean_token_accuracy": 0.17610020488500594,
      "num_tokens": 12243931.0,
      "step": 5485
    },
    {
      "entropy": 5.787777423858643,
      "epoch": 0.6166114449373842,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004971188820592656,
      "loss": 5.6509,
      "mean_token_accuracy": 0.17270122319459916,
      "num_tokens": 12254821.0,
      "step": 5490
    },
    {
      "entropy": 5.583421754837036,
      "epoch": 0.6171730218453417,
      "grad_norm": 1.203125,
      "learning_rate": 0.000497112474082569,
      "loss": 5.6572,
      "mean_token_accuracy": 0.17220793664455414,
      "num_tokens": 12265440.0,
      "step": 5495
    },
    {
      "entropy": 5.727280187606811,
      "epoch": 0.6177345987532993,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000497106059033725,
      "loss": 5.6329,
      "mean_token_accuracy": 0.17365415692329406,
      "num_tokens": 12276636.0,
      "step": 5500
    },
    {
      "entropy": 5.766627502441406,
      "epoch": 0.6182961756612568,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004970996369129381,
      "loss": 5.5133,
      "mean_token_accuracy": 0.17947188764810562,
      "num_tokens": 12287404.0,
      "step": 5505
    },
    {
      "entropy": 5.672186088562012,
      "epoch": 0.6188577525692144,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004970932077204125,
      "loss": 5.7397,
      "mean_token_accuracy": 0.1642993837594986,
      "num_tokens": 12299283.0,
      "step": 5510
    },
    {
      "entropy": 5.666852855682373,
      "epoch": 0.6194193294771719,
      "grad_norm": 1.296875,
      "learning_rate": 0.000497086771456353,
      "loss": 5.6404,
      "mean_token_accuracy": 0.16676075011491776,
      "num_tokens": 12310281.0,
      "step": 5515
    },
    {
      "entropy": 5.804997062683105,
      "epoch": 0.6199809063851295,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004970803281209646,
      "loss": 5.6661,
      "mean_token_accuracy": 0.17195305079221726,
      "num_tokens": 12320839.0,
      "step": 5520
    },
    {
      "entropy": 5.676700162887573,
      "epoch": 0.620542483293087,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004970738777144524,
      "loss": 5.6211,
      "mean_token_accuracy": 0.17245084047317505,
      "num_tokens": 12332845.0,
      "step": 5525
    },
    {
      "entropy": 5.554922676086425,
      "epoch": 0.6211040602010446,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004970674202370218,
      "loss": 5.5204,
      "mean_token_accuracy": 0.1806936517357826,
      "num_tokens": 12343327.0,
      "step": 5530
    },
    {
      "entropy": 5.72095046043396,
      "epoch": 0.621665637109002,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004970609556888785,
      "loss": 5.6445,
      "mean_token_accuracy": 0.17047364860773087,
      "num_tokens": 12353224.0,
      "step": 5535
    },
    {
      "entropy": 5.682199335098266,
      "epoch": 0.6222272140169596,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004970544840702282,
      "loss": 5.5634,
      "mean_token_accuracy": 0.17255714684724807,
      "num_tokens": 12363981.0,
      "step": 5540
    },
    {
      "entropy": 5.584892797470093,
      "epoch": 0.6227887909249171,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004970480053812771,
      "loss": 5.5853,
      "mean_token_accuracy": 0.1800747200846672,
      "num_tokens": 12374769.0,
      "step": 5545
    },
    {
      "entropy": 5.814059019088745,
      "epoch": 0.6233503678328747,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004970415196222312,
      "loss": 5.8038,
      "mean_token_accuracy": 0.16921334117650985,
      "num_tokens": 12387247.0,
      "step": 5550
    },
    {
      "entropy": 5.793169641494751,
      "epoch": 0.6239119447408322,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004970350267932973,
      "loss": 5.6566,
      "mean_token_accuracy": 0.17780971080064772,
      "num_tokens": 12398347.0,
      "step": 5555
    },
    {
      "entropy": 5.531301593780517,
      "epoch": 0.6244735216487898,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000497028526894682,
      "loss": 5.5502,
      "mean_token_accuracy": 0.17427618205547332,
      "num_tokens": 12409763.0,
      "step": 5560
    },
    {
      "entropy": 5.740422582626342,
      "epoch": 0.6250350985567473,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004970220199265923,
      "loss": 5.6429,
      "mean_token_accuracy": 0.17239355593919753,
      "num_tokens": 12420658.0,
      "step": 5565
    },
    {
      "entropy": 5.830804109573364,
      "epoch": 0.6255966754647049,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004970155058892354,
      "loss": 5.7334,
      "mean_token_accuracy": 0.16390733271837235,
      "num_tokens": 12433959.0,
      "step": 5570
    },
    {
      "entropy": 5.738416051864624,
      "epoch": 0.6261582523726624,
      "grad_norm": 1.5,
      "learning_rate": 0.0004970089847828187,
      "loss": 5.7218,
      "mean_token_accuracy": 0.16629742383956908,
      "num_tokens": 12444756.0,
      "step": 5575
    },
    {
      "entropy": 5.73380479812622,
      "epoch": 0.62671982928062,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004970024566075497,
      "loss": 5.6352,
      "mean_token_accuracy": 0.17590337246656418,
      "num_tokens": 12456114.0,
      "step": 5580
    },
    {
      "entropy": 5.736455249786377,
      "epoch": 0.6272814061885775,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004969959213636365,
      "loss": 5.6718,
      "mean_token_accuracy": 0.17344282269477845,
      "num_tokens": 12466927.0,
      "step": 5585
    },
    {
      "entropy": 5.701116847991943,
      "epoch": 0.6278429830965351,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004969893790512871,
      "loss": 5.723,
      "mean_token_accuracy": 0.16070540398359298,
      "num_tokens": 12477834.0,
      "step": 5590
    },
    {
      "entropy": 5.750099849700928,
      "epoch": 0.6284045600044926,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004969828296707097,
      "loss": 5.6247,
      "mean_token_accuracy": 0.1745440736413002,
      "num_tokens": 12488219.0,
      "step": 5595
    },
    {
      "entropy": 5.8359832763671875,
      "epoch": 0.6289661369124502,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004969762732221129,
      "loss": 5.7385,
      "mean_token_accuracy": 0.17294516116380693,
      "num_tokens": 12498947.0,
      "step": 5600
    },
    {
      "entropy": 5.66269965171814,
      "epoch": 0.6295277138204077,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004969697097057054,
      "loss": 5.6343,
      "mean_token_accuracy": 0.1720503583550453,
      "num_tokens": 12510142.0,
      "step": 5605
    },
    {
      "entropy": 5.79200439453125,
      "epoch": 0.6300892907283653,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004969631391216964,
      "loss": 5.6472,
      "mean_token_accuracy": 0.1704936742782593,
      "num_tokens": 12521350.0,
      "step": 5610
    },
    {
      "entropy": 5.521368408203125,
      "epoch": 0.6306508676363228,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004969565614702948,
      "loss": 5.4855,
      "mean_token_accuracy": 0.1774906650185585,
      "num_tokens": 12532579.0,
      "step": 5615
    },
    {
      "entropy": 5.734045362472534,
      "epoch": 0.6312124445442804,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004969499767517104,
      "loss": 5.6044,
      "mean_token_accuracy": 0.17279993891716003,
      "num_tokens": 12542885.0,
      "step": 5620
    },
    {
      "entropy": 5.80007586479187,
      "epoch": 0.6317740214522379,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004969433849661525,
      "loss": 5.7554,
      "mean_token_accuracy": 0.1622144266963005,
      "num_tokens": 12554881.0,
      "step": 5625
    },
    {
      "entropy": 5.737228202819824,
      "epoch": 0.6323355983601955,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004969367861138313,
      "loss": 5.6993,
      "mean_token_accuracy": 0.16859257966279984,
      "num_tokens": 12564626.0,
      "step": 5630
    },
    {
      "entropy": 5.745629692077637,
      "epoch": 0.6328971752681529,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004969301801949565,
      "loss": 5.5502,
      "mean_token_accuracy": 0.17777176648378373,
      "num_tokens": 12576180.0,
      "step": 5635
    },
    {
      "entropy": 5.658581924438477,
      "epoch": 0.6334587521761105,
      "grad_norm": 1.234375,
      "learning_rate": 0.000496923567209739,
      "loss": 5.6427,
      "mean_token_accuracy": 0.16927312016487123,
      "num_tokens": 12587487.0,
      "step": 5640
    },
    {
      "entropy": 5.788752126693725,
      "epoch": 0.634020329084068,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004969169471583889,
      "loss": 5.7908,
      "mean_token_accuracy": 0.16504471004009247,
      "num_tokens": 12598268.0,
      "step": 5645
    },
    {
      "entropy": 5.807561635971069,
      "epoch": 0.6345819059920256,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004969103200411171,
      "loss": 5.6686,
      "mean_token_accuracy": 0.1691132053732872,
      "num_tokens": 12609287.0,
      "step": 5650
    },
    {
      "entropy": 5.669478559494019,
      "epoch": 0.6351434828999831,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004969036858581346,
      "loss": 5.5726,
      "mean_token_accuracy": 0.17225826978683473,
      "num_tokens": 12620317.0,
      "step": 5655
    },
    {
      "entropy": 5.70239109992981,
      "epoch": 0.6357050598079407,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004968970446096527,
      "loss": 5.6395,
      "mean_token_accuracy": 0.17853050231933593,
      "num_tokens": 12631420.0,
      "step": 5660
    },
    {
      "entropy": 5.694774198532104,
      "epoch": 0.6362666367158982,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000496890396295883,
      "loss": 5.5798,
      "mean_token_accuracy": 0.1807517871260643,
      "num_tokens": 12641319.0,
      "step": 5665
    },
    {
      "entropy": 5.760483884811402,
      "epoch": 0.6368282136238558,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004968837409170368,
      "loss": 5.6805,
      "mean_token_accuracy": 0.1660526767373085,
      "num_tokens": 12651662.0,
      "step": 5670
    },
    {
      "entropy": 5.768562316894531,
      "epoch": 0.6373897905318133,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004968770784733263,
      "loss": 5.7284,
      "mean_token_accuracy": 0.1657192662358284,
      "num_tokens": 12663488.0,
      "step": 5675
    },
    {
      "entropy": 5.687350702285767,
      "epoch": 0.6379513674397709,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004968704089649636,
      "loss": 5.6362,
      "mean_token_accuracy": 0.1720639333128929,
      "num_tokens": 12675848.0,
      "step": 5680
    },
    {
      "entropy": 5.735997581481934,
      "epoch": 0.6385129443477284,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004968637323921611,
      "loss": 5.6178,
      "mean_token_accuracy": 0.1734489381313324,
      "num_tokens": 12686541.0,
      "step": 5685
    },
    {
      "entropy": 5.710645627975464,
      "epoch": 0.639074521255686,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004968570487551312,
      "loss": 5.6499,
      "mean_token_accuracy": 0.17180876433849335,
      "num_tokens": 12698001.0,
      "step": 5690
    },
    {
      "entropy": 5.720677328109741,
      "epoch": 0.6396360981636435,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004968503580540869,
      "loss": 5.6477,
      "mean_token_accuracy": 0.1650005429983139,
      "num_tokens": 12709069.0,
      "step": 5695
    },
    {
      "entropy": 5.759660863876343,
      "epoch": 0.6401976750716011,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004968436602892411,
      "loss": 5.6475,
      "mean_token_accuracy": 0.16808155924081802,
      "num_tokens": 12719916.0,
      "step": 5700
    },
    {
      "entropy": 5.743867063522339,
      "epoch": 0.6407592519795586,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004968369554608071,
      "loss": 5.7584,
      "mean_token_accuracy": 0.16313536167144777,
      "num_tokens": 12732079.0,
      "step": 5705
    },
    {
      "entropy": 5.852132272720337,
      "epoch": 0.6413208288875162,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004968302435689985,
      "loss": 5.7338,
      "mean_token_accuracy": 0.16560664474964143,
      "num_tokens": 12742708.0,
      "step": 5710
    },
    {
      "entropy": 5.792535924911499,
      "epoch": 0.6418824057954737,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004968235246140288,
      "loss": 5.6781,
      "mean_token_accuracy": 0.1690981075167656,
      "num_tokens": 12753952.0,
      "step": 5715
    },
    {
      "entropy": 5.677698373794556,
      "epoch": 0.6424439827034313,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004968167985961121,
      "loss": 5.6078,
      "mean_token_accuracy": 0.17256635576486587,
      "num_tokens": 12765226.0,
      "step": 5720
    },
    {
      "entropy": 5.602059602737427,
      "epoch": 0.6430055596113888,
      "grad_norm": 1.5,
      "learning_rate": 0.0004968100655154625,
      "loss": 5.5366,
      "mean_token_accuracy": 0.17826073467731476,
      "num_tokens": 12776455.0,
      "step": 5725
    },
    {
      "entropy": 5.672917366027832,
      "epoch": 0.6435671365193463,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004968033253722943,
      "loss": 5.5564,
      "mean_token_accuracy": 0.17404053062200547,
      "num_tokens": 12787054.0,
      "step": 5730
    },
    {
      "entropy": 5.719582462310791,
      "epoch": 0.6441287134273038,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004967965781668223,
      "loss": 5.6866,
      "mean_token_accuracy": 0.170250004529953,
      "num_tokens": 12798223.0,
      "step": 5735
    },
    {
      "entropy": 5.715972423553467,
      "epoch": 0.6446902903352614,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004967898238992613,
      "loss": 5.5732,
      "mean_token_accuracy": 0.17579154819250106,
      "num_tokens": 12809388.0,
      "step": 5740
    },
    {
      "entropy": 5.686679077148438,
      "epoch": 0.6452518672432189,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004967830625698262,
      "loss": 5.6769,
      "mean_token_accuracy": 0.17444596737623214,
      "num_tokens": 12820116.0,
      "step": 5745
    },
    {
      "entropy": 5.6671665668487545,
      "epoch": 0.6458134441511765,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004967762941787323,
      "loss": 5.5967,
      "mean_token_accuracy": 0.17919077277183532,
      "num_tokens": 12831590.0,
      "step": 5750
    },
    {
      "entropy": 5.747342586517334,
      "epoch": 0.646375021059134,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004967695187261952,
      "loss": 5.7564,
      "mean_token_accuracy": 0.16459435522556304,
      "num_tokens": 12843510.0,
      "step": 5755
    },
    {
      "entropy": 5.7117125511169435,
      "epoch": 0.6469365979670916,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004967627362124306,
      "loss": 5.6307,
      "mean_token_accuracy": 0.1705598846077919,
      "num_tokens": 12854907.0,
      "step": 5760
    },
    {
      "entropy": 5.658813190460205,
      "epoch": 0.6474981748750491,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004967559466376546,
      "loss": 5.5959,
      "mean_token_accuracy": 0.1710912510752678,
      "num_tokens": 12865753.0,
      "step": 5765
    },
    {
      "entropy": 5.6811473846435545,
      "epoch": 0.6480597517830067,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004967491500020831,
      "loss": 5.5035,
      "mean_token_accuracy": 0.17676436007022858,
      "num_tokens": 12877116.0,
      "step": 5770
    },
    {
      "entropy": 5.5508228778839115,
      "epoch": 0.6486213286909642,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004967423463059328,
      "loss": 5.4959,
      "mean_token_accuracy": 0.17579397261142732,
      "num_tokens": 12887685.0,
      "step": 5775
    },
    {
      "entropy": 5.713850355148315,
      "epoch": 0.6491829055989218,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004967355355494201,
      "loss": 5.6584,
      "mean_token_accuracy": 0.1692407101392746,
      "num_tokens": 12898519.0,
      "step": 5780
    },
    {
      "entropy": 5.742898797988891,
      "epoch": 0.6497444825068793,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000496728717732762,
      "loss": 5.586,
      "mean_token_accuracy": 0.17314111143350602,
      "num_tokens": 12908400.0,
      "step": 5785
    },
    {
      "entropy": 5.602768039703369,
      "epoch": 0.6503060594148369,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004967218928561755,
      "loss": 5.56,
      "mean_token_accuracy": 0.17106268554925919,
      "num_tokens": 12919159.0,
      "step": 5790
    },
    {
      "entropy": 5.631039237976074,
      "epoch": 0.6508676363227944,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004967150609198779,
      "loss": 5.523,
      "mean_token_accuracy": 0.1793055310845375,
      "num_tokens": 12929838.0,
      "step": 5795
    },
    {
      "entropy": 5.732568550109863,
      "epoch": 0.651429213230752,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004967082219240869,
      "loss": 5.6285,
      "mean_token_accuracy": 0.1695741206407547,
      "num_tokens": 12939976.0,
      "step": 5800
    },
    {
      "entropy": 5.758488702774048,
      "epoch": 0.6519907901387095,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004967013758690199,
      "loss": 5.6404,
      "mean_token_accuracy": 0.1749062106013298,
      "num_tokens": 12950493.0,
      "step": 5805
    },
    {
      "entropy": 5.654942989349365,
      "epoch": 0.6525523670466671,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004966945227548953,
      "loss": 5.6052,
      "mean_token_accuracy": 0.17201246023178102,
      "num_tokens": 12961267.0,
      "step": 5810
    },
    {
      "entropy": 5.66741909980774,
      "epoch": 0.6531139439546246,
      "grad_norm": 1.25,
      "learning_rate": 0.0004966876625819309,
      "loss": 5.6493,
      "mean_token_accuracy": 0.16587674468755723,
      "num_tokens": 12972720.0,
      "step": 5815
    },
    {
      "entropy": 5.8147331237792965,
      "epoch": 0.6536755208625822,
      "grad_norm": 1.375,
      "learning_rate": 0.0004966807953503456,
      "loss": 5.7294,
      "mean_token_accuracy": 0.16530007719993592,
      "num_tokens": 12984025.0,
      "step": 5820
    },
    {
      "entropy": 5.751352405548095,
      "epoch": 0.6542370977705396,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004966739210603578,
      "loss": 5.6064,
      "mean_token_accuracy": 0.17820507735013963,
      "num_tokens": 12994422.0,
      "step": 5825
    },
    {
      "entropy": 5.753329849243164,
      "epoch": 0.6547986746784972,
      "grad_norm": 1.5,
      "learning_rate": 0.0004966670397121862,
      "loss": 5.7651,
      "mean_token_accuracy": 0.1632770538330078,
      "num_tokens": 13005040.0,
      "step": 5830
    },
    {
      "entropy": 5.65441665649414,
      "epoch": 0.6553602515864547,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004966601513060502,
      "loss": 5.5259,
      "mean_token_accuracy": 0.17237963378429413,
      "num_tokens": 13015720.0,
      "step": 5835
    },
    {
      "entropy": 5.689071893692017,
      "epoch": 0.6559218284944123,
      "grad_norm": 1.25,
      "learning_rate": 0.000496653255842169,
      "loss": 5.4925,
      "mean_token_accuracy": 0.17613681703805922,
      "num_tokens": 13026269.0,
      "step": 5840
    },
    {
      "entropy": 5.638681697845459,
      "epoch": 0.6564834054023698,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004966463533207621,
      "loss": 5.6659,
      "mean_token_accuracy": 0.16914643421769143,
      "num_tokens": 13037102.0,
      "step": 5845
    },
    {
      "entropy": 5.634537220001221,
      "epoch": 0.6570449823103274,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004966394437420493,
      "loss": 5.5301,
      "mean_token_accuracy": 0.1786402016878128,
      "num_tokens": 13047149.0,
      "step": 5850
    },
    {
      "entropy": 5.6855401515960695,
      "epoch": 0.6576065592182849,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004966325271062508,
      "loss": 5.5626,
      "mean_token_accuracy": 0.17522660344839097,
      "num_tokens": 13058196.0,
      "step": 5855
    },
    {
      "entropy": 5.617813682556152,
      "epoch": 0.6581681361262425,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004966256034135866,
      "loss": 5.575,
      "mean_token_accuracy": 0.17152782380580903,
      "num_tokens": 13067897.0,
      "step": 5860
    },
    {
      "entropy": 5.704078626632691,
      "epoch": 0.6587297130342,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004966186726642772,
      "loss": 5.6755,
      "mean_token_accuracy": 0.1717715382575989,
      "num_tokens": 13079720.0,
      "step": 5865
    },
    {
      "entropy": 5.634116268157959,
      "epoch": 0.6592912899421576,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004966117348585434,
      "loss": 5.5258,
      "mean_token_accuracy": 0.167197148501873,
      "num_tokens": 13090995.0,
      "step": 5870
    },
    {
      "entropy": 5.605711269378662,
      "epoch": 0.6598528668501151,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004966047899966059,
      "loss": 5.572,
      "mean_token_accuracy": 0.17753074020147325,
      "num_tokens": 13101945.0,
      "step": 5875
    },
    {
      "entropy": 5.679044914245606,
      "epoch": 0.6604144437580727,
      "grad_norm": 1.3125,
      "learning_rate": 0.000496597838078686,
      "loss": 5.5155,
      "mean_token_accuracy": 0.1752203419804573,
      "num_tokens": 13112869.0,
      "step": 5880
    },
    {
      "entropy": 5.710168552398682,
      "epoch": 0.6609760206660302,
      "grad_norm": 1.390625,
      "learning_rate": 0.000496590879105005,
      "loss": 5.6074,
      "mean_token_accuracy": 0.16586030423641204,
      "num_tokens": 13123242.0,
      "step": 5885
    },
    {
      "entropy": 5.776198816299439,
      "epoch": 0.6615375975739878,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004965839130757845,
      "loss": 5.7182,
      "mean_token_accuracy": 0.1584482118487358,
      "num_tokens": 13133981.0,
      "step": 5890
    },
    {
      "entropy": 5.67058835029602,
      "epoch": 0.6620991744819453,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004965769399912462,
      "loss": 5.6804,
      "mean_token_accuracy": 0.16925962790846824,
      "num_tokens": 13144696.0,
      "step": 5895
    },
    {
      "entropy": 5.770441007614136,
      "epoch": 0.6626607513899029,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004965699598516122,
      "loss": 5.7219,
      "mean_token_accuracy": 0.17378196567296983,
      "num_tokens": 13156674.0,
      "step": 5900
    },
    {
      "entropy": 5.704008340835571,
      "epoch": 0.6632223282978604,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004965629726571047,
      "loss": 5.5632,
      "mean_token_accuracy": 0.17381847351789476,
      "num_tokens": 13167757.0,
      "step": 5905
    },
    {
      "entropy": 5.748448419570923,
      "epoch": 0.663783905205818,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004965559784079463,
      "loss": 5.6488,
      "mean_token_accuracy": 0.16911055743694306,
      "num_tokens": 13177701.0,
      "step": 5910
    },
    {
      "entropy": 5.785614109039306,
      "epoch": 0.6643454821137755,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004965489771043596,
      "loss": 5.6658,
      "mean_token_accuracy": 0.16642796993255615,
      "num_tokens": 13189840.0,
      "step": 5915
    },
    {
      "entropy": 5.5934004306793215,
      "epoch": 0.664907059021733,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004965419687465676,
      "loss": 5.4739,
      "mean_token_accuracy": 0.17768412381410598,
      "num_tokens": 13200206.0,
      "step": 5920
    },
    {
      "entropy": 5.586026859283447,
      "epoch": 0.6654686359296905,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004965349533347934,
      "loss": 5.588,
      "mean_token_accuracy": 0.17564436197280883,
      "num_tokens": 13211474.0,
      "step": 5925
    },
    {
      "entropy": 5.733428430557251,
      "epoch": 0.6660302128376481,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004965279308692602,
      "loss": 5.6484,
      "mean_token_accuracy": 0.1843220919370651,
      "num_tokens": 13223337.0,
      "step": 5930
    },
    {
      "entropy": 5.801119422912597,
      "epoch": 0.6665917897456056,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004965209013501919,
      "loss": 5.7285,
      "mean_token_accuracy": 0.17110735177993774,
      "num_tokens": 13234977.0,
      "step": 5935
    },
    {
      "entropy": 5.70616512298584,
      "epoch": 0.6671533666535632,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004965138647778122,
      "loss": 5.6768,
      "mean_token_accuracy": 0.17690228521823884,
      "num_tokens": 13245315.0,
      "step": 5940
    },
    {
      "entropy": 5.685105466842652,
      "epoch": 0.6677149435615207,
      "grad_norm": 1.453125,
      "learning_rate": 0.000496506821152345,
      "loss": 5.5582,
      "mean_token_accuracy": 0.1749824121594429,
      "num_tokens": 13255829.0,
      "step": 5945
    },
    {
      "entropy": 5.66689453125,
      "epoch": 0.6682765204694783,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004964997704740147,
      "loss": 5.6256,
      "mean_token_accuracy": 0.170874160528183,
      "num_tokens": 13266964.0,
      "step": 5950
    },
    {
      "entropy": 5.796186685562134,
      "epoch": 0.6688380973774358,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000496492712743046,
      "loss": 5.6614,
      "mean_token_accuracy": 0.1689567595720291,
      "num_tokens": 13279416.0,
      "step": 5955
    },
    {
      "entropy": 5.723680639266968,
      "epoch": 0.6693996742853934,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004964856479596632,
      "loss": 5.6919,
      "mean_token_accuracy": 0.1676722764968872,
      "num_tokens": 13290633.0,
      "step": 5960
    },
    {
      "entropy": 5.7067337989807125,
      "epoch": 0.6699612511933509,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004964785761240915,
      "loss": 5.5734,
      "mean_token_accuracy": 0.17108363509178162,
      "num_tokens": 13302236.0,
      "step": 5965
    },
    {
      "entropy": 5.7176580905914305,
      "epoch": 0.6705228281013085,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004964714972365562,
      "loss": 5.6724,
      "mean_token_accuracy": 0.17194082885980605,
      "num_tokens": 13315356.0,
      "step": 5970
    },
    {
      "entropy": 5.516757583618164,
      "epoch": 0.671084405009266,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004964644112972824,
      "loss": 5.5327,
      "mean_token_accuracy": 0.1755928486585617,
      "num_tokens": 13326037.0,
      "step": 5975
    },
    {
      "entropy": 5.632549905776978,
      "epoch": 0.6716459819172236,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004964573183064959,
      "loss": 5.5914,
      "mean_token_accuracy": 0.17360655069351197,
      "num_tokens": 13336278.0,
      "step": 5980
    },
    {
      "entropy": 5.736466360092163,
      "epoch": 0.6722075588251811,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004964502182644225,
      "loss": 5.7065,
      "mean_token_accuracy": 0.17162298858165742,
      "num_tokens": 13348147.0,
      "step": 5985
    },
    {
      "entropy": 5.71574764251709,
      "epoch": 0.6727691357331387,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004964431111712883,
      "loss": 5.6413,
      "mean_token_accuracy": 0.17189154624938965,
      "num_tokens": 13360181.0,
      "step": 5990
    },
    {
      "entropy": 5.774587678909302,
      "epoch": 0.6733307126410962,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004964359970273195,
      "loss": 5.7266,
      "mean_token_accuracy": 0.16633295267820358,
      "num_tokens": 13372258.0,
      "step": 5995
    },
    {
      "entropy": 5.71021146774292,
      "epoch": 0.6738922895490538,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004964288758327426,
      "loss": 5.6378,
      "mean_token_accuracy": 0.16402822285890578,
      "num_tokens": 13384332.0,
      "step": 6000
    },
    {
      "epoch": 0.6738922895490538,
      "eval_entropy": 5.5617586172269675,
      "eval_loss": 5.65424108505249,
      "eval_mean_token_accuracy": 0.17736698931749925,
      "eval_num_tokens": 13384332.0,
      "eval_runtime": 23.6897,
      "eval_samples_per_second": 1309.134,
      "eval_steps_per_second": 163.658,
      "step": 6000
    },
    {
      "entropy": 5.7171698093414305,
      "epoch": 0.6744538664570113,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004964217475877846,
      "loss": 5.6314,
      "mean_token_accuracy": 0.16942399442195893,
      "num_tokens": 13396345.0,
      "step": 6005
    },
    {
      "entropy": 5.591657829284668,
      "epoch": 0.6750154433649689,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000496414612292672,
      "loss": 5.4644,
      "mean_token_accuracy": 0.17689456790685654,
      "num_tokens": 13407325.0,
      "step": 6010
    },
    {
      "entropy": 5.689947080612183,
      "epoch": 0.6755770202729264,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004964074699476325,
      "loss": 5.6021,
      "mean_token_accuracy": 0.17259112149477004,
      "num_tokens": 13417476.0,
      "step": 6015
    },
    {
      "entropy": 5.656712484359741,
      "epoch": 0.6761385971808839,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004964003205528933,
      "loss": 5.6096,
      "mean_token_accuracy": 0.1726845309138298,
      "num_tokens": 13428792.0,
      "step": 6020
    },
    {
      "entropy": 5.700129842758178,
      "epoch": 0.6767001740888414,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004963931641086819,
      "loss": 5.6682,
      "mean_token_accuracy": 0.1713205248117447,
      "num_tokens": 13440043.0,
      "step": 6025
    },
    {
      "entropy": 5.7401378631591795,
      "epoch": 0.677261750996799,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004963860006152262,
      "loss": 5.6353,
      "mean_token_accuracy": 0.17090512961149215,
      "num_tokens": 13451637.0,
      "step": 6030
    },
    {
      "entropy": 5.824400568008423,
      "epoch": 0.6778233279047565,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004963788300727546,
      "loss": 5.6046,
      "mean_token_accuracy": 0.17313029766082763,
      "num_tokens": 13463205.0,
      "step": 6035
    },
    {
      "entropy": 5.660120248794556,
      "epoch": 0.6783849048127141,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004963716524814951,
      "loss": 5.6569,
      "mean_token_accuracy": 0.17166511565446854,
      "num_tokens": 13474201.0,
      "step": 6040
    },
    {
      "entropy": 5.62141170501709,
      "epoch": 0.6789464817206716,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004963644678416763,
      "loss": 5.5434,
      "mean_token_accuracy": 0.17378048449754716,
      "num_tokens": 13485439.0,
      "step": 6045
    },
    {
      "entropy": 5.730491065979004,
      "epoch": 0.6795080586286292,
      "grad_norm": 1.1875,
      "learning_rate": 0.000496357276153527,
      "loss": 5.7492,
      "mean_token_accuracy": 0.16852689236402513,
      "num_tokens": 13497257.0,
      "step": 6050
    },
    {
      "entropy": 5.730785655975342,
      "epoch": 0.6800696355365867,
      "grad_norm": 1.125,
      "learning_rate": 0.0004963500774172763,
      "loss": 5.5958,
      "mean_token_accuracy": 0.17269592732191086,
      "num_tokens": 13508689.0,
      "step": 6055
    },
    {
      "entropy": 5.698699808120727,
      "epoch": 0.6806312124445443,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000496342871633153,
      "loss": 5.5497,
      "mean_token_accuracy": 0.18048133552074433,
      "num_tokens": 13519392.0,
      "step": 6060
    },
    {
      "entropy": 5.62058744430542,
      "epoch": 0.6811927893525018,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004963356588013871,
      "loss": 5.605,
      "mean_token_accuracy": 0.1779961258172989,
      "num_tokens": 13530503.0,
      "step": 6065
    },
    {
      "entropy": 5.707237100601196,
      "epoch": 0.6817543662604594,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004963284389222078,
      "loss": 5.674,
      "mean_token_accuracy": 0.1740332290530205,
      "num_tokens": 13541668.0,
      "step": 6070
    },
    {
      "entropy": 5.699452447891235,
      "epoch": 0.6823159431684169,
      "grad_norm": 1.25,
      "learning_rate": 0.0004963212119958453,
      "loss": 5.5067,
      "mean_token_accuracy": 0.1787132978439331,
      "num_tokens": 13552692.0,
      "step": 6075
    },
    {
      "entropy": 5.577724456787109,
      "epoch": 0.6828775200763745,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004963139780225295,
      "loss": 5.6074,
      "mean_token_accuracy": 0.1706698417663574,
      "num_tokens": 13563813.0,
      "step": 6080
    },
    {
      "entropy": 5.6306346416473385,
      "epoch": 0.683439096984332,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004963067370024909,
      "loss": 5.5537,
      "mean_token_accuracy": 0.17718004286289216,
      "num_tokens": 13573818.0,
      "step": 6085
    },
    {
      "entropy": 5.760622024536133,
      "epoch": 0.6840006738922896,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004962994889359598,
      "loss": 5.6771,
      "mean_token_accuracy": 0.16787791550159453,
      "num_tokens": 13585416.0,
      "step": 6090
    },
    {
      "entropy": 5.631905651092529,
      "epoch": 0.6845622508002471,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004962922338231673,
      "loss": 5.5727,
      "mean_token_accuracy": 0.17555365860462188,
      "num_tokens": 13596645.0,
      "step": 6095
    },
    {
      "entropy": 5.68451681137085,
      "epoch": 0.6851238277082047,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004962849716643441,
      "loss": 5.5218,
      "mean_token_accuracy": 0.177066071331501,
      "num_tokens": 13607206.0,
      "step": 6100
    },
    {
      "entropy": 5.7633359909057615,
      "epoch": 0.6856854046161622,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004962777024597216,
      "loss": 5.7316,
      "mean_token_accuracy": 0.16855940222740173,
      "num_tokens": 13619418.0,
      "step": 6105
    },
    {
      "entropy": 5.643983459472656,
      "epoch": 0.6862469815241198,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004962704262095314,
      "loss": 5.5224,
      "mean_token_accuracy": 0.17984748929738997,
      "num_tokens": 13630098.0,
      "step": 6110
    },
    {
      "entropy": 5.625010013580322,
      "epoch": 0.6868085584320772,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004962631429140048,
      "loss": 5.6491,
      "mean_token_accuracy": 0.17171284705400466,
      "num_tokens": 13641401.0,
      "step": 6115
    },
    {
      "entropy": 5.739342164993286,
      "epoch": 0.6873701353400348,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000496255852573374,
      "loss": 5.6075,
      "mean_token_accuracy": 0.1670444905757904,
      "num_tokens": 13653099.0,
      "step": 6120
    },
    {
      "entropy": 5.795357084274292,
      "epoch": 0.6879317122479923,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004962485551878711,
      "loss": 5.622,
      "mean_token_accuracy": 0.16640905737876893,
      "num_tokens": 13665381.0,
      "step": 6125
    },
    {
      "entropy": 5.661467742919922,
      "epoch": 0.6884932891559499,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004962412507577283,
      "loss": 5.6346,
      "mean_token_accuracy": 0.17221850603818895,
      "num_tokens": 13676767.0,
      "step": 6130
    },
    {
      "entropy": 5.627943468093872,
      "epoch": 0.6890548660639074,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004962339392831783,
      "loss": 5.6228,
      "mean_token_accuracy": 0.17215524911880492,
      "num_tokens": 13688552.0,
      "step": 6135
    },
    {
      "entropy": 5.70760407447815,
      "epoch": 0.689616442971865,
      "grad_norm": 1.46875,
      "learning_rate": 0.000496226620764454,
      "loss": 5.5953,
      "mean_token_accuracy": 0.17065925300121307,
      "num_tokens": 13700160.0,
      "step": 6140
    },
    {
      "entropy": 5.671698188781738,
      "epoch": 0.6901780198798225,
      "grad_norm": 2.015625,
      "learning_rate": 0.000496219295201788,
      "loss": 5.6011,
      "mean_token_accuracy": 0.1736286088824272,
      "num_tokens": 13710828.0,
      "step": 6145
    },
    {
      "entropy": 5.744783687591553,
      "epoch": 0.6907395967877801,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004962119625954141,
      "loss": 5.7163,
      "mean_token_accuracy": 0.16494496762752534,
      "num_tokens": 13721912.0,
      "step": 6150
    },
    {
      "entropy": 5.773995733261108,
      "epoch": 0.6913011736957376,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004962046229455653,
      "loss": 5.7312,
      "mean_token_accuracy": 0.1703624963760376,
      "num_tokens": 13733662.0,
      "step": 6155
    },
    {
      "entropy": 5.625052499771118,
      "epoch": 0.6918627506036952,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004961972762524757,
      "loss": 5.549,
      "mean_token_accuracy": 0.1760735273361206,
      "num_tokens": 13743699.0,
      "step": 6160
    },
    {
      "entropy": 5.629493951797485,
      "epoch": 0.6924243275116527,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000496189922516379,
      "loss": 5.5554,
      "mean_token_accuracy": 0.17835567742586136,
      "num_tokens": 13754918.0,
      "step": 6165
    },
    {
      "entropy": 5.802048397064209,
      "epoch": 0.6929859044196103,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004961825617375093,
      "loss": 5.589,
      "mean_token_accuracy": 0.1810780867934227,
      "num_tokens": 13764432.0,
      "step": 6170
    },
    {
      "entropy": 5.56612720489502,
      "epoch": 0.6935474813275678,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004961751939161011,
      "loss": 5.5101,
      "mean_token_accuracy": 0.1665957048535347,
      "num_tokens": 13777049.0,
      "step": 6175
    },
    {
      "entropy": 5.613159084320069,
      "epoch": 0.6941090582355254,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000496167819052389,
      "loss": 5.6285,
      "mean_token_accuracy": 0.17187547385692598,
      "num_tokens": 13788324.0,
      "step": 6180
    },
    {
      "entropy": 5.774720954895019,
      "epoch": 0.6946706351434829,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004961604371466077,
      "loss": 5.7071,
      "mean_token_accuracy": 0.1715664029121399,
      "num_tokens": 13800443.0,
      "step": 6185
    },
    {
      "entropy": 5.813463592529297,
      "epoch": 0.6952322120514405,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004961530481989923,
      "loss": 5.7471,
      "mean_token_accuracy": 0.1697026953101158,
      "num_tokens": 13811982.0,
      "step": 6190
    },
    {
      "entropy": 5.664430713653564,
      "epoch": 0.695793788959398,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004961456522097782,
      "loss": 5.5307,
      "mean_token_accuracy": 0.17545289546251297,
      "num_tokens": 13822005.0,
      "step": 6195
    },
    {
      "entropy": 5.580399084091186,
      "epoch": 0.6963553658673556,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004961382491792006,
      "loss": 5.5458,
      "mean_token_accuracy": 0.1745297133922577,
      "num_tokens": 13831943.0,
      "step": 6200
    },
    {
      "entropy": 5.800264644622803,
      "epoch": 0.6969169427753131,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004961308391074953,
      "loss": 5.7382,
      "mean_token_accuracy": 0.16787344366312026,
      "num_tokens": 13842517.0,
      "step": 6205
    },
    {
      "entropy": 5.726728725433349,
      "epoch": 0.6974785196832706,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004961234219948986,
      "loss": 5.6109,
      "mean_token_accuracy": 0.1686943292617798,
      "num_tokens": 13853636.0,
      "step": 6210
    },
    {
      "entropy": 5.655916929244995,
      "epoch": 0.6980400965912281,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004961159978416461,
      "loss": 5.5513,
      "mean_token_accuracy": 0.1721298947930336,
      "num_tokens": 13864378.0,
      "step": 6215
    },
    {
      "entropy": 5.713779544830322,
      "epoch": 0.6986016734991857,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004961085666479746,
      "loss": 5.5778,
      "mean_token_accuracy": 0.17656196355819703,
      "num_tokens": 13874964.0,
      "step": 6220
    },
    {
      "entropy": 5.689186382293701,
      "epoch": 0.6991632504071432,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004961011284141206,
      "loss": 5.5644,
      "mean_token_accuracy": 0.16974303126335144,
      "num_tokens": 13885852.0,
      "step": 6225
    },
    {
      "entropy": 5.630877304077148,
      "epoch": 0.6997248273151008,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004960936831403209,
      "loss": 5.6134,
      "mean_token_accuracy": 0.17150009721517562,
      "num_tokens": 13896101.0,
      "step": 6230
    },
    {
      "entropy": 5.648992586135864,
      "epoch": 0.7002864042230583,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004960862308268126,
      "loss": 5.5052,
      "mean_token_accuracy": 0.17886769622564316,
      "num_tokens": 13907129.0,
      "step": 6235
    },
    {
      "entropy": 5.720948791503906,
      "epoch": 0.7008479811310159,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004960787714738329,
      "loss": 5.6951,
      "mean_token_accuracy": 0.1737249031662941,
      "num_tokens": 13918662.0,
      "step": 6240
    },
    {
      "entropy": 5.685187721252442,
      "epoch": 0.7014095580389734,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004960713050816194,
      "loss": 5.6562,
      "mean_token_accuracy": 0.17299176901578903,
      "num_tokens": 13930717.0,
      "step": 6245
    },
    {
      "entropy": 5.633158874511719,
      "epoch": 0.701971134946931,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004960638316504098,
      "loss": 5.5279,
      "mean_token_accuracy": 0.17130255252122878,
      "num_tokens": 13941299.0,
      "step": 6250
    },
    {
      "entropy": 5.780350589752198,
      "epoch": 0.7025327118548885,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004960563511804421,
      "loss": 5.6119,
      "mean_token_accuracy": 0.17323518246412278,
      "num_tokens": 13951807.0,
      "step": 6255
    },
    {
      "entropy": 5.650942611694336,
      "epoch": 0.7030942887628461,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004960488636719545,
      "loss": 5.5209,
      "mean_token_accuracy": 0.17925544530153276,
      "num_tokens": 13962994.0,
      "step": 6260
    },
    {
      "entropy": 5.566241312026977,
      "epoch": 0.7036558656708036,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004960413691251852,
      "loss": 5.5587,
      "mean_token_accuracy": 0.17189284712076186,
      "num_tokens": 13973731.0,
      "step": 6265
    },
    {
      "entropy": 5.686014032363891,
      "epoch": 0.7042174425787612,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004960338675403731,
      "loss": 5.6194,
      "mean_token_accuracy": 0.16850770264863968,
      "num_tokens": 13983772.0,
      "step": 6270
    },
    {
      "entropy": 5.6831377983093265,
      "epoch": 0.7047790194867187,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004960263589177568,
      "loss": 5.6285,
      "mean_token_accuracy": 0.17523877322673798,
      "num_tokens": 13994935.0,
      "step": 6275
    },
    {
      "entropy": 5.693459796905517,
      "epoch": 0.7053405963946763,
      "grad_norm": 1.625,
      "learning_rate": 0.0004960188432575757,
      "loss": 5.5889,
      "mean_token_accuracy": 0.17878624945878982,
      "num_tokens": 14006225.0,
      "step": 6280
    },
    {
      "entropy": 5.69208722114563,
      "epoch": 0.7059021733026338,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004960113205600688,
      "loss": 5.6025,
      "mean_token_accuracy": 0.18202067613601686,
      "num_tokens": 14017693.0,
      "step": 6285
    },
    {
      "entropy": 5.67942214012146,
      "epoch": 0.7064637502105914,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004960037908254757,
      "loss": 5.5771,
      "mean_token_accuracy": 0.17393409013748168,
      "num_tokens": 14027264.0,
      "step": 6290
    },
    {
      "entropy": 5.7325080871582035,
      "epoch": 0.7070253271185489,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004959962540540361,
      "loss": 5.6781,
      "mean_token_accuracy": 0.17509326040744783,
      "num_tokens": 14039422.0,
      "step": 6295
    },
    {
      "entropy": 5.694810819625855,
      "epoch": 0.7075869040265065,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004959887102459903,
      "loss": 5.6026,
      "mean_token_accuracy": 0.17309763431549072,
      "num_tokens": 14051619.0,
      "step": 6300
    },
    {
      "entropy": 5.518226671218872,
      "epoch": 0.7081484809344639,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004959811594015781,
      "loss": 5.6008,
      "mean_token_accuracy": 0.17118653059005737,
      "num_tokens": 14061752.0,
      "step": 6305
    },
    {
      "entropy": 5.7187727928161625,
      "epoch": 0.7087100578424215,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004959736015210401,
      "loss": 5.5499,
      "mean_token_accuracy": 0.17870223075151442,
      "num_tokens": 14072931.0,
      "step": 6310
    },
    {
      "entropy": 5.670480346679687,
      "epoch": 0.709271634750379,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004959660366046169,
      "loss": 5.6354,
      "mean_token_accuracy": 0.1812950477004051,
      "num_tokens": 14084137.0,
      "step": 6315
    },
    {
      "entropy": 5.733542203903198,
      "epoch": 0.7098332116583366,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004959584646525494,
      "loss": 5.643,
      "mean_token_accuracy": 0.16594423800706865,
      "num_tokens": 14096251.0,
      "step": 6320
    },
    {
      "entropy": 5.834647226333618,
      "epoch": 0.7103947885662941,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004959508856650788,
      "loss": 5.747,
      "mean_token_accuracy": 0.15990046858787538,
      "num_tokens": 14107844.0,
      "step": 6325
    },
    {
      "entropy": 5.701244258880616,
      "epoch": 0.7109563654742517,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004959432996424463,
      "loss": 5.5159,
      "mean_token_accuracy": 0.1764766052365303,
      "num_tokens": 14118344.0,
      "step": 6330
    },
    {
      "entropy": 5.683355712890625,
      "epoch": 0.7115179423822092,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004959357065848932,
      "loss": 5.5731,
      "mean_token_accuracy": 0.17560436129570006,
      "num_tokens": 14128767.0,
      "step": 6335
    },
    {
      "entropy": 5.676946210861206,
      "epoch": 0.7120795192901668,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004959281064926618,
      "loss": 5.7624,
      "mean_token_accuracy": 0.1625056892633438,
      "num_tokens": 14141593.0,
      "step": 6340
    },
    {
      "entropy": 5.787940835952758,
      "epoch": 0.7126410961981243,
      "grad_norm": 1.375,
      "learning_rate": 0.0004959204993659936,
      "loss": 5.658,
      "mean_token_accuracy": 0.16435816437005996,
      "num_tokens": 14151488.0,
      "step": 6345
    },
    {
      "entropy": 5.709534454345703,
      "epoch": 0.7132026731060819,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004959128852051311,
      "loss": 5.6446,
      "mean_token_accuracy": 0.17000896334648133,
      "num_tokens": 14162399.0,
      "step": 6350
    },
    {
      "entropy": 5.681110572814942,
      "epoch": 0.7137642500140394,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004959052640103167,
      "loss": 5.6154,
      "mean_token_accuracy": 0.1655770495533943,
      "num_tokens": 14173528.0,
      "step": 6355
    },
    {
      "entropy": 5.737189197540284,
      "epoch": 0.714325826921997,
      "grad_norm": 1.578125,
      "learning_rate": 0.000495897635781793,
      "loss": 5.5977,
      "mean_token_accuracy": 0.1738772988319397,
      "num_tokens": 14184119.0,
      "step": 6360
    },
    {
      "entropy": 5.783509731292725,
      "epoch": 0.7148874038299545,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004958900005198029,
      "loss": 5.6966,
      "mean_token_accuracy": 0.16666410118341446,
      "num_tokens": 14196225.0,
      "step": 6365
    },
    {
      "entropy": 5.5992560386657715,
      "epoch": 0.7154489807379121,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004958823582245894,
      "loss": 5.5122,
      "mean_token_accuracy": 0.1770908787846565,
      "num_tokens": 14207541.0,
      "step": 6370
    },
    {
      "entropy": 5.713044595718384,
      "epoch": 0.7160105576458696,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004958747088963961,
      "loss": 5.62,
      "mean_token_accuracy": 0.17931632846593856,
      "num_tokens": 14218713.0,
      "step": 6375
    },
    {
      "entropy": 5.576667070388794,
      "epoch": 0.7165721345538272,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004958670525354663,
      "loss": 5.4876,
      "mean_token_accuracy": 0.17556754648685455,
      "num_tokens": 14229626.0,
      "step": 6380
    },
    {
      "entropy": 5.67002272605896,
      "epoch": 0.7171337114617847,
      "grad_norm": 1.421875,
      "learning_rate": 0.000495859389142044,
      "loss": 5.6633,
      "mean_token_accuracy": 0.16710093021392822,
      "num_tokens": 14240320.0,
      "step": 6385
    },
    {
      "entropy": 5.830556774139405,
      "epoch": 0.7176952883697423,
      "grad_norm": 1.390625,
      "learning_rate": 0.000495851718716373,
      "loss": 5.719,
      "mean_token_accuracy": 0.17310031801462172,
      "num_tokens": 14252233.0,
      "step": 6390
    },
    {
      "entropy": 5.79133095741272,
      "epoch": 0.7182568652776998,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004958440412586977,
      "loss": 5.7307,
      "mean_token_accuracy": 0.15904613435268403,
      "num_tokens": 14264421.0,
      "step": 6395
    },
    {
      "entropy": 5.731802177429199,
      "epoch": 0.7188184421856574,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004958363567692624,
      "loss": 5.6031,
      "mean_token_accuracy": 0.17360225617885588,
      "num_tokens": 14275320.0,
      "step": 6400
    },
    {
      "entropy": 5.638220739364624,
      "epoch": 0.7193800190936148,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004958286652483118,
      "loss": 5.5668,
      "mean_token_accuracy": 0.1780466377735138,
      "num_tokens": 14286099.0,
      "step": 6405
    },
    {
      "entropy": 5.66323037147522,
      "epoch": 0.7199415960015724,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004958209666960909,
      "loss": 5.6835,
      "mean_token_accuracy": 0.16914550662040712,
      "num_tokens": 14297795.0,
      "step": 6410
    },
    {
      "entropy": 5.638380813598633,
      "epoch": 0.7205031729095299,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004958132611128448,
      "loss": 5.5828,
      "mean_token_accuracy": 0.17390582710504532,
      "num_tokens": 14309424.0,
      "step": 6415
    },
    {
      "entropy": 5.784746742248535,
      "epoch": 0.7210647498174875,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004958055484988188,
      "loss": 5.6623,
      "mean_token_accuracy": 0.16943105459213256,
      "num_tokens": 14320763.0,
      "step": 6420
    },
    {
      "entropy": 5.7342438220977785,
      "epoch": 0.721626326725445,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004957978288542585,
      "loss": 5.6264,
      "mean_token_accuracy": 0.1723559558391571,
      "num_tokens": 14331704.0,
      "step": 6425
    },
    {
      "entropy": 5.756981945037841,
      "epoch": 0.7221879036334026,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004957901021794098,
      "loss": 5.675,
      "mean_token_accuracy": 0.16486780792474748,
      "num_tokens": 14343694.0,
      "step": 6430
    },
    {
      "entropy": 5.680773448944092,
      "epoch": 0.7227494805413601,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004957823684745184,
      "loss": 5.572,
      "mean_token_accuracy": 0.17167100310325623,
      "num_tokens": 14355408.0,
      "step": 6435
    },
    {
      "entropy": 5.703755664825439,
      "epoch": 0.7233110574493177,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004957746277398309,
      "loss": 5.7254,
      "mean_token_accuracy": 0.16364069730043412,
      "num_tokens": 14366778.0,
      "step": 6440
    },
    {
      "entropy": 5.712000751495362,
      "epoch": 0.7238726343572752,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004957668799755937,
      "loss": 5.5799,
      "mean_token_accuracy": 0.17299263626337053,
      "num_tokens": 14377947.0,
      "step": 6445
    },
    {
      "entropy": 5.705758762359619,
      "epoch": 0.7244342112652328,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004957591251820532,
      "loss": 5.6688,
      "mean_token_accuracy": 0.1714080259203911,
      "num_tokens": 14389571.0,
      "step": 6450
    },
    {
      "entropy": 5.8451415538787845,
      "epoch": 0.7249957881731903,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004957513633594567,
      "loss": 5.7455,
      "mean_token_accuracy": 0.1655985251069069,
      "num_tokens": 14401865.0,
      "step": 6455
    },
    {
      "entropy": 5.643240690231323,
      "epoch": 0.7255573650811479,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004957435945080512,
      "loss": 5.576,
      "mean_token_accuracy": 0.17257309556007386,
      "num_tokens": 14413676.0,
      "step": 6460
    },
    {
      "entropy": 5.6537106990814205,
      "epoch": 0.7261189419891054,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004957358186280839,
      "loss": 5.6016,
      "mean_token_accuracy": 0.1794086754322052,
      "num_tokens": 14424675.0,
      "step": 6465
    },
    {
      "entropy": 5.618784666061401,
      "epoch": 0.726680518897063,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004957280357198026,
      "loss": 5.5149,
      "mean_token_accuracy": 0.17647840976715087,
      "num_tokens": 14437238.0,
      "step": 6470
    },
    {
      "entropy": 5.549806547164917,
      "epoch": 0.7272420958050205,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004957202457834551,
      "loss": 5.4673,
      "mean_token_accuracy": 0.17938644886016847,
      "num_tokens": 14448003.0,
      "step": 6475
    },
    {
      "entropy": 5.528613138198852,
      "epoch": 0.7278036727129781,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004957124488192893,
      "loss": 5.4236,
      "mean_token_accuracy": 0.18780238777399064,
      "num_tokens": 14458440.0,
      "step": 6480
    },
    {
      "entropy": 5.688983345031739,
      "epoch": 0.7283652496209356,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004957046448275535,
      "loss": 5.6076,
      "mean_token_accuracy": 0.17096126079559326,
      "num_tokens": 14470127.0,
      "step": 6485
    },
    {
      "entropy": 5.652023792266846,
      "epoch": 0.7289268265288932,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004956968338084961,
      "loss": 5.5194,
      "mean_token_accuracy": 0.17799320966005325,
      "num_tokens": 14480993.0,
      "step": 6490
    },
    {
      "entropy": 5.667353630065918,
      "epoch": 0.7294884034368507,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004956890157623661,
      "loss": 5.6216,
      "mean_token_accuracy": 0.16608051508665084,
      "num_tokens": 14492789.0,
      "step": 6495
    },
    {
      "entropy": 5.735092353820801,
      "epoch": 0.7300499803448082,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004956811906894122,
      "loss": 5.7142,
      "mean_token_accuracy": 0.16898949593305587,
      "num_tokens": 14504528.0,
      "step": 6500
    },
    {
      "entropy": 5.777708959579468,
      "epoch": 0.7306115572527657,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004956733585898836,
      "loss": 5.6052,
      "mean_token_accuracy": 0.17461458444595337,
      "num_tokens": 14515739.0,
      "step": 6505
    },
    {
      "entropy": 5.7372722148895265,
      "epoch": 0.7311731341607233,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004956655194640297,
      "loss": 5.7253,
      "mean_token_accuracy": 0.165912227332592,
      "num_tokens": 14527120.0,
      "step": 6510
    },
    {
      "entropy": 5.734869432449341,
      "epoch": 0.7317347110686808,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004956576733121,
      "loss": 5.6732,
      "mean_token_accuracy": 0.17188447639346122,
      "num_tokens": 14539390.0,
      "step": 6515
    },
    {
      "entropy": 5.7428192615509035,
      "epoch": 0.7322962879766384,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004956498201343444,
      "loss": 5.6542,
      "mean_token_accuracy": 0.17126702666282653,
      "num_tokens": 14549905.0,
      "step": 6520
    },
    {
      "entropy": 5.7277122974395756,
      "epoch": 0.7328578648845959,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000495641959931013,
      "loss": 5.7013,
      "mean_token_accuracy": 0.17059755474328994,
      "num_tokens": 14561519.0,
      "step": 6525
    },
    {
      "entropy": 5.656032657623291,
      "epoch": 0.7334194417925535,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004956340927023561,
      "loss": 5.4721,
      "mean_token_accuracy": 0.17097557485103607,
      "num_tokens": 14572704.0,
      "step": 6530
    },
    {
      "entropy": 5.733268117904663,
      "epoch": 0.733981018700511,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000495626218448624,
      "loss": 5.6477,
      "mean_token_accuracy": 0.17309343367815017,
      "num_tokens": 14584055.0,
      "step": 6535
    },
    {
      "entropy": 5.528829097747803,
      "epoch": 0.7345425956084686,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004956183371700676,
      "loss": 5.4578,
      "mean_token_accuracy": 0.17761017978191376,
      "num_tokens": 14595271.0,
      "step": 6540
    },
    {
      "entropy": 5.687887334823609,
      "epoch": 0.7351041725164261,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004956104488669377,
      "loss": 5.6217,
      "mean_token_accuracy": 0.1771188572049141,
      "num_tokens": 14605748.0,
      "step": 6545
    },
    {
      "entropy": 5.658285093307495,
      "epoch": 0.7356657494243837,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004956025535394856,
      "loss": 5.5326,
      "mean_token_accuracy": 0.18142949491739274,
      "num_tokens": 14616295.0,
      "step": 6550
    },
    {
      "entropy": 5.633347415924073,
      "epoch": 0.7362273263323412,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004955946511879626,
      "loss": 5.625,
      "mean_token_accuracy": 0.16703985631465912,
      "num_tokens": 14626690.0,
      "step": 6555
    },
    {
      "entropy": 5.678591918945313,
      "epoch": 0.7367889032402988,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004955867418126205,
      "loss": 5.6293,
      "mean_token_accuracy": 0.17487271279096603,
      "num_tokens": 14637685.0,
      "step": 6560
    },
    {
      "entropy": 5.738638973236084,
      "epoch": 0.7373504801482563,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004955788254137108,
      "loss": 5.5306,
      "mean_token_accuracy": 0.17113898396492006,
      "num_tokens": 14648098.0,
      "step": 6565
    },
    {
      "entropy": 5.746050071716309,
      "epoch": 0.7379120570562139,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004955709019914857,
      "loss": 5.6727,
      "mean_token_accuracy": 0.16656773388385773,
      "num_tokens": 14661134.0,
      "step": 6570
    },
    {
      "entropy": 5.746538877487183,
      "epoch": 0.7384736339641714,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004955629715461977,
      "loss": 5.6445,
      "mean_token_accuracy": 0.17216980457305908,
      "num_tokens": 14672699.0,
      "step": 6575
    },
    {
      "entropy": 5.747246170043946,
      "epoch": 0.739035210872129,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004955550340780991,
      "loss": 5.6143,
      "mean_token_accuracy": 0.16882283240556717,
      "num_tokens": 14682994.0,
      "step": 6580
    },
    {
      "entropy": 5.672818374633789,
      "epoch": 0.7395967877800865,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004955470895874427,
      "loss": 5.6186,
      "mean_token_accuracy": 0.17457711100578308,
      "num_tokens": 14694828.0,
      "step": 6585
    },
    {
      "entropy": 5.648428583145142,
      "epoch": 0.7401583646880441,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004955391380744813,
      "loss": 5.6108,
      "mean_token_accuracy": 0.17359036207199097,
      "num_tokens": 14705975.0,
      "step": 6590
    },
    {
      "entropy": 5.726032495498657,
      "epoch": 0.7407199415960015,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004955311795394684,
      "loss": 5.5944,
      "mean_token_accuracy": 0.1734750136733055,
      "num_tokens": 14717717.0,
      "step": 6595
    },
    {
      "entropy": 5.711815071105957,
      "epoch": 0.7412815185039591,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000495523213982657,
      "loss": 5.5662,
      "mean_token_accuracy": 0.1741805464029312,
      "num_tokens": 14729254.0,
      "step": 6600
    },
    {
      "entropy": 5.6931883811950685,
      "epoch": 0.7418430954119166,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004955152414043011,
      "loss": 5.7119,
      "mean_token_accuracy": 0.1718486949801445,
      "num_tokens": 14739733.0,
      "step": 6605
    },
    {
      "entropy": 5.752079057693481,
      "epoch": 0.7424046723198742,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004955072618046543,
      "loss": 5.6666,
      "mean_token_accuracy": 0.17087932378053666,
      "num_tokens": 14750855.0,
      "step": 6610
    },
    {
      "entropy": 5.71418023109436,
      "epoch": 0.7429662492278317,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004954992751839708,
      "loss": 5.6614,
      "mean_token_accuracy": 0.1698993593454361,
      "num_tokens": 14762762.0,
      "step": 6615
    },
    {
      "entropy": 5.688072967529297,
      "epoch": 0.7435278261357893,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004954912815425048,
      "loss": 5.5299,
      "mean_token_accuracy": 0.1700862854719162,
      "num_tokens": 14773690.0,
      "step": 6620
    },
    {
      "entropy": 5.640633058547974,
      "epoch": 0.7440894030437468,
      "grad_norm": 1.390625,
      "learning_rate": 0.000495483280880511,
      "loss": 5.6447,
      "mean_token_accuracy": 0.17218244969844818,
      "num_tokens": 14785019.0,
      "step": 6625
    },
    {
      "entropy": 5.643097877502441,
      "epoch": 0.7446509799517044,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000495475273198244,
      "loss": 5.6206,
      "mean_token_accuracy": 0.16789701133966445,
      "num_tokens": 14795638.0,
      "step": 6630
    },
    {
      "entropy": 5.715163087844848,
      "epoch": 0.7452125568596619,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004954672584959588,
      "loss": 5.549,
      "mean_token_accuracy": 0.17813322842121124,
      "num_tokens": 14806546.0,
      "step": 6635
    },
    {
      "entropy": 5.596341991424561,
      "epoch": 0.7457741337676195,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004954592367739104,
      "loss": 5.5626,
      "mean_token_accuracy": 0.17497475147247316,
      "num_tokens": 14816918.0,
      "step": 6640
    },
    {
      "entropy": 5.6133809089660645,
      "epoch": 0.746335710675577,
      "grad_norm": 1.75,
      "learning_rate": 0.0004954512080323546,
      "loss": 5.5571,
      "mean_token_accuracy": 0.18172363042831421,
      "num_tokens": 14827420.0,
      "step": 6645
    },
    {
      "entropy": 5.7528026580810545,
      "epoch": 0.7468972875835346,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004954431722715468,
      "loss": 5.7731,
      "mean_token_accuracy": 0.17284041196107863,
      "num_tokens": 14839209.0,
      "step": 6650
    },
    {
      "entropy": 5.772123146057129,
      "epoch": 0.7474588644914921,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004954351294917428,
      "loss": 5.6546,
      "mean_token_accuracy": 0.16983917206525803,
      "num_tokens": 14849916.0,
      "step": 6655
    },
    {
      "entropy": 5.593883991241455,
      "epoch": 0.7480204413994497,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004954270796931989,
      "loss": 5.5467,
      "mean_token_accuracy": 0.17545504868030548,
      "num_tokens": 14860946.0,
      "step": 6660
    },
    {
      "entropy": 5.773710012435913,
      "epoch": 0.7485820183074072,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004954190228761712,
      "loss": 5.6383,
      "mean_token_accuracy": 0.16717524081468582,
      "num_tokens": 14872750.0,
      "step": 6665
    },
    {
      "entropy": 5.68846116065979,
      "epoch": 0.7491435952153648,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004954109590409164,
      "loss": 5.6227,
      "mean_token_accuracy": 0.17093631327152253,
      "num_tokens": 14883714.0,
      "step": 6670
    },
    {
      "entropy": 5.633935451507568,
      "epoch": 0.7497051721233223,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004954028881876911,
      "loss": 5.6325,
      "mean_token_accuracy": 0.17225925624370575,
      "num_tokens": 14895147.0,
      "step": 6675
    },
    {
      "entropy": 5.698961019515991,
      "epoch": 0.7502667490312799,
      "grad_norm": 1.375,
      "learning_rate": 0.0004953948103167524,
      "loss": 5.5536,
      "mean_token_accuracy": 0.1762816697359085,
      "num_tokens": 14905914.0,
      "step": 6680
    },
    {
      "entropy": 5.582836866378784,
      "epoch": 0.7508283259392374,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004953867254283575,
      "loss": 5.5428,
      "mean_token_accuracy": 0.17713676244020463,
      "num_tokens": 14916988.0,
      "step": 6685
    },
    {
      "entropy": 5.6897834777832035,
      "epoch": 0.7513899028471949,
      "grad_norm": 1.5,
      "learning_rate": 0.0004953786335227636,
      "loss": 5.6518,
      "mean_token_accuracy": 0.17040350288152695,
      "num_tokens": 14928675.0,
      "step": 6690
    },
    {
      "entropy": 5.795252656936645,
      "epoch": 0.7519514797551524,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004953705346002287,
      "loss": 5.6581,
      "mean_token_accuracy": 0.17117824405431747,
      "num_tokens": 14939422.0,
      "step": 6695
    },
    {
      "entropy": 5.6437677383422855,
      "epoch": 0.75251305666311,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004953624286610106,
      "loss": 5.6086,
      "mean_token_accuracy": 0.17704180479049683,
      "num_tokens": 14950156.0,
      "step": 6700
    },
    {
      "entropy": 5.65733003616333,
      "epoch": 0.7530746335710675,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004953543157053672,
      "loss": 5.5983,
      "mean_token_accuracy": 0.1764406979084015,
      "num_tokens": 14960535.0,
      "step": 6705
    },
    {
      "entropy": 5.767850303649903,
      "epoch": 0.7536362104790251,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004953461957335569,
      "loss": 5.6239,
      "mean_token_accuracy": 0.1750498056411743,
      "num_tokens": 14970975.0,
      "step": 6710
    },
    {
      "entropy": 5.632910871505738,
      "epoch": 0.7541977873869826,
      "grad_norm": 1.375,
      "learning_rate": 0.0004953380687458382,
      "loss": 5.4731,
      "mean_token_accuracy": 0.18237796127796174,
      "num_tokens": 14981204.0,
      "step": 6715
    },
    {
      "entropy": 5.653511953353882,
      "epoch": 0.7547593642949402,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004953299347424701,
      "loss": 5.6468,
      "mean_token_accuracy": 0.17152471095323563,
      "num_tokens": 14991631.0,
      "step": 6720
    },
    {
      "entropy": 5.657448101043701,
      "epoch": 0.7553209412028977,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004953217937237114,
      "loss": 5.5735,
      "mean_token_accuracy": 0.17640465199947358,
      "num_tokens": 15003873.0,
      "step": 6725
    },
    {
      "entropy": 5.69847002029419,
      "epoch": 0.7558825181108553,
      "grad_norm": 1.5,
      "learning_rate": 0.0004953136456898212,
      "loss": 5.6493,
      "mean_token_accuracy": 0.17125397473573684,
      "num_tokens": 15014890.0,
      "step": 6730
    },
    {
      "entropy": 5.604665327072143,
      "epoch": 0.7564440950188128,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004953054906410593,
      "loss": 5.5194,
      "mean_token_accuracy": 0.18065904527902604,
      "num_tokens": 15025037.0,
      "step": 6735
    },
    {
      "entropy": 5.690511035919189,
      "epoch": 0.7570056719267704,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004952973285776848,
      "loss": 5.6233,
      "mean_token_accuracy": 0.17375240176916124,
      "num_tokens": 15035722.0,
      "step": 6740
    },
    {
      "entropy": 5.696527576446533,
      "epoch": 0.7575672488347279,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004952891594999582,
      "loss": 5.6224,
      "mean_token_accuracy": 0.16907656490802764,
      "num_tokens": 15047138.0,
      "step": 6745
    },
    {
      "entropy": 5.616460132598877,
      "epoch": 0.7581288257426855,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004952809834081392,
      "loss": 5.5799,
      "mean_token_accuracy": 0.17677342891693115,
      "num_tokens": 15058559.0,
      "step": 6750
    },
    {
      "entropy": 5.669347715377808,
      "epoch": 0.758690402650643,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004952728003024883,
      "loss": 5.6034,
      "mean_token_accuracy": 0.17386410832405091,
      "num_tokens": 15070481.0,
      "step": 6755
    },
    {
      "entropy": 5.746075916290283,
      "epoch": 0.7592519795586006,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004952646101832659,
      "loss": 5.525,
      "mean_token_accuracy": 0.1773831456899643,
      "num_tokens": 15079844.0,
      "step": 6760
    },
    {
      "entropy": 5.575839519500732,
      "epoch": 0.7598135564665581,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004952564130507331,
      "loss": 5.5459,
      "mean_token_accuracy": 0.1829729363322258,
      "num_tokens": 15091197.0,
      "step": 6765
    },
    {
      "entropy": 5.63728141784668,
      "epoch": 0.7603751333745157,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004952482089051506,
      "loss": 5.6395,
      "mean_token_accuracy": 0.16757640540599822,
      "num_tokens": 15102130.0,
      "step": 6770
    },
    {
      "entropy": 5.80513825416565,
      "epoch": 0.7609367102824732,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004952399977467796,
      "loss": 5.7159,
      "mean_token_accuracy": 0.164860138297081,
      "num_tokens": 15113886.0,
      "step": 6775
    },
    {
      "entropy": 5.695002603530884,
      "epoch": 0.7614982871904308,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004952317795758817,
      "loss": 5.5309,
      "mean_token_accuracy": 0.1707255095243454,
      "num_tokens": 15125284.0,
      "step": 6780
    },
    {
      "entropy": 5.693516254425049,
      "epoch": 0.7620598640983883,
      "grad_norm": 1.375,
      "learning_rate": 0.0004952235543927185,
      "loss": 5.6013,
      "mean_token_accuracy": 0.17872631102800368,
      "num_tokens": 15136486.0,
      "step": 6785
    },
    {
      "entropy": 5.5827912330627445,
      "epoch": 0.7626214410063458,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000495215322197552,
      "loss": 5.4942,
      "mean_token_accuracy": 0.1766755223274231,
      "num_tokens": 15145937.0,
      "step": 6790
    },
    {
      "entropy": 5.606060981750488,
      "epoch": 0.7631830179143033,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004952070829906442,
      "loss": 5.5703,
      "mean_token_accuracy": 0.17622093558311464,
      "num_tokens": 15156923.0,
      "step": 6795
    },
    {
      "entropy": 5.764186716079712,
      "epoch": 0.7637445948222609,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004951988367722574,
      "loss": 5.701,
      "mean_token_accuracy": 0.17148951441049576,
      "num_tokens": 15169793.0,
      "step": 6800
    },
    {
      "entropy": 5.666046524047852,
      "epoch": 0.7643061717302184,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004951905835426544,
      "loss": 5.4701,
      "mean_token_accuracy": 0.18407936692237853,
      "num_tokens": 15180679.0,
      "step": 6805
    },
    {
      "entropy": 5.658257055282593,
      "epoch": 0.764867748638176,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004951823233020976,
      "loss": 5.6771,
      "mean_token_accuracy": 0.1721682906150818,
      "num_tokens": 15192298.0,
      "step": 6810
    },
    {
      "entropy": 5.657361793518066,
      "epoch": 0.7654293255461335,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004951740560508504,
      "loss": 5.7427,
      "mean_token_accuracy": 0.1654340222477913,
      "num_tokens": 15204759.0,
      "step": 6815
    },
    {
      "entropy": 5.795834302902222,
      "epoch": 0.7659909024540911,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004951657817891758,
      "loss": 5.6181,
      "mean_token_accuracy": 0.17172770649194719,
      "num_tokens": 15217166.0,
      "step": 6820
    },
    {
      "entropy": 5.596086502075195,
      "epoch": 0.7665524793620486,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004951575005173373,
      "loss": 5.5908,
      "mean_token_accuracy": 0.17772889286279678,
      "num_tokens": 15227022.0,
      "step": 6825
    },
    {
      "entropy": 5.7071445941925045,
      "epoch": 0.7671140562700062,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004951492122355986,
      "loss": 5.6438,
      "mean_token_accuracy": 0.16916121244430543,
      "num_tokens": 15239088.0,
      "step": 6830
    },
    {
      "entropy": 5.74067268371582,
      "epoch": 0.7676756331779637,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004951409169442236,
      "loss": 5.6636,
      "mean_token_accuracy": 0.17377155274152756,
      "num_tokens": 15250263.0,
      "step": 6835
    },
    {
      "entropy": 5.614690923690796,
      "epoch": 0.7682372100859213,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004951326146434765,
      "loss": 5.56,
      "mean_token_accuracy": 0.17508067339658737,
      "num_tokens": 15263014.0,
      "step": 6840
    },
    {
      "entropy": 5.698054409027099,
      "epoch": 0.7687987869938788,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004951243053336214,
      "loss": 5.5488,
      "mean_token_accuracy": 0.17616119384765624,
      "num_tokens": 15274261.0,
      "step": 6845
    },
    {
      "entropy": 5.617717838287353,
      "epoch": 0.7693603639018364,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004951159890149232,
      "loss": 5.4812,
      "mean_token_accuracy": 0.17868284434080123,
      "num_tokens": 15284929.0,
      "step": 6850
    },
    {
      "entropy": 5.601969814300537,
      "epoch": 0.7699219408097939,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004951076656876466,
      "loss": 5.5905,
      "mean_token_accuracy": 0.1784390926361084,
      "num_tokens": 15296622.0,
      "step": 6855
    },
    {
      "entropy": 5.671678590774536,
      "epoch": 0.7704835177177515,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004950993353520564,
      "loss": 5.6416,
      "mean_token_accuracy": 0.16959952116012572,
      "num_tokens": 15308373.0,
      "step": 6860
    },
    {
      "entropy": 5.633020448684692,
      "epoch": 0.771045094625709,
      "grad_norm": 1.515625,
      "learning_rate": 0.000495090998008418,
      "loss": 5.5145,
      "mean_token_accuracy": 0.18185887783765792,
      "num_tokens": 15319705.0,
      "step": 6865
    },
    {
      "entropy": 5.718202543258667,
      "epoch": 0.7716066715336666,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000495082653656997,
      "loss": 5.6745,
      "mean_token_accuracy": 0.16872720867395402,
      "num_tokens": 15330883.0,
      "step": 6870
    },
    {
      "entropy": 5.695891237258911,
      "epoch": 0.7721682484416241,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004950743022980587,
      "loss": 5.5826,
      "mean_token_accuracy": 0.17275900691747664,
      "num_tokens": 15342294.0,
      "step": 6875
    },
    {
      "entropy": 5.698135662078857,
      "epoch": 0.7727298253495817,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004950659439318696,
      "loss": 5.5901,
      "mean_token_accuracy": 0.1722102403640747,
      "num_tokens": 15352615.0,
      "step": 6880
    },
    {
      "entropy": 5.6508818626403805,
      "epoch": 0.7732914022575391,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004950575785586952,
      "loss": 5.6341,
      "mean_token_accuracy": 0.168660968542099,
      "num_tokens": 15364516.0,
      "step": 6885
    },
    {
      "entropy": 5.67797999382019,
      "epoch": 0.7738529791654967,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004950492061788023,
      "loss": 5.527,
      "mean_token_accuracy": 0.173749241232872,
      "num_tokens": 15374877.0,
      "step": 6890
    },
    {
      "entropy": 5.673302030563354,
      "epoch": 0.7744145560734542,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004950408267924573,
      "loss": 5.4688,
      "mean_token_accuracy": 0.18849669992923737,
      "num_tokens": 15385250.0,
      "step": 6895
    },
    {
      "entropy": 5.630706882476806,
      "epoch": 0.7749761329814118,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004950324403999271,
      "loss": 5.6217,
      "mean_token_accuracy": 0.178818379342556,
      "num_tokens": 15397074.0,
      "step": 6900
    },
    {
      "entropy": 5.6523487091064455,
      "epoch": 0.7755377098893693,
      "grad_norm": 1.5,
      "learning_rate": 0.0004950240470014785,
      "loss": 5.6633,
      "mean_token_accuracy": 0.16847054213285445,
      "num_tokens": 15408575.0,
      "step": 6905
    },
    {
      "entropy": 5.721060371398925,
      "epoch": 0.7760992867973269,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004950156465973791,
      "loss": 5.5786,
      "mean_token_accuracy": 0.17714389264583588,
      "num_tokens": 15419071.0,
      "step": 6910
    },
    {
      "entropy": 5.595917701721191,
      "epoch": 0.7766608637052844,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000495007239187896,
      "loss": 5.4627,
      "mean_token_accuracy": 0.1753948897123337,
      "num_tokens": 15430137.0,
      "step": 6915
    },
    {
      "entropy": 5.549996185302734,
      "epoch": 0.777222440613242,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004949988247732972,
      "loss": 5.536,
      "mean_token_accuracy": 0.18008177876472473,
      "num_tokens": 15440661.0,
      "step": 6920
    },
    {
      "entropy": 5.577071666717529,
      "epoch": 0.7777840175211995,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004949904033538504,
      "loss": 5.5443,
      "mean_token_accuracy": 0.18011982291936873,
      "num_tokens": 15451909.0,
      "step": 6925
    },
    {
      "entropy": 5.744392728805542,
      "epoch": 0.7783455944291571,
      "grad_norm": 1.390625,
      "learning_rate": 0.000494981974929824,
      "loss": 5.6133,
      "mean_token_accuracy": 0.17083452343940736,
      "num_tokens": 15462244.0,
      "step": 6930
    },
    {
      "entropy": 5.644606256484986,
      "epoch": 0.7789071713371146,
      "grad_norm": 1.375,
      "learning_rate": 0.000494973539501486,
      "loss": 5.6585,
      "mean_token_accuracy": 0.17281380891799927,
      "num_tokens": 15473284.0,
      "step": 6935
    },
    {
      "entropy": 5.587286949157715,
      "epoch": 0.7794687482450722,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004949650970691052,
      "loss": 5.5591,
      "mean_token_accuracy": 0.1845861554145813,
      "num_tokens": 15484964.0,
      "step": 6940
    },
    {
      "entropy": 5.646401405334473,
      "epoch": 0.7800303251530297,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004949566476329505,
      "loss": 5.5877,
      "mean_token_accuracy": 0.17619993090629577,
      "num_tokens": 15495922.0,
      "step": 6945
    },
    {
      "entropy": 5.780960130691528,
      "epoch": 0.7805919020609873,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004949481911932908,
      "loss": 5.6492,
      "mean_token_accuracy": 0.17256713807582855,
      "num_tokens": 15507173.0,
      "step": 6950
    },
    {
      "entropy": 5.70377368927002,
      "epoch": 0.7811534789689448,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004949397277503954,
      "loss": 5.6209,
      "mean_token_accuracy": 0.17474339306354522,
      "num_tokens": 15517720.0,
      "step": 6955
    },
    {
      "entropy": 5.71235990524292,
      "epoch": 0.7817150558769024,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004949312573045338,
      "loss": 5.6394,
      "mean_token_accuracy": 0.1703775092959404,
      "num_tokens": 15529482.0,
      "step": 6960
    },
    {
      "entropy": 5.76483187675476,
      "epoch": 0.7822766327848599,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004949227798559757,
      "loss": 5.6429,
      "mean_token_accuracy": 0.1658545657992363,
      "num_tokens": 15540522.0,
      "step": 6965
    },
    {
      "entropy": 5.700478982925415,
      "epoch": 0.7828382096928175,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004949142954049909,
      "loss": 5.6267,
      "mean_token_accuracy": 0.17000363171100616,
      "num_tokens": 15552443.0,
      "step": 6970
    },
    {
      "entropy": 5.766734600067139,
      "epoch": 0.783399786600775,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004949058039518496,
      "loss": 5.664,
      "mean_token_accuracy": 0.17442309558391572,
      "num_tokens": 15563418.0,
      "step": 6975
    },
    {
      "entropy": 5.721600914001465,
      "epoch": 0.7839613635087325,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004948973054968222,
      "loss": 5.6809,
      "mean_token_accuracy": 0.1688017174601555,
      "num_tokens": 15574301.0,
      "step": 6980
    },
    {
      "entropy": 5.780105543136597,
      "epoch": 0.78452294041669,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004948888000401794,
      "loss": 5.5844,
      "mean_token_accuracy": 0.1704667940735817,
      "num_tokens": 15585125.0,
      "step": 6985
    },
    {
      "entropy": 5.7562014102935795,
      "epoch": 0.7850845173246476,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004948802875821918,
      "loss": 5.6937,
      "mean_token_accuracy": 0.17066754549741744,
      "num_tokens": 15595822.0,
      "step": 6990
    },
    {
      "entropy": 5.5500236511230465,
      "epoch": 0.7856460942326051,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004948717681231308,
      "loss": 5.5279,
      "mean_token_accuracy": 0.17664736062288283,
      "num_tokens": 15606835.0,
      "step": 6995
    },
    {
      "entropy": 5.632187032699585,
      "epoch": 0.7862076711405627,
      "grad_norm": 3.125,
      "learning_rate": 0.0004948632416632672,
      "loss": 5.6124,
      "mean_token_accuracy": 0.17274107187986373,
      "num_tokens": 15617287.0,
      "step": 7000
    },
    {
      "entropy": 5.8710390567779545,
      "epoch": 0.7867692480485202,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004948547082028728,
      "loss": 5.6573,
      "mean_token_accuracy": 0.16424071788787842,
      "num_tokens": 15628440.0,
      "step": 7005
    },
    {
      "entropy": 5.7740159034729,
      "epoch": 0.7873308249564778,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004948461677422193,
      "loss": 5.7659,
      "mean_token_accuracy": 0.16086588203907012,
      "num_tokens": 15640677.0,
      "step": 7010
    },
    {
      "entropy": 5.702903985977173,
      "epoch": 0.7878924018644353,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004948376202815785,
      "loss": 5.5799,
      "mean_token_accuracy": 0.17863162457942963,
      "num_tokens": 15651941.0,
      "step": 7015
    },
    {
      "entropy": 5.643584585189819,
      "epoch": 0.7884539787723929,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004948290658212225,
      "loss": 5.5872,
      "mean_token_accuracy": 0.1744334802031517,
      "num_tokens": 15663513.0,
      "step": 7020
    },
    {
      "entropy": 5.612797451019287,
      "epoch": 0.7890155556803504,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004948205043614239,
      "loss": 5.5694,
      "mean_token_accuracy": 0.17362559586763382,
      "num_tokens": 15675548.0,
      "step": 7025
    },
    {
      "entropy": 5.7275231838226315,
      "epoch": 0.789577132588308,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004948119359024551,
      "loss": 5.6234,
      "mean_token_accuracy": 0.17916473597288132,
      "num_tokens": 15685365.0,
      "step": 7030
    },
    {
      "entropy": 5.686577701568604,
      "epoch": 0.7901387094962655,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004948033604445891,
      "loss": 5.6126,
      "mean_token_accuracy": 0.17018286287784576,
      "num_tokens": 15696112.0,
      "step": 7035
    },
    {
      "entropy": 5.621775579452515,
      "epoch": 0.7907002864042231,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004947947779880988,
      "loss": 5.5484,
      "mean_token_accuracy": 0.1771180063486099,
      "num_tokens": 15707414.0,
      "step": 7040
    },
    {
      "entropy": 5.734884214401245,
      "epoch": 0.7912618633121806,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004947861885332576,
      "loss": 5.6507,
      "mean_token_accuracy": 0.17650367319583893,
      "num_tokens": 15719077.0,
      "step": 7045
    },
    {
      "entropy": 5.712475728988648,
      "epoch": 0.7918234402201382,
      "grad_norm": 1.375,
      "learning_rate": 0.0004947775920803388,
      "loss": 5.6426,
      "mean_token_accuracy": 0.1715145543217659,
      "num_tokens": 15730497.0,
      "step": 7050
    },
    {
      "entropy": 5.666735076904297,
      "epoch": 0.7923850171280957,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004947689886296162,
      "loss": 5.5464,
      "mean_token_accuracy": 0.1741068556904793,
      "num_tokens": 15741724.0,
      "step": 7055
    },
    {
      "entropy": 5.691272640228272,
      "epoch": 0.7929465940360533,
      "grad_norm": 1.8125,
      "learning_rate": 0.000494760378181364,
      "loss": 5.627,
      "mean_token_accuracy": 0.1703273355960846,
      "num_tokens": 15751868.0,
      "step": 7060
    },
    {
      "entropy": 5.659696960449219,
      "epoch": 0.7935081709440108,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004947517607358559,
      "loss": 5.6401,
      "mean_token_accuracy": 0.16680581122636795,
      "num_tokens": 15762783.0,
      "step": 7065
    },
    {
      "entropy": 5.722996759414673,
      "epoch": 0.7940697478519684,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004947431362933666,
      "loss": 5.6208,
      "mean_token_accuracy": 0.17095878720283508,
      "num_tokens": 15774155.0,
      "step": 7070
    },
    {
      "entropy": 5.82449369430542,
      "epoch": 0.7946313247599258,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004947345048541706,
      "loss": 5.6404,
      "mean_token_accuracy": 0.1689218834042549,
      "num_tokens": 15785034.0,
      "step": 7075
    },
    {
      "entropy": 5.595941543579102,
      "epoch": 0.7951929016678834,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004947258664185427,
      "loss": 5.5844,
      "mean_token_accuracy": 0.17003095299005508,
      "num_tokens": 15795959.0,
      "step": 7080
    },
    {
      "entropy": 5.658422422409058,
      "epoch": 0.7957544785758409,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004947172209867581,
      "loss": 5.4638,
      "mean_token_accuracy": 0.18406398892402648,
      "num_tokens": 15805561.0,
      "step": 7085
    },
    {
      "entropy": 5.713173961639404,
      "epoch": 0.7963160554837985,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004947085685590919,
      "loss": 5.5692,
      "mean_token_accuracy": 0.17338566929101945,
      "num_tokens": 15816883.0,
      "step": 7090
    },
    {
      "entropy": 5.580157470703125,
      "epoch": 0.796877632391756,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004946999091358196,
      "loss": 5.5261,
      "mean_token_accuracy": 0.17477951496839522,
      "num_tokens": 15827753.0,
      "step": 7095
    },
    {
      "entropy": 5.665834951400757,
      "epoch": 0.7974392092997136,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000494691242717217,
      "loss": 5.6449,
      "mean_token_accuracy": 0.1671520546078682,
      "num_tokens": 15840346.0,
      "step": 7100
    },
    {
      "entropy": 5.718213987350464,
      "epoch": 0.7980007862076711,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004946825693035601,
      "loss": 5.5982,
      "mean_token_accuracy": 0.17325277775526046,
      "num_tokens": 15851036.0,
      "step": 7105
    },
    {
      "entropy": 5.6614522457122805,
      "epoch": 0.7985623631156287,
      "grad_norm": 1.25,
      "learning_rate": 0.000494673888895125,
      "loss": 5.5575,
      "mean_token_accuracy": 0.16837076246738433,
      "num_tokens": 15863154.0,
      "step": 7110
    },
    {
      "entropy": 5.720604181289673,
      "epoch": 0.7991239400235862,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004946652014921881,
      "loss": 5.6639,
      "mean_token_accuracy": 0.16528711915016175,
      "num_tokens": 15874543.0,
      "step": 7115
    },
    {
      "entropy": 5.706193590164185,
      "epoch": 0.7996855169315438,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004946565070950259,
      "loss": 5.5968,
      "mean_token_accuracy": 0.17797693461179734,
      "num_tokens": 15886221.0,
      "step": 7120
    },
    {
      "entropy": 5.709640121459961,
      "epoch": 0.8002470938395013,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004946478057039153,
      "loss": 5.6053,
      "mean_token_accuracy": 0.17097271531820296,
      "num_tokens": 15897602.0,
      "step": 7125
    },
    {
      "entropy": 5.709883308410644,
      "epoch": 0.8008086707474589,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004946390973191335,
      "loss": 5.6058,
      "mean_token_accuracy": 0.171327905356884,
      "num_tokens": 15908030.0,
      "step": 7130
    },
    {
      "entropy": 5.781009912490845,
      "epoch": 0.8013702476554164,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004946303819409575,
      "loss": 5.6467,
      "mean_token_accuracy": 0.17163679897785186,
      "num_tokens": 15918341.0,
      "step": 7135
    },
    {
      "entropy": 5.623637104034424,
      "epoch": 0.801931824563374,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004946216595696652,
      "loss": 5.6264,
      "mean_token_accuracy": 0.17764584273099898,
      "num_tokens": 15929144.0,
      "step": 7140
    },
    {
      "entropy": 5.7017920970916744,
      "epoch": 0.8024934014713315,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004946129302055338,
      "loss": 5.5534,
      "mean_token_accuracy": 0.1748026505112648,
      "num_tokens": 15940013.0,
      "step": 7145
    },
    {
      "entropy": 5.746622562408447,
      "epoch": 0.8030549783792891,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004946041938488417,
      "loss": 5.6151,
      "mean_token_accuracy": 0.17079914808273317,
      "num_tokens": 15950680.0,
      "step": 7150
    },
    {
      "entropy": 5.644058322906494,
      "epoch": 0.8036165552872466,
      "grad_norm": 1.625,
      "learning_rate": 0.0004945954504998668,
      "loss": 5.6012,
      "mean_token_accuracy": 0.17851518392562865,
      "num_tokens": 15961462.0,
      "step": 7155
    },
    {
      "entropy": 5.67428412437439,
      "epoch": 0.8041781321952042,
      "grad_norm": 1.375,
      "learning_rate": 0.0004945867001588877,
      "loss": 5.5232,
      "mean_token_accuracy": 0.17970660775899888,
      "num_tokens": 15972788.0,
      "step": 7160
    },
    {
      "entropy": 5.655921173095703,
      "epoch": 0.8047397091031617,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004945779428261827,
      "loss": 5.5807,
      "mean_token_accuracy": 0.17457723915576934,
      "num_tokens": 15983244.0,
      "step": 7165
    },
    {
      "entropy": 5.596137046813965,
      "epoch": 0.8053012860111193,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000494569178502031,
      "loss": 5.6844,
      "mean_token_accuracy": 0.16675399839878083,
      "num_tokens": 15994325.0,
      "step": 7170
    },
    {
      "entropy": 5.77509298324585,
      "epoch": 0.8058628629190767,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004945604071867114,
      "loss": 5.6754,
      "mean_token_accuracy": 0.1712609902024269,
      "num_tokens": 16005237.0,
      "step": 7175
    },
    {
      "entropy": 5.679333925247192,
      "epoch": 0.8064244398270343,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004945516288805035,
      "loss": 5.5028,
      "mean_token_accuracy": 0.180667944252491,
      "num_tokens": 16015943.0,
      "step": 7180
    },
    {
      "entropy": 5.576353931427002,
      "epoch": 0.8069860167349918,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004945428435836864,
      "loss": 5.5114,
      "mean_token_accuracy": 0.17413717806339263,
      "num_tokens": 16026727.0,
      "step": 7185
    },
    {
      "entropy": 5.622631072998047,
      "epoch": 0.8075475936429494,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004945340512965401,
      "loss": 5.5532,
      "mean_token_accuracy": 0.16718171387910843,
      "num_tokens": 16039021.0,
      "step": 7190
    },
    {
      "entropy": 5.626061344146729,
      "epoch": 0.8081091705509069,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004945252520193443,
      "loss": 5.5031,
      "mean_token_accuracy": 0.17845340073108673,
      "num_tokens": 16050312.0,
      "step": 7195
    },
    {
      "entropy": 5.668435192108154,
      "epoch": 0.8086707474588645,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004945164457523795,
      "loss": 5.4561,
      "mean_token_accuracy": 0.18104152232408524,
      "num_tokens": 16060009.0,
      "step": 7200
    },
    {
      "entropy": 5.589456796646118,
      "epoch": 0.809232324366822,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004945076324959258,
      "loss": 5.5798,
      "mean_token_accuracy": 0.17277259826660157,
      "num_tokens": 16072025.0,
      "step": 7205
    },
    {
      "entropy": 5.730244207382202,
      "epoch": 0.8097939012747796,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004944988122502641,
      "loss": 5.6001,
      "mean_token_accuracy": 0.17296409606933594,
      "num_tokens": 16083229.0,
      "step": 7210
    },
    {
      "entropy": 5.6034921169281,
      "epoch": 0.8103554781827371,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000494489985015675,
      "loss": 5.4848,
      "mean_token_accuracy": 0.17338486462831498,
      "num_tokens": 16094936.0,
      "step": 7215
    },
    {
      "entropy": 5.629249382019043,
      "epoch": 0.8109170550906947,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004944811507924396,
      "loss": 5.5966,
      "mean_token_accuracy": 0.17871529906988143,
      "num_tokens": 16106182.0,
      "step": 7220
    },
    {
      "entropy": 5.612615060806275,
      "epoch": 0.8114786319986522,
      "grad_norm": 1.25,
      "learning_rate": 0.0004944723095808393,
      "loss": 5.5095,
      "mean_token_accuracy": 0.18377710580825807,
      "num_tokens": 16117006.0,
      "step": 7225
    },
    {
      "entropy": 5.6077577590942385,
      "epoch": 0.8120402089066098,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004944634613811555,
      "loss": 5.5422,
      "mean_token_accuracy": 0.1714983433485031,
      "num_tokens": 16128892.0,
      "step": 7230
    },
    {
      "entropy": 5.727390718460083,
      "epoch": 0.8126017858145673,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00049445460619367,
      "loss": 5.6503,
      "mean_token_accuracy": 0.17741401940584184,
      "num_tokens": 16139953.0,
      "step": 7235
    },
    {
      "entropy": 5.771072483062744,
      "epoch": 0.8131633627225249,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004944457440186648,
      "loss": 5.6187,
      "mean_token_accuracy": 0.1663561910390854,
      "num_tokens": 16150103.0,
      "step": 7240
    },
    {
      "entropy": 5.621788215637207,
      "epoch": 0.8137249396304824,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004944368748564219,
      "loss": 5.7002,
      "mean_token_accuracy": 0.167274634540081,
      "num_tokens": 16161666.0,
      "step": 7245
    },
    {
      "entropy": 5.6615965366363525,
      "epoch": 0.81428651653844,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004944279987072238,
      "loss": 5.6388,
      "mean_token_accuracy": 0.16890483051538469,
      "num_tokens": 16172524.0,
      "step": 7250
    },
    {
      "entropy": 5.7637022018432615,
      "epoch": 0.8148480934463975,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004944191155713531,
      "loss": 5.5756,
      "mean_token_accuracy": 0.1765427216887474,
      "num_tokens": 16183023.0,
      "step": 7255
    },
    {
      "entropy": 5.686658239364624,
      "epoch": 0.8154096703543551,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004944102254490928,
      "loss": 5.5918,
      "mean_token_accuracy": 0.17796089202165605,
      "num_tokens": 16193504.0,
      "step": 7260
    },
    {
      "entropy": 5.630303049087525,
      "epoch": 0.8159712472623126,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004944013283407257,
      "loss": 5.5496,
      "mean_token_accuracy": 0.17709226608276368,
      "num_tokens": 16204730.0,
      "step": 7265
    },
    {
      "entropy": 5.6650426387786865,
      "epoch": 0.8165328241702701,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004943924242465352,
      "loss": 5.5613,
      "mean_token_accuracy": 0.1778118908405304,
      "num_tokens": 16214496.0,
      "step": 7270
    },
    {
      "entropy": 5.608745145797729,
      "epoch": 0.8170944010782276,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004943835131668049,
      "loss": 5.5023,
      "mean_token_accuracy": 0.18091920465230943,
      "num_tokens": 16225494.0,
      "step": 7275
    },
    {
      "entropy": 5.569278240203857,
      "epoch": 0.8176559779861852,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004943745951018184,
      "loss": 5.4483,
      "mean_token_accuracy": 0.17797216773033142,
      "num_tokens": 16236385.0,
      "step": 7280
    },
    {
      "entropy": 5.640453624725342,
      "epoch": 0.8182175548941427,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004943656700518597,
      "loss": 5.6211,
      "mean_token_accuracy": 0.1828884169459343,
      "num_tokens": 16247300.0,
      "step": 7285
    },
    {
      "entropy": 5.715027952194214,
      "epoch": 0.8187791318021003,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004943567380172129,
      "loss": 5.6553,
      "mean_token_accuracy": 0.17173643559217452,
      "num_tokens": 16259326.0,
      "step": 7290
    },
    {
      "entropy": 5.711592769622802,
      "epoch": 0.8193407087100578,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004943477989981627,
      "loss": 5.6291,
      "mean_token_accuracy": 0.17496313005685807,
      "num_tokens": 16270559.0,
      "step": 7295
    },
    {
      "entropy": 5.681230926513672,
      "epoch": 0.8199022856180154,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004943388529949933,
      "loss": 5.6011,
      "mean_token_accuracy": 0.16669342219829558,
      "num_tokens": 16283120.0,
      "step": 7300
    },
    {
      "entropy": 5.798020315170288,
      "epoch": 0.8204638625259729,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004943299000079899,
      "loss": 5.6213,
      "mean_token_accuracy": 0.17745826691389083,
      "num_tokens": 16293449.0,
      "step": 7305
    },
    {
      "entropy": 5.665607500076294,
      "epoch": 0.8210254394339305,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004943209400374373,
      "loss": 5.6342,
      "mean_token_accuracy": 0.16801470518112183,
      "num_tokens": 16305144.0,
      "step": 7310
    },
    {
      "entropy": 5.592559432983398,
      "epoch": 0.821587016341888,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000494311973083621,
      "loss": 5.4582,
      "mean_token_accuracy": 0.17840937823057174,
      "num_tokens": 16316728.0,
      "step": 7315
    },
    {
      "entropy": 5.7071305274963375,
      "epoch": 0.8221485932498456,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004943029991468264,
      "loss": 5.7375,
      "mean_token_accuracy": 0.16938112676143646,
      "num_tokens": 16329117.0,
      "step": 7320
    },
    {
      "entropy": 5.769921779632568,
      "epoch": 0.8227101701578031,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004942940182273393,
      "loss": 5.6999,
      "mean_token_accuracy": 0.1706177443265915,
      "num_tokens": 16341651.0,
      "step": 7325
    },
    {
      "entropy": 5.79766092300415,
      "epoch": 0.8232717470657607,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004942850303254455,
      "loss": 5.6691,
      "mean_token_accuracy": 0.17348309159278869,
      "num_tokens": 16353220.0,
      "step": 7330
    },
    {
      "entropy": 5.774486017227173,
      "epoch": 0.8238333239737182,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004942760354414313,
      "loss": 5.706,
      "mean_token_accuracy": 0.1695014789700508,
      "num_tokens": 16363748.0,
      "step": 7335
    },
    {
      "entropy": 5.521135091781616,
      "epoch": 0.8243949008816758,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004942670335755831,
      "loss": 5.3854,
      "mean_token_accuracy": 0.18615321964025497,
      "num_tokens": 16374086.0,
      "step": 7340
    },
    {
      "entropy": 5.6541694641113285,
      "epoch": 0.8249564777896333,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004942580247281876,
      "loss": 5.5564,
      "mean_token_accuracy": 0.17574853003025054,
      "num_tokens": 16384140.0,
      "step": 7345
    },
    {
      "entropy": 5.707195806503296,
      "epoch": 0.8255180546975909,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004942490088995315,
      "loss": 5.6636,
      "mean_token_accuracy": 0.1692671984434128,
      "num_tokens": 16395940.0,
      "step": 7350
    },
    {
      "entropy": 5.629947328567505,
      "epoch": 0.8260796316055484,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004942399860899018,
      "loss": 5.5087,
      "mean_token_accuracy": 0.18015119284391404,
      "num_tokens": 16406667.0,
      "step": 7355
    },
    {
      "entropy": 5.7489300727844235,
      "epoch": 0.826641208513506,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004942309562995862,
      "loss": 5.6606,
      "mean_token_accuracy": 0.165461665391922,
      "num_tokens": 16418621.0,
      "step": 7360
    },
    {
      "entropy": 5.661215448379517,
      "epoch": 0.8272027854214634,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004942219195288716,
      "loss": 5.5629,
      "mean_token_accuracy": 0.1782558485865593,
      "num_tokens": 16429576.0,
      "step": 7365
    },
    {
      "entropy": 5.675850677490234,
      "epoch": 0.827764362329421,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004942128757780463,
      "loss": 5.6791,
      "mean_token_accuracy": 0.16648157685995102,
      "num_tokens": 16441004.0,
      "step": 7370
    },
    {
      "entropy": 5.695897388458252,
      "epoch": 0.8283259392373785,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004942038250473979,
      "loss": 5.6208,
      "mean_token_accuracy": 0.16576655954122543,
      "num_tokens": 16452183.0,
      "step": 7375
    },
    {
      "entropy": 5.650100517272949,
      "epoch": 0.8288875161453361,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004941947673372147,
      "loss": 5.5744,
      "mean_token_accuracy": 0.17556263953447343,
      "num_tokens": 16463552.0,
      "step": 7380
    },
    {
      "entropy": 5.723178291320801,
      "epoch": 0.8294490930532936,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004941857026477851,
      "loss": 5.6099,
      "mean_token_accuracy": 0.1799635797739029,
      "num_tokens": 16474062.0,
      "step": 7385
    },
    {
      "entropy": 5.752410173416138,
      "epoch": 0.8300106699612512,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004941766309793978,
      "loss": 5.6976,
      "mean_token_accuracy": 0.17354531735181808,
      "num_tokens": 16487377.0,
      "step": 7390
    },
    {
      "entropy": 5.755760908126831,
      "epoch": 0.8305722468692087,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004941675523323414,
      "loss": 5.6332,
      "mean_token_accuracy": 0.18097470849752426,
      "num_tokens": 16499704.0,
      "step": 7395
    },
    {
      "entropy": 5.6458765983581545,
      "epoch": 0.8311338237771663,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004941584667069052,
      "loss": 5.6438,
      "mean_token_accuracy": 0.17571662366390228,
      "num_tokens": 16511954.0,
      "step": 7400
    },
    {
      "entropy": 5.5650115489959715,
      "epoch": 0.8316954006851238,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004941493741033784,
      "loss": 5.479,
      "mean_token_accuracy": 0.1725131541490555,
      "num_tokens": 16522602.0,
      "step": 7405
    },
    {
      "entropy": 5.740706253051758,
      "epoch": 0.8322569775930814,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004941402745220507,
      "loss": 5.6325,
      "mean_token_accuracy": 0.16962281316518785,
      "num_tokens": 16534233.0,
      "step": 7410
    },
    {
      "entropy": 5.7556068897247314,
      "epoch": 0.8328185545010389,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004941311679632115,
      "loss": 5.7121,
      "mean_token_accuracy": 0.17037609368562698,
      "num_tokens": 16546120.0,
      "step": 7415
    },
    {
      "entropy": 5.709545755386353,
      "epoch": 0.8333801314089965,
      "grad_norm": 1.421875,
      "learning_rate": 0.000494122054427151,
      "loss": 5.5935,
      "mean_token_accuracy": 0.1746015027165413,
      "num_tokens": 16556830.0,
      "step": 7420
    },
    {
      "entropy": 5.644628953933716,
      "epoch": 0.833941708316954,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004941129339141592,
      "loss": 5.5573,
      "mean_token_accuracy": 0.17804006040096282,
      "num_tokens": 16567436.0,
      "step": 7425
    },
    {
      "entropy": 5.660109615325927,
      "epoch": 0.8345032852249116,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004941038064245267,
      "loss": 5.5257,
      "mean_token_accuracy": 0.1747888594865799,
      "num_tokens": 16577807.0,
      "step": 7430
    },
    {
      "entropy": 5.707765102386475,
      "epoch": 0.8350648621328691,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000494094671958544,
      "loss": 5.6211,
      "mean_token_accuracy": 0.17427547425031661,
      "num_tokens": 16589745.0,
      "step": 7435
    },
    {
      "entropy": 5.779169940948487,
      "epoch": 0.8356264390408267,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004940855305165019,
      "loss": 5.6572,
      "mean_token_accuracy": 0.17683361768722533,
      "num_tokens": 16600727.0,
      "step": 7440
    },
    {
      "entropy": 5.661066389083862,
      "epoch": 0.8361880159487842,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004940763820986916,
      "loss": 5.5597,
      "mean_token_accuracy": 0.17430997490882874,
      "num_tokens": 16611091.0,
      "step": 7445
    },
    {
      "entropy": 5.582389688491821,
      "epoch": 0.8367495928567418,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004940672267054044,
      "loss": 5.5401,
      "mean_token_accuracy": 0.17366688549518586,
      "num_tokens": 16622550.0,
      "step": 7450
    },
    {
      "entropy": 5.662583303451538,
      "epoch": 0.8373111697646993,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004940580643369316,
      "loss": 5.5482,
      "mean_token_accuracy": 0.17264899015426635,
      "num_tokens": 16634243.0,
      "step": 7455
    },
    {
      "entropy": 5.637042140960693,
      "epoch": 0.8378727466726568,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004940488949935651,
      "loss": 5.5622,
      "mean_token_accuracy": 0.1711127832531929,
      "num_tokens": 16645266.0,
      "step": 7460
    },
    {
      "entropy": 5.686684608459473,
      "epoch": 0.8384343235806143,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004940397186755968,
      "loss": 5.5659,
      "mean_token_accuracy": 0.17493919879198075,
      "num_tokens": 16654365.0,
      "step": 7465
    },
    {
      "entropy": 5.683757019042969,
      "epoch": 0.8389959004885719,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004940305353833191,
      "loss": 5.5409,
      "mean_token_accuracy": 0.18146927505731583,
      "num_tokens": 16665251.0,
      "step": 7470
    },
    {
      "entropy": 5.676362562179565,
      "epoch": 0.8395574773965294,
      "grad_norm": 1.390625,
      "learning_rate": 0.000494021345117024,
      "loss": 5.5938,
      "mean_token_accuracy": 0.17529145181179046,
      "num_tokens": 16676209.0,
      "step": 7475
    },
    {
      "entropy": 5.618336915969849,
      "epoch": 0.840119054304487,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004940121478770044,
      "loss": 5.5403,
      "mean_token_accuracy": 0.17493845522403717,
      "num_tokens": 16687213.0,
      "step": 7480
    },
    {
      "entropy": 5.630944728851318,
      "epoch": 0.8406806312124445,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004940029436635531,
      "loss": 5.5802,
      "mean_token_accuracy": 0.18078657239675522,
      "num_tokens": 16697920.0,
      "step": 7485
    },
    {
      "entropy": 5.627004718780517,
      "epoch": 0.8412422081204021,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004939937324769631,
      "loss": 5.5623,
      "mean_token_accuracy": 0.17084627151489257,
      "num_tokens": 16707815.0,
      "step": 7490
    },
    {
      "entropy": 5.632286834716797,
      "epoch": 0.8418037850283596,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004939845143175276,
      "loss": 5.4714,
      "mean_token_accuracy": 0.1783707097172737,
      "num_tokens": 16719221.0,
      "step": 7495
    },
    {
      "entropy": 5.597582626342773,
      "epoch": 0.8423653619363172,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004939752891855404,
      "loss": 5.5105,
      "mean_token_accuracy": 0.17155885994434356,
      "num_tokens": 16731741.0,
      "step": 7500
    },
    {
      "entropy": 5.613684749603271,
      "epoch": 0.8429269388442747,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004939660570812951,
      "loss": 5.5279,
      "mean_token_accuracy": 0.18077764958143233,
      "num_tokens": 16742795.0,
      "step": 7505
    },
    {
      "entropy": 5.636168098449707,
      "epoch": 0.8434885157522323,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004939568180050855,
      "loss": 5.524,
      "mean_token_accuracy": 0.177456533908844,
      "num_tokens": 16753139.0,
      "step": 7510
    },
    {
      "entropy": 5.653849363327026,
      "epoch": 0.8440500926601898,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004939475719572059,
      "loss": 5.5368,
      "mean_token_accuracy": 0.17530351132154465,
      "num_tokens": 16764246.0,
      "step": 7515
    },
    {
      "entropy": 5.70451078414917,
      "epoch": 0.8446116695681474,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004939383189379507,
      "loss": 5.6509,
      "mean_token_accuracy": 0.1700239211320877,
      "num_tokens": 16776039.0,
      "step": 7520
    },
    {
      "entropy": 5.718337535858154,
      "epoch": 0.8451732464761049,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004939290589476145,
      "loss": 5.6344,
      "mean_token_accuracy": 0.17975671738386154,
      "num_tokens": 16786979.0,
      "step": 7525
    },
    {
      "entropy": 5.705946540832519,
      "epoch": 0.8457348233840625,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004939197919864922,
      "loss": 5.6571,
      "mean_token_accuracy": 0.17143665701150895,
      "num_tokens": 16798113.0,
      "step": 7530
    },
    {
      "entropy": 5.6824140548706055,
      "epoch": 0.84629640029202,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004939105180548788,
      "loss": 5.5921,
      "mean_token_accuracy": 0.17321973741054536,
      "num_tokens": 16809072.0,
      "step": 7535
    },
    {
      "entropy": 5.61365909576416,
      "epoch": 0.8468579771999776,
      "grad_norm": 1.5,
      "learning_rate": 0.0004939012371530696,
      "loss": 5.3964,
      "mean_token_accuracy": 0.17975719422101974,
      "num_tokens": 16819548.0,
      "step": 7540
    },
    {
      "entropy": 5.733366680145264,
      "epoch": 0.8474195541079351,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004938919492813601,
      "loss": 5.7082,
      "mean_token_accuracy": 0.16560092866420745,
      "num_tokens": 16831935.0,
      "step": 7545
    },
    {
      "entropy": 5.675737428665161,
      "epoch": 0.8479811310158927,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004938826544400461,
      "loss": 5.604,
      "mean_token_accuracy": 0.17931524962186812,
      "num_tokens": 16842961.0,
      "step": 7550
    },
    {
      "entropy": 5.6335454940795895,
      "epoch": 0.8485427079238502,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004938733526294234,
      "loss": 5.5092,
      "mean_token_accuracy": 0.17333014458417892,
      "num_tokens": 16854392.0,
      "step": 7555
    },
    {
      "entropy": 5.620349311828614,
      "epoch": 0.8491042848318077,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004938640438497883,
      "loss": 5.5988,
      "mean_token_accuracy": 0.16811076998710633,
      "num_tokens": 16865443.0,
      "step": 7560
    },
    {
      "entropy": 5.677863216400146,
      "epoch": 0.8496658617397652,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004938547281014372,
      "loss": 5.5517,
      "mean_token_accuracy": 0.17474559247493743,
      "num_tokens": 16876854.0,
      "step": 7565
    },
    {
      "entropy": 5.7703766345977785,
      "epoch": 0.8502274386477228,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004938454053846666,
      "loss": 5.6388,
      "mean_token_accuracy": 0.1716684192419052,
      "num_tokens": 16888823.0,
      "step": 7570
    },
    {
      "entropy": 5.61264853477478,
      "epoch": 0.8507890155556803,
      "grad_norm": 1.875,
      "learning_rate": 0.0004938360756997735,
      "loss": 5.565,
      "mean_token_accuracy": 0.1738715797662735,
      "num_tokens": 16898807.0,
      "step": 7575
    },
    {
      "entropy": 5.690545082092285,
      "epoch": 0.8513505924636379,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004938267390470547,
      "loss": 5.6812,
      "mean_token_accuracy": 0.16348067820072174,
      "num_tokens": 16911258.0,
      "step": 7580
    },
    {
      "entropy": 5.783576726913452,
      "epoch": 0.8519121693715954,
      "grad_norm": 1.625,
      "learning_rate": 0.0004938173954268079,
      "loss": 5.6417,
      "mean_token_accuracy": 0.17234276682138444,
      "num_tokens": 16922557.0,
      "step": 7585
    },
    {
      "entropy": 5.723300170898438,
      "epoch": 0.852473746279553,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004938080448393302,
      "loss": 5.66,
      "mean_token_accuracy": 0.16851891279220582,
      "num_tokens": 16935330.0,
      "step": 7590
    },
    {
      "entropy": 5.723967742919922,
      "epoch": 0.8530353231875105,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004937986872849195,
      "loss": 5.5626,
      "mean_token_accuracy": 0.1776669666171074,
      "num_tokens": 16945361.0,
      "step": 7595
    },
    {
      "entropy": 5.565654373168945,
      "epoch": 0.8535969000954681,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004937893227638738,
      "loss": 5.4383,
      "mean_token_accuracy": 0.18152663558721543,
      "num_tokens": 16955363.0,
      "step": 7600
    },
    {
      "entropy": 5.56234769821167,
      "epoch": 0.8541584770034256,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004937799512764912,
      "loss": 5.5148,
      "mean_token_accuracy": 0.17852590829133988,
      "num_tokens": 16965267.0,
      "step": 7605
    },
    {
      "entropy": 5.671128654479981,
      "epoch": 0.8547200539113832,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00049377057282307,
      "loss": 5.7321,
      "mean_token_accuracy": 0.17176787555217743,
      "num_tokens": 16977528.0,
      "step": 7610
    },
    {
      "entropy": 5.715606927871704,
      "epoch": 0.8552816308193407,
      "grad_norm": 1.796875,
      "learning_rate": 0.000493761187403909,
      "loss": 5.5643,
      "mean_token_accuracy": 0.17219118773937225,
      "num_tokens": 16987326.0,
      "step": 7615
    },
    {
      "entropy": 5.799017238616943,
      "epoch": 0.8558432077272983,
      "grad_norm": 2.078125,
      "learning_rate": 0.000493751795019307,
      "loss": 5.6114,
      "mean_token_accuracy": 0.1744072198867798,
      "num_tokens": 16999127.0,
      "step": 7620
    },
    {
      "entropy": 5.582177782058716,
      "epoch": 0.8564047846352558,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004937423956695629,
      "loss": 5.5232,
      "mean_token_accuracy": 0.18371396213769914,
      "num_tokens": 17009929.0,
      "step": 7625
    },
    {
      "entropy": 5.713424921035767,
      "epoch": 0.8569663615432134,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004937329893549762,
      "loss": 5.6514,
      "mean_token_accuracy": 0.17402349263429642,
      "num_tokens": 17020909.0,
      "step": 7630
    },
    {
      "entropy": 5.744926929473877,
      "epoch": 0.8575279384511709,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004937235760758462,
      "loss": 5.5807,
      "mean_token_accuracy": 0.17321668565273285,
      "num_tokens": 17031358.0,
      "step": 7635
    },
    {
      "entropy": 5.654985761642456,
      "epoch": 0.8580895153591285,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004937141558324726,
      "loss": 5.6242,
      "mean_token_accuracy": 0.17349520176649094,
      "num_tokens": 17042566.0,
      "step": 7640
    },
    {
      "entropy": 5.677562236785889,
      "epoch": 0.858651092267086,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004937047286251556,
      "loss": 5.6472,
      "mean_token_accuracy": 0.16312987059354783,
      "num_tokens": 17054998.0,
      "step": 7645
    },
    {
      "entropy": 5.6565553665161135,
      "epoch": 0.8592126691750436,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004936952944541952,
      "loss": 5.512,
      "mean_token_accuracy": 0.17477733939886092,
      "num_tokens": 17064876.0,
      "step": 7650
    },
    {
      "entropy": 5.647423267364502,
      "epoch": 0.859774246083001,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004936858533198917,
      "loss": 5.5621,
      "mean_token_accuracy": 0.17949966639280318,
      "num_tokens": 17075310.0,
      "step": 7655
    },
    {
      "entropy": 5.519606304168701,
      "epoch": 0.8603358229909586,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000493676405222546,
      "loss": 5.4676,
      "mean_token_accuracy": 0.17772774249315262,
      "num_tokens": 17086023.0,
      "step": 7660
    },
    {
      "entropy": 5.5962433338165285,
      "epoch": 0.8608973998989161,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004936669501624586,
      "loss": 5.5539,
      "mean_token_accuracy": 0.17585509717464448,
      "num_tokens": 17097554.0,
      "step": 7665
    },
    {
      "entropy": 5.8958173274993895,
      "epoch": 0.8614589768068737,
      "grad_norm": 6.15625,
      "learning_rate": 0.0004936574881399308,
      "loss": 5.6789,
      "mean_token_accuracy": 0.17030972838401795,
      "num_tokens": 17108890.0,
      "step": 7670
    },
    {
      "entropy": 5.694438219070435,
      "epoch": 0.8620205537148312,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004936480191552637,
      "loss": 5.5134,
      "mean_token_accuracy": 0.17637402266263963,
      "num_tokens": 17120651.0,
      "step": 7675
    },
    {
      "entropy": 5.666385555267334,
      "epoch": 0.8625821306227888,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004936385432087589,
      "loss": 5.651,
      "mean_token_accuracy": 0.17565001398324967,
      "num_tokens": 17131306.0,
      "step": 7680
    },
    {
      "entropy": 5.812044906616211,
      "epoch": 0.8631437075307463,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004936290603007181,
      "loss": 5.798,
      "mean_token_accuracy": 0.16602026224136351,
      "num_tokens": 17142779.0,
      "step": 7685
    },
    {
      "entropy": 5.860309505462647,
      "epoch": 0.8637052844387039,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004936195704314432,
      "loss": 5.6687,
      "mean_token_accuracy": 0.18110954910516738,
      "num_tokens": 17154992.0,
      "step": 7690
    },
    {
      "entropy": 5.662728500366211,
      "epoch": 0.8642668613466614,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004936100736012363,
      "loss": 5.6427,
      "mean_token_accuracy": 0.17859661132097243,
      "num_tokens": 17166089.0,
      "step": 7695
    },
    {
      "entropy": 5.610977554321289,
      "epoch": 0.864828438254619,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004936005698104,
      "loss": 5.5468,
      "mean_token_accuracy": 0.1707144245505333,
      "num_tokens": 17175691.0,
      "step": 7700
    },
    {
      "entropy": 5.6774708271026615,
      "epoch": 0.8653900151625765,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004935910590592367,
      "loss": 5.602,
      "mean_token_accuracy": 0.16899661272764205,
      "num_tokens": 17186466.0,
      "step": 7705
    },
    {
      "entropy": 5.582030820846557,
      "epoch": 0.8659515920705341,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004935815413480494,
      "loss": 5.3873,
      "mean_token_accuracy": 0.18827227801084517,
      "num_tokens": 17198520.0,
      "step": 7710
    },
    {
      "entropy": 5.670034170150757,
      "epoch": 0.8665131689784916,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000493572016677141,
      "loss": 5.6234,
      "mean_token_accuracy": 0.16890666335821153,
      "num_tokens": 17210496.0,
      "step": 7715
    },
    {
      "entropy": 5.68838415145874,
      "epoch": 0.8670747458864492,
      "grad_norm": 3.328125,
      "learning_rate": 0.0004935624850468148,
      "loss": 5.6152,
      "mean_token_accuracy": 0.17695807665586472,
      "num_tokens": 17221631.0,
      "step": 7720
    },
    {
      "entropy": 5.682717370986938,
      "epoch": 0.8676363227944067,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004935529464573742,
      "loss": 5.65,
      "mean_token_accuracy": 0.16548071503639222,
      "num_tokens": 17233316.0,
      "step": 7725
    },
    {
      "entropy": 5.678200769424438,
      "epoch": 0.8681978997023643,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004935434009091232,
      "loss": 5.5573,
      "mean_token_accuracy": 0.1785666510462761,
      "num_tokens": 17245761.0,
      "step": 7730
    },
    {
      "entropy": 5.697708415985107,
      "epoch": 0.8687594766103218,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004935338484023655,
      "loss": 5.5596,
      "mean_token_accuracy": 0.1804763749241829,
      "num_tokens": 17255007.0,
      "step": 7735
    },
    {
      "entropy": 5.644683218002319,
      "epoch": 0.8693210535182794,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004935242889374052,
      "loss": 5.536,
      "mean_token_accuracy": 0.17835524082183837,
      "num_tokens": 17265378.0,
      "step": 7740
    },
    {
      "entropy": 5.724565315246582,
      "epoch": 0.8698826304262369,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004935147225145469,
      "loss": 5.6105,
      "mean_token_accuracy": 0.17023085206747054,
      "num_tokens": 17277202.0,
      "step": 7745
    },
    {
      "entropy": 5.6139226913452145,
      "epoch": 0.8704442073341944,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004935051491340951,
      "loss": 5.5543,
      "mean_token_accuracy": 0.1805954396724701,
      "num_tokens": 17287006.0,
      "step": 7750
    },
    {
      "entropy": 5.649082994461059,
      "epoch": 0.8710057842421519,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004934955687963546,
      "loss": 5.52,
      "mean_token_accuracy": 0.18563045263290406,
      "num_tokens": 17297387.0,
      "step": 7755
    },
    {
      "entropy": 5.797161531448364,
      "epoch": 0.8715673611501095,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004934859815016304,
      "loss": 5.7537,
      "mean_token_accuracy": 0.16585738137364386,
      "num_tokens": 17310364.0,
      "step": 7760
    },
    {
      "entropy": 5.6904988288879395,
      "epoch": 0.872128938058067,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004934763872502278,
      "loss": 5.5897,
      "mean_token_accuracy": 0.17995432764291763,
      "num_tokens": 17321282.0,
      "step": 7765
    },
    {
      "entropy": 5.684155511856079,
      "epoch": 0.8726905149660246,
      "grad_norm": 1.625,
      "learning_rate": 0.0004934667860424523,
      "loss": 5.632,
      "mean_token_accuracy": 0.1700143799185753,
      "num_tokens": 17331755.0,
      "step": 7770
    },
    {
      "entropy": 5.656625986099243,
      "epoch": 0.8732520918739821,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004934571778786096,
      "loss": 5.5745,
      "mean_token_accuracy": 0.17331324964761735,
      "num_tokens": 17342542.0,
      "step": 7775
    },
    {
      "entropy": 5.652045154571534,
      "epoch": 0.8738136687819397,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004934475627590057,
      "loss": 5.5215,
      "mean_token_accuracy": 0.17376778274774551,
      "num_tokens": 17353150.0,
      "step": 7780
    },
    {
      "entropy": 5.65459680557251,
      "epoch": 0.8743752456898972,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004934379406839466,
      "loss": 5.5644,
      "mean_token_accuracy": 0.17380731254816056,
      "num_tokens": 17363901.0,
      "step": 7785
    },
    {
      "entropy": 5.67426061630249,
      "epoch": 0.8749368225978548,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004934283116537387,
      "loss": 5.488,
      "mean_token_accuracy": 0.17831778526306152,
      "num_tokens": 17374429.0,
      "step": 7790
    },
    {
      "entropy": 5.673512268066406,
      "epoch": 0.8754983995058123,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004934186756686888,
      "loss": 5.6291,
      "mean_token_accuracy": 0.17127451598644255,
      "num_tokens": 17384952.0,
      "step": 7795
    },
    {
      "entropy": 5.670864820480347,
      "epoch": 0.8760599764137699,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004934090327291034,
      "loss": 5.5948,
      "mean_token_accuracy": 0.17161913961172104,
      "num_tokens": 17397771.0,
      "step": 7800
    },
    {
      "entropy": 5.670876312255859,
      "epoch": 0.8766215533217274,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004933993828352898,
      "loss": 5.5027,
      "mean_token_accuracy": 0.17967257499694825,
      "num_tokens": 17407748.0,
      "step": 7805
    },
    {
      "entropy": 5.611726522445679,
      "epoch": 0.877183130229685,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004933897259875551,
      "loss": 5.4843,
      "mean_token_accuracy": 0.175080007314682,
      "num_tokens": 17418572.0,
      "step": 7810
    },
    {
      "entropy": 5.52510027885437,
      "epoch": 0.8777447071376425,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004933800621862069,
      "loss": 5.5306,
      "mean_token_accuracy": 0.17571943253278732,
      "num_tokens": 17429909.0,
      "step": 7815
    },
    {
      "entropy": 5.638761234283447,
      "epoch": 0.8783062840456001,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004933703914315527,
      "loss": 5.5553,
      "mean_token_accuracy": 0.1714665785431862,
      "num_tokens": 17441143.0,
      "step": 7820
    },
    {
      "entropy": 5.736855459213257,
      "epoch": 0.8788678609535576,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004933607137239006,
      "loss": 5.6724,
      "mean_token_accuracy": 0.16394690573215484,
      "num_tokens": 17452156.0,
      "step": 7825
    },
    {
      "entropy": 5.618521738052368,
      "epoch": 0.8794294378615152,
      "grad_norm": 1.5,
      "learning_rate": 0.0004933510290635587,
      "loss": 5.574,
      "mean_token_accuracy": 0.17211691588163375,
      "num_tokens": 17463873.0,
      "step": 7830
    },
    {
      "entropy": 5.598127937316894,
      "epoch": 0.8799910147694727,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004933413374508353,
      "loss": 5.5071,
      "mean_token_accuracy": 0.17668081372976302,
      "num_tokens": 17474180.0,
      "step": 7835
    },
    {
      "entropy": 5.652832412719727,
      "epoch": 0.8805525916774303,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004933316388860391,
      "loss": 5.539,
      "mean_token_accuracy": 0.1772014617919922,
      "num_tokens": 17486903.0,
      "step": 7840
    },
    {
      "entropy": 5.573214387893676,
      "epoch": 0.8811141685853877,
      "grad_norm": 1.75,
      "learning_rate": 0.0004933219333694787,
      "loss": 5.567,
      "mean_token_accuracy": 0.17032884508371354,
      "num_tokens": 17498790.0,
      "step": 7845
    },
    {
      "entropy": 5.723428106307983,
      "epoch": 0.8816757454933453,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004933122209014634,
      "loss": 5.5983,
      "mean_token_accuracy": 0.16774654984474183,
      "num_tokens": 17510061.0,
      "step": 7850
    },
    {
      "entropy": 5.679811573028564,
      "epoch": 0.8822373224013028,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004933025014823022,
      "loss": 5.6114,
      "mean_token_accuracy": 0.17310027331113814,
      "num_tokens": 17522414.0,
      "step": 7855
    },
    {
      "entropy": 5.8511148452758786,
      "epoch": 0.8827988993092604,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004932927751123048,
      "loss": 5.7443,
      "mean_token_accuracy": 0.16237847954034806,
      "num_tokens": 17532745.0,
      "step": 7860
    },
    {
      "entropy": 5.631840419769287,
      "epoch": 0.8833604762172179,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004932830417917808,
      "loss": 5.5358,
      "mean_token_accuracy": 0.17639891058206558,
      "num_tokens": 17545329.0,
      "step": 7865
    },
    {
      "entropy": 5.729234743118286,
      "epoch": 0.8839220531251755,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00049327330152104,
      "loss": 5.6476,
      "mean_token_accuracy": 0.1724901929497719,
      "num_tokens": 17557789.0,
      "step": 7870
    },
    {
      "entropy": 5.672587299346924,
      "epoch": 0.884483630033133,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004932635543003927,
      "loss": 5.4769,
      "mean_token_accuracy": 0.18150333017110826,
      "num_tokens": 17568338.0,
      "step": 7875
    },
    {
      "entropy": 5.638689804077148,
      "epoch": 0.8850452069410906,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004932538001301491,
      "loss": 5.5781,
      "mean_token_accuracy": 0.1726062461733818,
      "num_tokens": 17581199.0,
      "step": 7880
    },
    {
      "entropy": 5.544777345657349,
      "epoch": 0.8856067838490481,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004932440390106199,
      "loss": 5.5334,
      "mean_token_accuracy": 0.1764208197593689,
      "num_tokens": 17592937.0,
      "step": 7885
    },
    {
      "entropy": 5.629625701904297,
      "epoch": 0.8861683607570057,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004932342709421159,
      "loss": 5.5926,
      "mean_token_accuracy": 0.17708465456962585,
      "num_tokens": 17604003.0,
      "step": 7890
    },
    {
      "entropy": 5.90011510848999,
      "epoch": 0.8867299376649632,
      "grad_norm": 1.3125,
      "learning_rate": 0.000493224495924948,
      "loss": 5.7302,
      "mean_token_accuracy": 0.16930215954780578,
      "num_tokens": 17617283.0,
      "step": 7895
    },
    {
      "entropy": 5.677170133590698,
      "epoch": 0.8872915145729208,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004932147139594276,
      "loss": 5.5707,
      "mean_token_accuracy": 0.1770807087421417,
      "num_tokens": 17628161.0,
      "step": 7900
    },
    {
      "entropy": 5.728165149688721,
      "epoch": 0.8878530914808783,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004932049250458659,
      "loss": 5.6503,
      "mean_token_accuracy": 0.16944217383861543,
      "num_tokens": 17639226.0,
      "step": 7905
    },
    {
      "entropy": 5.592314767837524,
      "epoch": 0.8884146683888359,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004931951291845749,
      "loss": 5.4893,
      "mean_token_accuracy": 0.18333953768014907,
      "num_tokens": 17652114.0,
      "step": 7910
    },
    {
      "entropy": 5.5825001239776615,
      "epoch": 0.8889762452967934,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004931853263758663,
      "loss": 5.529,
      "mean_token_accuracy": 0.18106401562690735,
      "num_tokens": 17663280.0,
      "step": 7915
    },
    {
      "entropy": 5.623987007141113,
      "epoch": 0.889537822204751,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004931755166200523,
      "loss": 5.6155,
      "mean_token_accuracy": 0.17368122637271882,
      "num_tokens": 17674445.0,
      "step": 7920
    },
    {
      "entropy": 5.691004753112793,
      "epoch": 0.8900993991127085,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004931656999174452,
      "loss": 5.5071,
      "mean_token_accuracy": 0.17749142944812774,
      "num_tokens": 17685046.0,
      "step": 7925
    },
    {
      "entropy": 5.591828203201294,
      "epoch": 0.8906609760206661,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004931558762683577,
      "loss": 5.507,
      "mean_token_accuracy": 0.17967617213726045,
      "num_tokens": 17695269.0,
      "step": 7930
    },
    {
      "entropy": 5.559813833236694,
      "epoch": 0.8912225529286236,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004931460456731024,
      "loss": 5.5466,
      "mean_token_accuracy": 0.17714540660381317,
      "num_tokens": 17706296.0,
      "step": 7935
    },
    {
      "entropy": 5.6044251918792725,
      "epoch": 0.8917841298365812,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004931362081319925,
      "loss": 5.4769,
      "mean_token_accuracy": 0.17915261387825013,
      "num_tokens": 17716262.0,
      "step": 7940
    },
    {
      "entropy": 5.66194977760315,
      "epoch": 0.8923457067445386,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004931263636453409,
      "loss": 5.5201,
      "mean_token_accuracy": 0.17992712408304215,
      "num_tokens": 17726818.0,
      "step": 7945
    },
    {
      "entropy": 5.589525651931763,
      "epoch": 0.8929072836524962,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004931165122134616,
      "loss": 5.5256,
      "mean_token_accuracy": 0.18117809742689134,
      "num_tokens": 17737112.0,
      "step": 7950
    },
    {
      "entropy": 5.660979413986206,
      "epoch": 0.8934688605604537,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004931066538366677,
      "loss": 5.5984,
      "mean_token_accuracy": 0.17052928656339644,
      "num_tokens": 17747558.0,
      "step": 7955
    },
    {
      "entropy": 5.712664318084717,
      "epoch": 0.8940304374684113,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004930967885152734,
      "loss": 5.536,
      "mean_token_accuracy": 0.17624609172344208,
      "num_tokens": 17757495.0,
      "step": 7960
    },
    {
      "entropy": 5.624363279342651,
      "epoch": 0.8945920143763688,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004930869162495929,
      "loss": 5.5294,
      "mean_token_accuracy": 0.1765795186161995,
      "num_tokens": 17767562.0,
      "step": 7965
    },
    {
      "entropy": 5.6667194843292235,
      "epoch": 0.8951535912843264,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004930770370399403,
      "loss": 5.6186,
      "mean_token_accuracy": 0.16974660009145737,
      "num_tokens": 17779800.0,
      "step": 7970
    },
    {
      "entropy": 5.610093069076538,
      "epoch": 0.8957151681922839,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004930671508866303,
      "loss": 5.4308,
      "mean_token_accuracy": 0.17751775681972504,
      "num_tokens": 17791183.0,
      "step": 7975
    },
    {
      "entropy": 5.681202840805054,
      "epoch": 0.8962767451002415,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004930572577899776,
      "loss": 5.5975,
      "mean_token_accuracy": 0.1753363460302353,
      "num_tokens": 17802547.0,
      "step": 7980
    },
    {
      "entropy": 5.511297702789307,
      "epoch": 0.896838322008199,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004930473577502972,
      "loss": 5.492,
      "mean_token_accuracy": 0.17547860294580458,
      "num_tokens": 17815406.0,
      "step": 7985
    },
    {
      "entropy": 5.631451606750488,
      "epoch": 0.8973998989161566,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004930374507679044,
      "loss": 5.521,
      "mean_token_accuracy": 0.17684534192085266,
      "num_tokens": 17826516.0,
      "step": 7990
    },
    {
      "entropy": 5.646472406387329,
      "epoch": 0.8979614758241141,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004930275368431146,
      "loss": 5.4978,
      "mean_token_accuracy": 0.1771871417760849,
      "num_tokens": 17838193.0,
      "step": 7995
    },
    {
      "entropy": 5.624480295181274,
      "epoch": 0.8985230527320717,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004930176159762435,
      "loss": 5.5205,
      "mean_token_accuracy": 0.1750570610165596,
      "num_tokens": 17849564.0,
      "step": 8000
    },
    {
      "entropy": 5.736857175827026,
      "epoch": 0.8990846296400292,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004930076881676069,
      "loss": 5.662,
      "mean_token_accuracy": 0.17136270999908448,
      "num_tokens": 17861874.0,
      "step": 8005
    },
    {
      "entropy": 5.575713205337524,
      "epoch": 0.8996462065479868,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004929977534175211,
      "loss": 5.5055,
      "mean_token_accuracy": 0.17983078062534333,
      "num_tokens": 17873139.0,
      "step": 8010
    },
    {
      "entropy": 5.736148262023926,
      "epoch": 0.9002077834559443,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004929878117263021,
      "loss": 5.572,
      "mean_token_accuracy": 0.17753154337406157,
      "num_tokens": 17883262.0,
      "step": 8015
    },
    {
      "entropy": 5.554032325744629,
      "epoch": 0.9007693603639019,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004929778630942668,
      "loss": 5.4604,
      "mean_token_accuracy": 0.18278993517160416,
      "num_tokens": 17895041.0,
      "step": 8020
    },
    {
      "entropy": 5.528722429275513,
      "epoch": 0.9013309372718594,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004929679075217317,
      "loss": 5.5034,
      "mean_token_accuracy": 0.17425708770751952,
      "num_tokens": 17905154.0,
      "step": 8025
    },
    {
      "entropy": 5.616612911224365,
      "epoch": 0.901892514179817,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004929579450090139,
      "loss": 5.5601,
      "mean_token_accuracy": 0.17341381311416626,
      "num_tokens": 17916287.0,
      "step": 8030
    },
    {
      "entropy": 5.7543384552001955,
      "epoch": 0.9024540910877745,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004929479755564305,
      "loss": 5.5673,
      "mean_token_accuracy": 0.17662589699029924,
      "num_tokens": 17927269.0,
      "step": 8035
    },
    {
      "entropy": 5.6863922595977785,
      "epoch": 0.903015667995732,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004929379991642991,
      "loss": 5.5818,
      "mean_token_accuracy": 0.1742201864719391,
      "num_tokens": 17938142.0,
      "step": 8040
    },
    {
      "entropy": 5.675807046890259,
      "epoch": 0.9035772449036895,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004929280158329373,
      "loss": 5.5645,
      "mean_token_accuracy": 0.1757965698838234,
      "num_tokens": 17950229.0,
      "step": 8045
    },
    {
      "entropy": 5.612723159790039,
      "epoch": 0.9041388218116471,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000492918025562663,
      "loss": 5.548,
      "mean_token_accuracy": 0.18230236321687698,
      "num_tokens": 17961757.0,
      "step": 8050
    },
    {
      "entropy": 5.685063695907592,
      "epoch": 0.9047003987196046,
      "grad_norm": 1.5,
      "learning_rate": 0.0004929080283537942,
      "loss": 5.6233,
      "mean_token_accuracy": 0.1663338214159012,
      "num_tokens": 17972564.0,
      "step": 8055
    },
    {
      "entropy": 5.735013771057129,
      "epoch": 0.9052619756275622,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004928980242066492,
      "loss": 5.5325,
      "mean_token_accuracy": 0.18212261348962783,
      "num_tokens": 17982776.0,
      "step": 8060
    },
    {
      "entropy": 5.688648509979248,
      "epoch": 0.9058235525355197,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004928880131215467,
      "loss": 5.6463,
      "mean_token_accuracy": 0.1749667137861252,
      "num_tokens": 17994460.0,
      "step": 8065
    },
    {
      "entropy": 5.624743270874023,
      "epoch": 0.9063851294434773,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004928779950988053,
      "loss": 5.62,
      "mean_token_accuracy": 0.17174448817968369,
      "num_tokens": 18005744.0,
      "step": 8070
    },
    {
      "entropy": 5.611772775650024,
      "epoch": 0.9069467063514348,
      "grad_norm": 1.4375,
      "learning_rate": 0.000492867970138744,
      "loss": 5.4754,
      "mean_token_accuracy": 0.18351303040981293,
      "num_tokens": 18016517.0,
      "step": 8075
    },
    {
      "entropy": 5.675594282150269,
      "epoch": 0.9075082832593924,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004928579382416821,
      "loss": 5.5975,
      "mean_token_accuracy": 0.16978041380643843,
      "num_tokens": 18027653.0,
      "step": 8080
    },
    {
      "entropy": 5.675237369537354,
      "epoch": 0.9080698601673499,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000492847899407939,
      "loss": 5.636,
      "mean_token_accuracy": 0.17045610696077346,
      "num_tokens": 18038741.0,
      "step": 8085
    },
    {
      "entropy": 5.7510435581207275,
      "epoch": 0.9086314370753075,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004928378536378341,
      "loss": 5.5802,
      "mean_token_accuracy": 0.17184377759695052,
      "num_tokens": 18049188.0,
      "step": 8090
    },
    {
      "entropy": 5.608703708648681,
      "epoch": 0.909193013983265,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004928278009316877,
      "loss": 5.482,
      "mean_token_accuracy": 0.17725870013237,
      "num_tokens": 18059979.0,
      "step": 8095
    },
    {
      "entropy": 5.6458415508270265,
      "epoch": 0.9097545908912226,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004928177412898195,
      "loss": 5.4967,
      "mean_token_accuracy": 0.1831892564892769,
      "num_tokens": 18070712.0,
      "step": 8100
    },
    {
      "entropy": 5.670635318756103,
      "epoch": 0.9103161677991801,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004928076747125499,
      "loss": 5.5632,
      "mean_token_accuracy": 0.18162725418806075,
      "num_tokens": 18081372.0,
      "step": 8105
    },
    {
      "entropy": 5.577960062026977,
      "epoch": 0.9108777447071377,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004927976012001995,
      "loss": 5.4888,
      "mean_token_accuracy": 0.17665131241083146,
      "num_tokens": 18093033.0,
      "step": 8110
    },
    {
      "entropy": 5.555771923065185,
      "epoch": 0.9114393216150952,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004927875207530889,
      "loss": 5.4461,
      "mean_token_accuracy": 0.18108128160238265,
      "num_tokens": 18104808.0,
      "step": 8115
    },
    {
      "entropy": 5.730599784851075,
      "epoch": 0.9120008985230528,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004927774333715394,
      "loss": 5.5846,
      "mean_token_accuracy": 0.17475593835115433,
      "num_tokens": 18115450.0,
      "step": 8120
    },
    {
      "entropy": 5.730729341506958,
      "epoch": 0.9125624754310103,
      "grad_norm": 2.90625,
      "learning_rate": 0.0004927673390558718,
      "loss": 5.6786,
      "mean_token_accuracy": 0.17357307821512222,
      "num_tokens": 18127437.0,
      "step": 8125
    },
    {
      "entropy": 5.705981349945068,
      "epoch": 0.9131240523389679,
      "grad_norm": 1.5,
      "learning_rate": 0.0004927572378064077,
      "loss": 5.5123,
      "mean_token_accuracy": 0.1819301575422287,
      "num_tokens": 18137371.0,
      "step": 8130
    },
    {
      "entropy": 5.618573713302612,
      "epoch": 0.9136856292469253,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004927471296234687,
      "loss": 5.5725,
      "mean_token_accuracy": 0.17275052070617675,
      "num_tokens": 18148332.0,
      "step": 8135
    },
    {
      "entropy": 5.588826656341553,
      "epoch": 0.9142472061548829,
      "grad_norm": 1.5,
      "learning_rate": 0.0004927370145073767,
      "loss": 5.5549,
      "mean_token_accuracy": 0.17383279651403427,
      "num_tokens": 18159918.0,
      "step": 8140
    },
    {
      "entropy": 5.65464768409729,
      "epoch": 0.9148087830628404,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004927268924584536,
      "loss": 5.5406,
      "mean_token_accuracy": 0.18160120993852616,
      "num_tokens": 18170911.0,
      "step": 8145
    },
    {
      "entropy": 5.667963266372681,
      "epoch": 0.915370359970798,
      "grad_norm": 1.375,
      "learning_rate": 0.0004927167634770219,
      "loss": 5.5562,
      "mean_token_accuracy": 0.17828303426504136,
      "num_tokens": 18183089.0,
      "step": 8150
    },
    {
      "entropy": 5.7810358047485355,
      "epoch": 0.9159319368787555,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000492706627563404,
      "loss": 5.8173,
      "mean_token_accuracy": 0.15935783535242082,
      "num_tokens": 18195467.0,
      "step": 8155
    },
    {
      "entropy": 5.740857410430908,
      "epoch": 0.9164935137867131,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004926964847179225,
      "loss": 5.572,
      "mean_token_accuracy": 0.1790512040257454,
      "num_tokens": 18206766.0,
      "step": 8160
    },
    {
      "entropy": 5.679031610488892,
      "epoch": 0.9170550906946706,
      "grad_norm": 1.375,
      "learning_rate": 0.0004926863349409006,
      "loss": 5.5629,
      "mean_token_accuracy": 0.17597046494483948,
      "num_tokens": 18217840.0,
      "step": 8165
    },
    {
      "entropy": 5.649716377258301,
      "epoch": 0.9176166676026282,
      "grad_norm": 2.0,
      "learning_rate": 0.0004926761782326613,
      "loss": 5.5758,
      "mean_token_accuracy": 0.17311379611492156,
      "num_tokens": 18229514.0,
      "step": 8170
    },
    {
      "entropy": 5.5854212760925295,
      "epoch": 0.9181782445105857,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004926660145935282,
      "loss": 5.5559,
      "mean_token_accuracy": 0.1705981135368347,
      "num_tokens": 18241949.0,
      "step": 8175
    },
    {
      "entropy": 5.651837682723999,
      "epoch": 0.9187398214185433,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004926558440238247,
      "loss": 5.5314,
      "mean_token_accuracy": 0.17833069413900376,
      "num_tokens": 18252938.0,
      "step": 8180
    },
    {
      "entropy": 5.624422931671143,
      "epoch": 0.9193013983265008,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004926456665238746,
      "loss": 5.479,
      "mean_token_accuracy": 0.18345132768154143,
      "num_tokens": 18263417.0,
      "step": 8185
    },
    {
      "entropy": 5.661493015289307,
      "epoch": 0.9198629752344584,
      "grad_norm": 1.375,
      "learning_rate": 0.0004926354820940023,
      "loss": 5.6485,
      "mean_token_accuracy": 0.17259145379066468,
      "num_tokens": 18275595.0,
      "step": 8190
    },
    {
      "entropy": 5.592373466491699,
      "epoch": 0.9204245521424159,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004926252907345317,
      "loss": 5.4836,
      "mean_token_accuracy": 0.17857701629400252,
      "num_tokens": 18287510.0,
      "step": 8195
    },
    {
      "entropy": 5.717923927307129,
      "epoch": 0.9209861290503735,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004926150924457874,
      "loss": 5.5838,
      "mean_token_accuracy": 0.17386380732059478,
      "num_tokens": 18298764.0,
      "step": 8200
    },
    {
      "entropy": 5.588229608535767,
      "epoch": 0.921547705958331,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004926048872280943,
      "loss": 5.4009,
      "mean_token_accuracy": 0.17802754789590836,
      "num_tokens": 18310427.0,
      "step": 8205
    },
    {
      "entropy": 5.580872583389282,
      "epoch": 0.9221092828662886,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000492594675081777,
      "loss": 5.532,
      "mean_token_accuracy": 0.17346930950880052,
      "num_tokens": 18322174.0,
      "step": 8210
    },
    {
      "entropy": 5.584685850143432,
      "epoch": 0.9226708597742461,
      "grad_norm": 2.8125,
      "learning_rate": 0.000492584456007161,
      "loss": 5.4795,
      "mean_token_accuracy": 0.17690377533435822,
      "num_tokens": 18332372.0,
      "step": 8215
    },
    {
      "entropy": 5.672969627380371,
      "epoch": 0.9232324366822037,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004925742300045716,
      "loss": 5.5459,
      "mean_token_accuracy": 0.1704122766852379,
      "num_tokens": 18343201.0,
      "step": 8220
    },
    {
      "entropy": 5.634353446960449,
      "epoch": 0.9237940135901612,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004925639970743341,
      "loss": 5.5613,
      "mean_token_accuracy": 0.1783763960003853,
      "num_tokens": 18354124.0,
      "step": 8225
    },
    {
      "entropy": 5.626930046081543,
      "epoch": 0.9243555904981187,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004925537572167746,
      "loss": 5.5138,
      "mean_token_accuracy": 0.18148281127214433,
      "num_tokens": 18365148.0,
      "step": 8230
    },
    {
      "entropy": 5.467441701889038,
      "epoch": 0.9249171674060762,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004925435104322192,
      "loss": 5.3908,
      "mean_token_accuracy": 0.18088495582342148,
      "num_tokens": 18376244.0,
      "step": 8235
    },
    {
      "entropy": 5.6002789497375485,
      "epoch": 0.9254787443140338,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000492533256720994,
      "loss": 5.5533,
      "mean_token_accuracy": 0.17563220411539077,
      "num_tokens": 18386728.0,
      "step": 8240
    },
    {
      "entropy": 5.666000461578369,
      "epoch": 0.9260403212219913,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004925229960834256,
      "loss": 5.5451,
      "mean_token_accuracy": 0.17448101937770844,
      "num_tokens": 18397008.0,
      "step": 8245
    },
    {
      "entropy": 5.637777423858642,
      "epoch": 0.9266018981299489,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004925127285198407,
      "loss": 5.5305,
      "mean_token_accuracy": 0.18294473737478256,
      "num_tokens": 18408348.0,
      "step": 8250
    },
    {
      "entropy": 5.5927502632141115,
      "epoch": 0.9271634750379064,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000492502454030566,
      "loss": 5.5562,
      "mean_token_accuracy": 0.18002525418996812,
      "num_tokens": 18419239.0,
      "step": 8255
    },
    {
      "entropy": 5.727730941772461,
      "epoch": 0.927725051945864,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004924921726159288,
      "loss": 5.561,
      "mean_token_accuracy": 0.17431779056787491,
      "num_tokens": 18428918.0,
      "step": 8260
    },
    {
      "entropy": 5.725991296768188,
      "epoch": 0.9282866288538215,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004924818842762565,
      "loss": 5.5523,
      "mean_token_accuracy": 0.1816506192088127,
      "num_tokens": 18440361.0,
      "step": 8265
    },
    {
      "entropy": 5.559592771530151,
      "epoch": 0.9288482057617791,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004924715890118768,
      "loss": 5.4345,
      "mean_token_accuracy": 0.180571249127388,
      "num_tokens": 18450678.0,
      "step": 8270
    },
    {
      "entropy": 5.608138799667358,
      "epoch": 0.9294097826697366,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004924612868231173,
      "loss": 5.6334,
      "mean_token_accuracy": 0.17058404982089997,
      "num_tokens": 18462364.0,
      "step": 8275
    },
    {
      "entropy": 5.645278215408325,
      "epoch": 0.9299713595776942,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004924509777103059,
      "loss": 5.5226,
      "mean_token_accuracy": 0.17628366500139236,
      "num_tokens": 18472896.0,
      "step": 8280
    },
    {
      "entropy": 5.787207078933716,
      "epoch": 0.9305329364856517,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004924406616737712,
      "loss": 5.6589,
      "mean_token_accuracy": 0.16721024066209794,
      "num_tokens": 18484951.0,
      "step": 8285
    },
    {
      "entropy": 5.628914451599121,
      "epoch": 0.9310945133936093,
      "grad_norm": 1.375,
      "learning_rate": 0.0004924303387138415,
      "loss": 5.5925,
      "mean_token_accuracy": 0.174946691095829,
      "num_tokens": 18496006.0,
      "step": 8290
    },
    {
      "entropy": 5.748250484466553,
      "epoch": 0.9316560903015668,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004924200088308455,
      "loss": 5.5748,
      "mean_token_accuracy": 0.17472342401742935,
      "num_tokens": 18508069.0,
      "step": 8295
    },
    {
      "entropy": 5.7464500904083256,
      "epoch": 0.9322176672095244,
      "grad_norm": 1.390625,
      "learning_rate": 0.000492409672025112,
      "loss": 5.6889,
      "mean_token_accuracy": 0.16791522353887559,
      "num_tokens": 18519746.0,
      "step": 8300
    },
    {
      "entropy": 5.742476892471314,
      "epoch": 0.9327792441174819,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004923993282969703,
      "loss": 5.592,
      "mean_token_accuracy": 0.17625355273485183,
      "num_tokens": 18530249.0,
      "step": 8305
    },
    {
      "entropy": 5.6794898986816404,
      "epoch": 0.9333408210254395,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004923889776467495,
      "loss": 5.613,
      "mean_token_accuracy": 0.17979559898376465,
      "num_tokens": 18543306.0,
      "step": 8310
    },
    {
      "entropy": 5.634688186645508,
      "epoch": 0.933902397933397,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004923786200747795,
      "loss": 5.6575,
      "mean_token_accuracy": 0.17098734974861146,
      "num_tokens": 18555250.0,
      "step": 8315
    },
    {
      "entropy": 5.6697102069854735,
      "epoch": 0.9344639748413546,
      "grad_norm": 1.40625,
      "learning_rate": 0.00049236825558139,
      "loss": 5.402,
      "mean_token_accuracy": 0.1727229595184326,
      "num_tokens": 18566378.0,
      "step": 8320
    },
    {
      "entropy": 5.67743182182312,
      "epoch": 0.9350255517493121,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004923578841669108,
      "loss": 5.6108,
      "mean_token_accuracy": 0.17351700663566588,
      "num_tokens": 18576827.0,
      "step": 8325
    },
    {
      "entropy": 5.616914176940918,
      "epoch": 0.9355871286572696,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004923475058316723,
      "loss": 5.5397,
      "mean_token_accuracy": 0.17626303136348725,
      "num_tokens": 18589209.0,
      "step": 8330
    },
    {
      "entropy": 5.6495646953582765,
      "epoch": 0.9361487055652271,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004923371205760049,
      "loss": 5.5529,
      "mean_token_accuracy": 0.17969396114349365,
      "num_tokens": 18599367.0,
      "step": 8335
    },
    {
      "entropy": 5.666165113449097,
      "epoch": 0.9367102824731847,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004923267284002394,
      "loss": 5.5904,
      "mean_token_accuracy": 0.1729588732123375,
      "num_tokens": 18610505.0,
      "step": 8340
    },
    {
      "entropy": 5.692822408676148,
      "epoch": 0.9372718593811422,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004923163293047065,
      "loss": 5.5945,
      "mean_token_accuracy": 0.17593003809452057,
      "num_tokens": 18622616.0,
      "step": 8345
    },
    {
      "entropy": 5.594743490219116,
      "epoch": 0.9378334362890998,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004923059232897374,
      "loss": 5.4359,
      "mean_token_accuracy": 0.18015130013227462,
      "num_tokens": 18634127.0,
      "step": 8350
    },
    {
      "entropy": 5.634627008438111,
      "epoch": 0.9383950131970573,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004922955103556633,
      "loss": 5.6466,
      "mean_token_accuracy": 0.17177557200193405,
      "num_tokens": 18646196.0,
      "step": 8355
    },
    {
      "entropy": 5.645306730270386,
      "epoch": 0.9389565901050149,
      "grad_norm": 1.296875,
      "learning_rate": 0.000492285090502816,
      "loss": 5.552,
      "mean_token_accuracy": 0.1770355746150017,
      "num_tokens": 18656911.0,
      "step": 8360
    },
    {
      "entropy": 5.7078851699829105,
      "epoch": 0.9395181670129724,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004922746637315272,
      "loss": 5.5278,
      "mean_token_accuracy": 0.1755846068263054,
      "num_tokens": 18667553.0,
      "step": 8365
    },
    {
      "entropy": 5.620309591293335,
      "epoch": 0.94007974392093,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004922642300421288,
      "loss": 5.5011,
      "mean_token_accuracy": 0.1790532335639,
      "num_tokens": 18678087.0,
      "step": 8370
    },
    {
      "entropy": 5.6581525802612305,
      "epoch": 0.9406413208288875,
      "grad_norm": 1.484375,
      "learning_rate": 0.000492253789434953,
      "loss": 5.6158,
      "mean_token_accuracy": 0.1750505179166794,
      "num_tokens": 18689946.0,
      "step": 8375
    },
    {
      "entropy": 5.625391101837158,
      "epoch": 0.9412028977368451,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004922433419103323,
      "loss": 5.5269,
      "mean_token_accuracy": 0.18149401843547822,
      "num_tokens": 18700944.0,
      "step": 8380
    },
    {
      "entropy": 5.687411165237426,
      "epoch": 0.9417644746448026,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004922328874685993,
      "loss": 5.6265,
      "mean_token_accuracy": 0.17350946068763734,
      "num_tokens": 18712901.0,
      "step": 8385
    },
    {
      "entropy": 5.6175501346588135,
      "epoch": 0.9423260515527602,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004922224261100869,
      "loss": 5.4589,
      "mean_token_accuracy": 0.17959638684988022,
      "num_tokens": 18724322.0,
      "step": 8390
    },
    {
      "entropy": 5.593675136566162,
      "epoch": 0.9428876284607177,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004922119578351283,
      "loss": 5.5165,
      "mean_token_accuracy": 0.17416234016418458,
      "num_tokens": 18735315.0,
      "step": 8395
    },
    {
      "entropy": 5.61961555480957,
      "epoch": 0.9434492053686753,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004922014826440566,
      "loss": 5.5541,
      "mean_token_accuracy": 0.18015410602092743,
      "num_tokens": 18747497.0,
      "step": 8400
    },
    {
      "entropy": 5.691785049438477,
      "epoch": 0.9440107822766328,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004921910005372056,
      "loss": 5.5891,
      "mean_token_accuracy": 0.1704226091504097,
      "num_tokens": 18758835.0,
      "step": 8405
    },
    {
      "entropy": 5.671810340881348,
      "epoch": 0.9445723591845904,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004921805115149088,
      "loss": 5.483,
      "mean_token_accuracy": 0.18468721210956573,
      "num_tokens": 18770325.0,
      "step": 8410
    },
    {
      "entropy": 5.520982933044434,
      "epoch": 0.9451339360925479,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004921700155775002,
      "loss": 5.3773,
      "mean_token_accuracy": 0.18612601608037949,
      "num_tokens": 18781204.0,
      "step": 8415
    },
    {
      "entropy": 5.526636171340942,
      "epoch": 0.9456955130005055,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004921595127253142,
      "loss": 5.4905,
      "mean_token_accuracy": 0.18205440640449524,
      "num_tokens": 18791899.0,
      "step": 8420
    },
    {
      "entropy": 5.603455257415772,
      "epoch": 0.9462570899084629,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004921490029586852,
      "loss": 5.4779,
      "mean_token_accuracy": 0.1817940965294838,
      "num_tokens": 18803043.0,
      "step": 8425
    },
    {
      "entropy": 5.60422306060791,
      "epoch": 0.9468186668164205,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004921384862779475,
      "loss": 5.4969,
      "mean_token_accuracy": 0.17535765916109086,
      "num_tokens": 18814316.0,
      "step": 8430
    },
    {
      "entropy": 5.586996269226074,
      "epoch": 0.947380243724378,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004921279626834363,
      "loss": 5.5614,
      "mean_token_accuracy": 0.17469020187854767,
      "num_tokens": 18825490.0,
      "step": 8435
    },
    {
      "entropy": 5.758854627609253,
      "epoch": 0.9479418206323356,
      "grad_norm": 1.5,
      "learning_rate": 0.0004921174321754865,
      "loss": 5.6497,
      "mean_token_accuracy": 0.1785335808992386,
      "num_tokens": 18836479.0,
      "step": 8440
    },
    {
      "entropy": 5.655997276306152,
      "epoch": 0.9485033975402931,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004921068947544335,
      "loss": 5.5424,
      "mean_token_accuracy": 0.17748990803956985,
      "num_tokens": 18848634.0,
      "step": 8445
    },
    {
      "entropy": 5.593111944198609,
      "epoch": 0.9490649744482507,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004920963504206128,
      "loss": 5.5003,
      "mean_token_accuracy": 0.18238631188869475,
      "num_tokens": 18858451.0,
      "step": 8450
    },
    {
      "entropy": 5.757823705673218,
      "epoch": 0.9496265513562082,
      "grad_norm": 1.578125,
      "learning_rate": 0.00049208579917436,
      "loss": 5.6392,
      "mean_token_accuracy": 0.16730133891105653,
      "num_tokens": 18869561.0,
      "step": 8455
    },
    {
      "entropy": 5.777918672561645,
      "epoch": 0.9501881282641658,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004920752410160113,
      "loss": 5.6417,
      "mean_token_accuracy": 0.16759925112128257,
      "num_tokens": 18880232.0,
      "step": 8460
    },
    {
      "entropy": 5.7202564716339115,
      "epoch": 0.9507497051721233,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004920646759459026,
      "loss": 5.6543,
      "mean_token_accuracy": 0.17059721201658248,
      "num_tokens": 18892562.0,
      "step": 8465
    },
    {
      "entropy": 5.640997695922851,
      "epoch": 0.9513112820800809,
      "grad_norm": 2.125,
      "learning_rate": 0.0004920541039643705,
      "loss": 5.5844,
      "mean_token_accuracy": 0.17161750495433808,
      "num_tokens": 18903566.0,
      "step": 8470
    },
    {
      "entropy": 5.633873176574707,
      "epoch": 0.9518728589880384,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004920435250717515,
      "loss": 5.5761,
      "mean_token_accuracy": 0.17739747911691667,
      "num_tokens": 18914619.0,
      "step": 8475
    },
    {
      "entropy": 5.76814079284668,
      "epoch": 0.952434435895996,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004920329392683826,
      "loss": 5.6495,
      "mean_token_accuracy": 0.17262647897005082,
      "num_tokens": 18927039.0,
      "step": 8480
    },
    {
      "entropy": 5.777394104003906,
      "epoch": 0.9529960128039535,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004920223465546006,
      "loss": 5.7291,
      "mean_token_accuracy": 0.16832094341516496,
      "num_tokens": 18940449.0,
      "step": 8485
    },
    {
      "entropy": 5.695525121688843,
      "epoch": 0.9535575897119111,
      "grad_norm": 1.625,
      "learning_rate": 0.0004920117469307429,
      "loss": 5.6045,
      "mean_token_accuracy": 0.17119804620742798,
      "num_tokens": 18950798.0,
      "step": 8490
    },
    {
      "entropy": 5.686978483200074,
      "epoch": 0.9541191666198686,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004920011403971471,
      "loss": 5.4587,
      "mean_token_accuracy": 0.17551997900009156,
      "num_tokens": 18960958.0,
      "step": 8495
    },
    {
      "entropy": 5.588903331756592,
      "epoch": 0.9546807435278262,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004919905269541508,
      "loss": 5.4933,
      "mean_token_accuracy": 0.17877306193113326,
      "num_tokens": 18971890.0,
      "step": 8500
    },
    {
      "entropy": 5.560519123077393,
      "epoch": 0.9552423204357837,
      "grad_norm": 1.78125,
      "learning_rate": 0.000491979906602092,
      "loss": 5.4282,
      "mean_token_accuracy": 0.18534162938594817,
      "num_tokens": 18982309.0,
      "step": 8505
    },
    {
      "entropy": 5.67816572189331,
      "epoch": 0.9558038973437413,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004919692793413089,
      "loss": 5.6482,
      "mean_token_accuracy": 0.16850887835025788,
      "num_tokens": 18993648.0,
      "step": 8510
    },
    {
      "entropy": 5.667968082427978,
      "epoch": 0.9563654742516988,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004919586451721396,
      "loss": 5.4243,
      "mean_token_accuracy": 0.18356293588876724,
      "num_tokens": 19003613.0,
      "step": 8515
    },
    {
      "entropy": 5.685746192932129,
      "epoch": 0.9569270511596563,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004919480040949229,
      "loss": 5.5428,
      "mean_token_accuracy": 0.17324323505163192,
      "num_tokens": 19014720.0,
      "step": 8520
    },
    {
      "entropy": 5.643367195129395,
      "epoch": 0.9574886280676138,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004919373561099977,
      "loss": 5.5788,
      "mean_token_accuracy": 0.17302146852016448,
      "num_tokens": 19026635.0,
      "step": 8525
    },
    {
      "entropy": 5.673974132537841,
      "epoch": 0.9580502049755714,
      "grad_norm": 1.6875,
      "learning_rate": 0.000491926701217703,
      "loss": 5.5794,
      "mean_token_accuracy": 0.1681892067193985,
      "num_tokens": 19037433.0,
      "step": 8530
    },
    {
      "entropy": 5.6716055393219,
      "epoch": 0.9586117818835289,
      "grad_norm": 2.375,
      "learning_rate": 0.0004919160394183779,
      "loss": 5.4346,
      "mean_token_accuracy": 0.18276526629924775,
      "num_tokens": 19047351.0,
      "step": 8535
    },
    {
      "entropy": 5.513965892791748,
      "epoch": 0.9591733587914865,
      "grad_norm": 1.34375,
      "learning_rate": 0.000491905370712362,
      "loss": 5.4747,
      "mean_token_accuracy": 0.1769762381911278,
      "num_tokens": 19058561.0,
      "step": 8540
    },
    {
      "entropy": 5.654763650894165,
      "epoch": 0.959734935699444,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004918946950999949,
      "loss": 5.5741,
      "mean_token_accuracy": 0.16921738386154175,
      "num_tokens": 19069877.0,
      "step": 8545
    },
    {
      "entropy": 5.7251872539520265,
      "epoch": 0.9602965126074016,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004918840125816167,
      "loss": 5.5792,
      "mean_token_accuracy": 0.17826244831085206,
      "num_tokens": 19081962.0,
      "step": 8550
    },
    {
      "entropy": 5.646814775466919,
      "epoch": 0.9608580895153591,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004918733231575673,
      "loss": 5.6367,
      "mean_token_accuracy": 0.1751244530081749,
      "num_tokens": 19092955.0,
      "step": 8555
    },
    {
      "entropy": 5.690008735656738,
      "epoch": 0.9614196664233167,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004918626268281874,
      "loss": 5.4435,
      "mean_token_accuracy": 0.18832328021526337,
      "num_tokens": 19103521.0,
      "step": 8560
    },
    {
      "entropy": 5.521415567398071,
      "epoch": 0.9619812433312742,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004918519235938171,
      "loss": 5.4644,
      "mean_token_accuracy": 0.18721356838941575,
      "num_tokens": 19114190.0,
      "step": 8565
    },
    {
      "entropy": 5.628028392791748,
      "epoch": 0.9625428202392318,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004918412134547976,
      "loss": 5.5215,
      "mean_token_accuracy": 0.1752006560564041,
      "num_tokens": 19123815.0,
      "step": 8570
    },
    {
      "entropy": 5.675377082824707,
      "epoch": 0.9631043971471893,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004918304964114696,
      "loss": 5.6645,
      "mean_token_accuracy": 0.17109055519104005,
      "num_tokens": 19135805.0,
      "step": 8575
    },
    {
      "entropy": 5.735991382598877,
      "epoch": 0.9636659740551469,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004918197724641747,
      "loss": 5.4943,
      "mean_token_accuracy": 0.17655572593212127,
      "num_tokens": 19146656.0,
      "step": 8580
    },
    {
      "entropy": 5.620199251174927,
      "epoch": 0.9642275509631044,
      "grad_norm": 1.28125,
      "learning_rate": 0.000491809041613254,
      "loss": 5.5713,
      "mean_token_accuracy": 0.1704343155026436,
      "num_tokens": 19158271.0,
      "step": 8585
    },
    {
      "entropy": 5.580804252624512,
      "epoch": 0.964789127871062,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004917983038590495,
      "loss": 5.484,
      "mean_token_accuracy": 0.17846604883670808,
      "num_tokens": 19168439.0,
      "step": 8590
    },
    {
      "entropy": 5.6851584911346436,
      "epoch": 0.9653507047790195,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004917875592019028,
      "loss": 5.5721,
      "mean_token_accuracy": 0.17893614768981933,
      "num_tokens": 19179411.0,
      "step": 8595
    },
    {
      "entropy": 5.689495515823364,
      "epoch": 0.9659122816869771,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004917768076421561,
      "loss": 5.5917,
      "mean_token_accuracy": 0.174559023976326,
      "num_tokens": 19189377.0,
      "step": 8600
    },
    {
      "entropy": 5.615097713470459,
      "epoch": 0.9664738585949346,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004917660491801518,
      "loss": 5.5604,
      "mean_token_accuracy": 0.17394228130578995,
      "num_tokens": 19200662.0,
      "step": 8605
    },
    {
      "entropy": 5.648424863815308,
      "epoch": 0.9670354355028922,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004917552838162324,
      "loss": 5.4598,
      "mean_token_accuracy": 0.18265042901039125,
      "num_tokens": 19211535.0,
      "step": 8610
    },
    {
      "entropy": 5.628148460388184,
      "epoch": 0.9675970124108496,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004917445115507408,
      "loss": 5.4433,
      "mean_token_accuracy": 0.18630209118127822,
      "num_tokens": 19223402.0,
      "step": 8615
    },
    {
      "entropy": 5.487207746505737,
      "epoch": 0.9681585893188072,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004917337323840198,
      "loss": 5.5248,
      "mean_token_accuracy": 0.17707892805337905,
      "num_tokens": 19234976.0,
      "step": 8620
    },
    {
      "entropy": 5.468156814575195,
      "epoch": 0.9687201662267647,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004917229463164128,
      "loss": 5.3766,
      "mean_token_accuracy": 0.18640546053647994,
      "num_tokens": 19247328.0,
      "step": 8625
    },
    {
      "entropy": 5.678643655776978,
      "epoch": 0.9692817431347223,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004917121533482631,
      "loss": 5.5186,
      "mean_token_accuracy": 0.17082815170288085,
      "num_tokens": 19259021.0,
      "step": 8630
    },
    {
      "entropy": 5.5514545917510985,
      "epoch": 0.9698433200426798,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004917013534799144,
      "loss": 5.4829,
      "mean_token_accuracy": 0.17271644324064256,
      "num_tokens": 19270021.0,
      "step": 8635
    },
    {
      "entropy": 5.603173446655274,
      "epoch": 0.9704048969506374,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004916905467117104,
      "loss": 5.5294,
      "mean_token_accuracy": 0.18102877885103225,
      "num_tokens": 19281520.0,
      "step": 8640
    },
    {
      "entropy": 5.627556180953979,
      "epoch": 0.9709664738585949,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004916797330439955,
      "loss": 5.4723,
      "mean_token_accuracy": 0.1788710445165634,
      "num_tokens": 19292761.0,
      "step": 8645
    },
    {
      "entropy": 5.644547462463379,
      "epoch": 0.9715280507665525,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004916689124771138,
      "loss": 5.5805,
      "mean_token_accuracy": 0.17524223029613495,
      "num_tokens": 19305022.0,
      "step": 8650
    },
    {
      "entropy": 5.744730758666992,
      "epoch": 0.97208962767451,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004916580850114099,
      "loss": 5.6912,
      "mean_token_accuracy": 0.1697470173239708,
      "num_tokens": 19316162.0,
      "step": 8655
    },
    {
      "entropy": 5.853869438171387,
      "epoch": 0.9726512045824676,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004916472506472286,
      "loss": 5.6904,
      "mean_token_accuracy": 0.16199478805065154,
      "num_tokens": 19327015.0,
      "step": 8660
    },
    {
      "entropy": 5.628166675567627,
      "epoch": 0.9732127814904251,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004916364093849147,
      "loss": 5.5602,
      "mean_token_accuracy": 0.1737251363694668,
      "num_tokens": 19338384.0,
      "step": 8665
    },
    {
      "entropy": 5.600831985473633,
      "epoch": 0.9737743583983827,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004916255612248136,
      "loss": 5.4773,
      "mean_token_accuracy": 0.18174067735671998,
      "num_tokens": 19349846.0,
      "step": 8670
    },
    {
      "entropy": 5.7154128074646,
      "epoch": 0.9743359353063402,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004916147061672706,
      "loss": 5.5172,
      "mean_token_accuracy": 0.17738369554281236,
      "num_tokens": 19361399.0,
      "step": 8675
    },
    {
      "entropy": 5.6050217628479,
      "epoch": 0.9748975122142978,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004916038442126314,
      "loss": 5.4925,
      "mean_token_accuracy": 0.1719199925661087,
      "num_tokens": 19371990.0,
      "step": 8680
    },
    {
      "entropy": 5.553276252746582,
      "epoch": 0.9754590891222553,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004915929753612416,
      "loss": 5.4993,
      "mean_token_accuracy": 0.17748630344867705,
      "num_tokens": 19383135.0,
      "step": 8685
    },
    {
      "entropy": 5.666282701492309,
      "epoch": 0.9760206660302129,
      "grad_norm": 2.625,
      "learning_rate": 0.0004915820996134477,
      "loss": 5.5743,
      "mean_token_accuracy": 0.17527981400489806,
      "num_tokens": 19393942.0,
      "step": 8690
    },
    {
      "entropy": 5.59716305732727,
      "epoch": 0.9765822429381704,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004915712169695956,
      "loss": 5.411,
      "mean_token_accuracy": 0.1805658981204033,
      "num_tokens": 19404255.0,
      "step": 8695
    },
    {
      "entropy": 5.596176338195801,
      "epoch": 0.977143819846128,
      "grad_norm": 1.8515625,
      "learning_rate": 0.000491560327430032,
      "loss": 5.5122,
      "mean_token_accuracy": 0.17859420478343963,
      "num_tokens": 19415425.0,
      "step": 8700
    },
    {
      "entropy": 5.594195079803467,
      "epoch": 0.9777053967540855,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004915494309951035,
      "loss": 5.4588,
      "mean_token_accuracy": 0.18495271801948548,
      "num_tokens": 19426050.0,
      "step": 8705
    },
    {
      "entropy": 5.647588205337525,
      "epoch": 0.9782669736620431,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004915385276651573,
      "loss": 5.5586,
      "mean_token_accuracy": 0.1710285097360611,
      "num_tokens": 19437471.0,
      "step": 8710
    },
    {
      "entropy": 5.6465130805969235,
      "epoch": 0.9788285505700005,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004915276174405403,
      "loss": 5.587,
      "mean_token_accuracy": 0.17624323666095734,
      "num_tokens": 19449569.0,
      "step": 8715
    },
    {
      "entropy": 5.666909074783325,
      "epoch": 0.9793901274779581,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004915167003215999,
      "loss": 5.6053,
      "mean_token_accuracy": 0.17028293907642364,
      "num_tokens": 19461630.0,
      "step": 8720
    },
    {
      "entropy": 5.637751913070678,
      "epoch": 0.9799517043859156,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004915057763086838,
      "loss": 5.5337,
      "mean_token_accuracy": 0.17756096571683883,
      "num_tokens": 19472626.0,
      "step": 8725
    },
    {
      "entropy": 5.57206540107727,
      "epoch": 0.9805132812938732,
      "grad_norm": 1.78125,
      "learning_rate": 0.00049149484540214,
      "loss": 5.5253,
      "mean_token_accuracy": 0.1777624398469925,
      "num_tokens": 19483141.0,
      "step": 8730
    },
    {
      "entropy": 5.650867986679077,
      "epoch": 0.9810748582018307,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004914839076023161,
      "loss": 5.499,
      "mean_token_accuracy": 0.18341919481754304,
      "num_tokens": 19495052.0,
      "step": 8735
    },
    {
      "entropy": 5.702249383926391,
      "epoch": 0.9816364351097883,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004914729629095609,
      "loss": 5.5682,
      "mean_token_accuracy": 0.17288218587636947,
      "num_tokens": 19506911.0,
      "step": 8740
    },
    {
      "entropy": 5.562432861328125,
      "epoch": 0.9821980120177458,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004914620113242224,
      "loss": 5.4369,
      "mean_token_accuracy": 0.18190908581018447,
      "num_tokens": 19517812.0,
      "step": 8745
    },
    {
      "entropy": 5.673620414733887,
      "epoch": 0.9827595889257034,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004914510528466495,
      "loss": 5.5298,
      "mean_token_accuracy": 0.176532019674778,
      "num_tokens": 19529857.0,
      "step": 8750
    },
    {
      "entropy": 5.6598528861999515,
      "epoch": 0.9833211658336609,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004914400874771913,
      "loss": 5.5514,
      "mean_token_accuracy": 0.17500680088996887,
      "num_tokens": 19541082.0,
      "step": 8755
    },
    {
      "entropy": 5.5744036674499515,
      "epoch": 0.9838827427416185,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004914291152161967,
      "loss": 5.4743,
      "mean_token_accuracy": 0.18563976138830185,
      "num_tokens": 19551266.0,
      "step": 8760
    },
    {
      "entropy": 5.674076700210572,
      "epoch": 0.984444319649576,
      "grad_norm": 1.59375,
      "learning_rate": 0.000491418136064015,
      "loss": 5.662,
      "mean_token_accuracy": 0.17542957365512848,
      "num_tokens": 19563381.0,
      "step": 8765
    },
    {
      "entropy": 5.752225589752197,
      "epoch": 0.9850058965575336,
      "grad_norm": 1.421875,
      "learning_rate": 0.000491407150020996,
      "loss": 5.6875,
      "mean_token_accuracy": 0.174013951420784,
      "num_tokens": 19575078.0,
      "step": 8770
    },
    {
      "entropy": 5.6871308326721195,
      "epoch": 0.9855674734654911,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004913961570874895,
      "loss": 5.4838,
      "mean_token_accuracy": 0.1798838794231415,
      "num_tokens": 19585273.0,
      "step": 8775
    },
    {
      "entropy": 5.567116832733154,
      "epoch": 0.9861290503734487,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004913851572638453,
      "loss": 5.4983,
      "mean_token_accuracy": 0.17919607013463973,
      "num_tokens": 19594245.0,
      "step": 8780
    },
    {
      "entropy": 5.678967380523682,
      "epoch": 0.9866906272814062,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004913741505504138,
      "loss": 5.5321,
      "mean_token_accuracy": 0.17172959744930266,
      "num_tokens": 19604920.0,
      "step": 8785
    },
    {
      "entropy": 5.7392435550689695,
      "epoch": 0.9872522041893638,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004913631369475454,
      "loss": 5.6313,
      "mean_token_accuracy": 0.17030101865530015,
      "num_tokens": 19616592.0,
      "step": 8790
    },
    {
      "entropy": 5.726884555816651,
      "epoch": 0.9878137810973213,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004913521164555909,
      "loss": 5.5896,
      "mean_token_accuracy": 0.17396857142448424,
      "num_tokens": 19627359.0,
      "step": 8795
    },
    {
      "entropy": 5.6828028678894045,
      "epoch": 0.9883753580052789,
      "grad_norm": 1.859375,
      "learning_rate": 0.000491341089074901,
      "loss": 5.5506,
      "mean_token_accuracy": 0.1802465796470642,
      "num_tokens": 19639685.0,
      "step": 8800
    },
    {
      "entropy": 5.56529130935669,
      "epoch": 0.9889369349132364,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000491330054805827,
      "loss": 5.3945,
      "mean_token_accuracy": 0.18767881393432617,
      "num_tokens": 19650525.0,
      "step": 8805
    },
    {
      "entropy": 5.508602094650269,
      "epoch": 0.9894985118211939,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004913190136487201,
      "loss": 5.4621,
      "mean_token_accuracy": 0.1844386488199234,
      "num_tokens": 19662241.0,
      "step": 8810
    },
    {
      "entropy": 5.670568418502808,
      "epoch": 0.9900600887291514,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004913079656039318,
      "loss": 5.5329,
      "mean_token_accuracy": 0.17866535037755965,
      "num_tokens": 19673752.0,
      "step": 8815
    },
    {
      "entropy": 5.600825834274292,
      "epoch": 0.990621665637109,
      "grad_norm": 1.875,
      "learning_rate": 0.0004912969106718142,
      "loss": 5.5253,
      "mean_token_accuracy": 0.17793213129043578,
      "num_tokens": 19685309.0,
      "step": 8820
    },
    {
      "entropy": 5.5946839332580565,
      "epoch": 0.9911832425450665,
      "grad_norm": 1.46875,
      "learning_rate": 0.000491285848852719,
      "loss": 5.6214,
      "mean_token_accuracy": 0.1743100419640541,
      "num_tokens": 19697391.0,
      "step": 8825
    },
    {
      "entropy": 5.695317173004151,
      "epoch": 0.9917448194530241,
      "grad_norm": 1.5,
      "learning_rate": 0.0004912747801469984,
      "loss": 5.5893,
      "mean_token_accuracy": 0.17411051988601683,
      "num_tokens": 19707426.0,
      "step": 8830
    },
    {
      "entropy": 5.710868787765503,
      "epoch": 0.9923063963609816,
      "grad_norm": 1.65625,
      "learning_rate": 0.000491263704555005,
      "loss": 5.6053,
      "mean_token_accuracy": 0.17150882482528687,
      "num_tokens": 19719800.0,
      "step": 8835
    },
    {
      "entropy": 5.579995489120483,
      "epoch": 0.9928679732689392,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004912526220770913,
      "loss": 5.5116,
      "mean_token_accuracy": 0.17503908276557922,
      "num_tokens": 19731600.0,
      "step": 8840
    },
    {
      "entropy": 5.720822381973266,
      "epoch": 0.9934295501768967,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004912415327136103,
      "loss": 5.5118,
      "mean_token_accuracy": 0.1760959282517433,
      "num_tokens": 19742610.0,
      "step": 8845
    },
    {
      "entropy": 5.776559829711914,
      "epoch": 0.9939911270848543,
      "grad_norm": 1.453125,
      "learning_rate": 0.000491230436464915,
      "loss": 5.721,
      "mean_token_accuracy": 0.1640644207596779,
      "num_tokens": 19754207.0,
      "step": 8850
    },
    {
      "entropy": 5.636538934707642,
      "epoch": 0.9945527039928118,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004912193333313589,
      "loss": 5.4245,
      "mean_token_accuracy": 0.1815497785806656,
      "num_tokens": 19765894.0,
      "step": 8855
    },
    {
      "entropy": 5.662306404113769,
      "epoch": 0.9951142809007694,
      "grad_norm": 2.0,
      "learning_rate": 0.0004912082233132951,
      "loss": 5.5738,
      "mean_token_accuracy": 0.17184434235095977,
      "num_tokens": 19777867.0,
      "step": 8860
    },
    {
      "entropy": 5.616130352020264,
      "epoch": 0.9956758578087269,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004911971064110778,
      "loss": 5.461,
      "mean_token_accuracy": 0.1819846212863922,
      "num_tokens": 19788346.0,
      "step": 8865
    },
    {
      "entropy": 5.745035696029663,
      "epoch": 0.9962374347166845,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004911859826250607,
      "loss": 5.6139,
      "mean_token_accuracy": 0.18041548728942872,
      "num_tokens": 19799752.0,
      "step": 8870
    },
    {
      "entropy": 5.649567937850952,
      "epoch": 0.996799011624642,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004911748519555981,
      "loss": 5.5514,
      "mean_token_accuracy": 0.1762547880411148,
      "num_tokens": 19810387.0,
      "step": 8875
    },
    {
      "entropy": 5.572876071929931,
      "epoch": 0.9973605885325996,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004911637144030444,
      "loss": 5.521,
      "mean_token_accuracy": 0.17834608554840087,
      "num_tokens": 19822092.0,
      "step": 8880
    },
    {
      "entropy": 5.707835960388183,
      "epoch": 0.9979221654405571,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004911525699677541,
      "loss": 5.6279,
      "mean_token_accuracy": 0.16719121038913726,
      "num_tokens": 19833731.0,
      "step": 8885
    },
    {
      "entropy": 5.668497037887573,
      "epoch": 0.9984837423485147,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004911414186500822,
      "loss": 5.5636,
      "mean_token_accuracy": 0.1742876499891281,
      "num_tokens": 19845101.0,
      "step": 8890
    },
    {
      "entropy": 5.6994829177856445,
      "epoch": 0.9990453192564722,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004911302604503837,
      "loss": 5.6231,
      "mean_token_accuracy": 0.17030643969774245,
      "num_tokens": 19857006.0,
      "step": 8895
    },
    {
      "entropy": 5.674806213378906,
      "epoch": 0.9996068961644298,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004911190953690139,
      "loss": 5.4978,
      "mean_token_accuracy": 0.18147013634443282,
      "num_tokens": 19868007.0,
      "step": 8900
    },
    {
      "entropy": 5.7907646497090655,
      "epoch": 1.0001123153815914,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004911079234063284,
      "loss": 5.4717,
      "mean_token_accuracy": 0.18072778979937235,
      "num_tokens": 19877897.0,
      "step": 8905
    },
    {
      "entropy": 5.650493812561035,
      "epoch": 1.000673892289549,
      "grad_norm": 1.75,
      "learning_rate": 0.0004910967445626826,
      "loss": 5.496,
      "mean_token_accuracy": 0.17796591967344283,
      "num_tokens": 19888449.0,
      "step": 8910
    },
    {
      "entropy": 5.612166070938111,
      "epoch": 1.0012354691975065,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004910855588384327,
      "loss": 5.4678,
      "mean_token_accuracy": 0.17649304419755935,
      "num_tokens": 19898998.0,
      "step": 8915
    },
    {
      "entropy": 5.6738122463226315,
      "epoch": 1.0017970461054642,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004910743662339348,
      "loss": 5.4831,
      "mean_token_accuracy": 0.18333123475313187,
      "num_tokens": 19910369.0,
      "step": 8920
    },
    {
      "entropy": 5.721463680267334,
      "epoch": 1.0023586230134216,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004910631667495452,
      "loss": 5.5219,
      "mean_token_accuracy": 0.1823783427476883,
      "num_tokens": 19921936.0,
      "step": 8925
    },
    {
      "entropy": 5.657253789901733,
      "epoch": 1.0029201999213793,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004910519603856207,
      "loss": 5.5097,
      "mean_token_accuracy": 0.17559026032686234,
      "num_tokens": 19933943.0,
      "step": 8930
    },
    {
      "entropy": 5.494042158126831,
      "epoch": 1.0034817768293367,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004910407471425179,
      "loss": 5.3106,
      "mean_token_accuracy": 0.18985848128795624,
      "num_tokens": 19944377.0,
      "step": 8935
    },
    {
      "entropy": 5.521006155014038,
      "epoch": 1.0040433537372944,
      "grad_norm": 1.6875,
      "learning_rate": 0.000491029527020594,
      "loss": 5.4144,
      "mean_token_accuracy": 0.18066316545009614,
      "num_tokens": 19954865.0,
      "step": 8940
    },
    {
      "entropy": 5.543346166610718,
      "epoch": 1.0046049306452518,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004910183000202061,
      "loss": 5.3318,
      "mean_token_accuracy": 0.19103278070688248,
      "num_tokens": 19965021.0,
      "step": 8945
    },
    {
      "entropy": 5.634443902969361,
      "epoch": 1.0051665075532095,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004910070661417119,
      "loss": 5.4471,
      "mean_token_accuracy": 0.17768314927816392,
      "num_tokens": 19977068.0,
      "step": 8950
    },
    {
      "entropy": 5.5534903049469,
      "epoch": 1.005728084461167,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004909958253854687,
      "loss": 5.474,
      "mean_token_accuracy": 0.17667020857334137,
      "num_tokens": 19987994.0,
      "step": 8955
    },
    {
      "entropy": 5.651862049102784,
      "epoch": 1.0062896613691246,
      "grad_norm": 1.75,
      "learning_rate": 0.0004909845777518347,
      "loss": 5.4939,
      "mean_token_accuracy": 0.17481962591409683,
      "num_tokens": 19998598.0,
      "step": 8960
    },
    {
      "entropy": 5.622123050689697,
      "epoch": 1.006851238277082,
      "grad_norm": 2.984375,
      "learning_rate": 0.0004909733232411681,
      "loss": 5.385,
      "mean_token_accuracy": 0.1836555555462837,
      "num_tokens": 20008881.0,
      "step": 8965
    },
    {
      "entropy": 5.586540174484253,
      "epoch": 1.0074128151850397,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004909620618538271,
      "loss": 5.4177,
      "mean_token_accuracy": 0.17951451539993285,
      "num_tokens": 20019105.0,
      "step": 8970
    },
    {
      "entropy": 5.579048728942871,
      "epoch": 1.007974392092997,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004909507935901702,
      "loss": 5.4769,
      "mean_token_accuracy": 0.17605098336935043,
      "num_tokens": 20031805.0,
      "step": 8975
    },
    {
      "entropy": 5.62233076095581,
      "epoch": 1.0085359690009548,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004909395184505564,
      "loss": 5.4234,
      "mean_token_accuracy": 0.18534507006406784,
      "num_tokens": 20041493.0,
      "step": 8980
    },
    {
      "entropy": 5.607890892028808,
      "epoch": 1.0090975459089122,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004909282364353445,
      "loss": 5.4807,
      "mean_token_accuracy": 0.18110638111829758,
      "num_tokens": 20052489.0,
      "step": 8985
    },
    {
      "entropy": 5.46955451965332,
      "epoch": 1.0096591228168699,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004909169475448938,
      "loss": 5.4328,
      "mean_token_accuracy": 0.1797465994954109,
      "num_tokens": 20064252.0,
      "step": 8990
    },
    {
      "entropy": 5.621907567977905,
      "epoch": 1.0102206997248273,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004909056517795638,
      "loss": 5.3744,
      "mean_token_accuracy": 0.17497411817312242,
      "num_tokens": 20075171.0,
      "step": 8995
    },
    {
      "entropy": 5.624486780166626,
      "epoch": 1.010782276632785,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004908943491397141,
      "loss": 5.376,
      "mean_token_accuracy": 0.18161926418542862,
      "num_tokens": 20085746.0,
      "step": 9000
    },
    {
      "epoch": 1.010782276632785,
      "eval_entropy": 5.459080059872112,
      "eval_loss": 5.578691005706787,
      "eval_mean_token_accuracy": 0.1821541891581993,
      "eval_num_tokens": 20085746.0,
      "eval_runtime": 23.814,
      "eval_samples_per_second": 1302.304,
      "eval_steps_per_second": 162.804,
      "step": 9000
    },
    {
      "entropy": 5.554215002059936,
      "epoch": 1.0113438535407424,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004908830396257045,
      "loss": 5.4536,
      "mean_token_accuracy": 0.17544384002685548,
      "num_tokens": 20097435.0,
      "step": 9005
    },
    {
      "entropy": 5.565097761154175,
      "epoch": 1.0119054304486998,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004908717232378953,
      "loss": 5.4455,
      "mean_token_accuracy": 0.17849185615777968,
      "num_tokens": 20108100.0,
      "step": 9010
    },
    {
      "entropy": 5.682046175003052,
      "epoch": 1.0124670073566575,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004908603999766467,
      "loss": 5.4581,
      "mean_token_accuracy": 0.17797257453203202,
      "num_tokens": 20119042.0,
      "step": 9015
    },
    {
      "entropy": 5.565086078643799,
      "epoch": 1.013028584264615,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004908490698423191,
      "loss": 5.5158,
      "mean_token_accuracy": 0.17789216041564943,
      "num_tokens": 20130275.0,
      "step": 9020
    },
    {
      "entropy": 5.555480098724365,
      "epoch": 1.0135901611725726,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004908377328352736,
      "loss": 5.4629,
      "mean_token_accuracy": 0.17693307846784592,
      "num_tokens": 20142523.0,
      "step": 9025
    },
    {
      "entropy": 5.602287340164184,
      "epoch": 1.01415173808053,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004908263889558709,
      "loss": 5.4208,
      "mean_token_accuracy": 0.18608217239379882,
      "num_tokens": 20153732.0,
      "step": 9030
    },
    {
      "entropy": 5.6788112163543705,
      "epoch": 1.0147133149884877,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004908150382044722,
      "loss": 5.4983,
      "mean_token_accuracy": 0.17320408672094345,
      "num_tokens": 20165136.0,
      "step": 9035
    },
    {
      "entropy": 5.568822240829467,
      "epoch": 1.0152748918964452,
      "grad_norm": 1.359375,
      "learning_rate": 0.000490803680581439,
      "loss": 5.4291,
      "mean_token_accuracy": 0.18082266598939895,
      "num_tokens": 20175326.0,
      "step": 9040
    },
    {
      "entropy": 5.587617254257202,
      "epoch": 1.0158364688044028,
      "grad_norm": 1.84375,
      "learning_rate": 0.000490792316087133,
      "loss": 5.3695,
      "mean_token_accuracy": 0.18594837188720703,
      "num_tokens": 20186678.0,
      "step": 9045
    },
    {
      "entropy": 5.539478063583374,
      "epoch": 1.0163980457123603,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004907809447219158,
      "loss": 5.3786,
      "mean_token_accuracy": 0.18022306710481645,
      "num_tokens": 20196696.0,
      "step": 9050
    },
    {
      "entropy": 5.565349864959717,
      "epoch": 1.016959622620318,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004907695664861498,
      "loss": 5.4002,
      "mean_token_accuracy": 0.18411532640457154,
      "num_tokens": 20207566.0,
      "step": 9055
    },
    {
      "entropy": 5.537447357177735,
      "epoch": 1.0175211995282754,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004907581813801971,
      "loss": 5.4409,
      "mean_token_accuracy": 0.17487737089395522,
      "num_tokens": 20218233.0,
      "step": 9060
    },
    {
      "entropy": 5.616069173812866,
      "epoch": 1.018082776436233,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004907467894044202,
      "loss": 5.3577,
      "mean_token_accuracy": 0.18108492195606232,
      "num_tokens": 20228524.0,
      "step": 9065
    },
    {
      "entropy": 5.5232336044311525,
      "epoch": 1.0186443533441905,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004907353905591817,
      "loss": 5.4619,
      "mean_token_accuracy": 0.1800857111811638,
      "num_tokens": 20240821.0,
      "step": 9070
    },
    {
      "entropy": 5.537438440322876,
      "epoch": 1.0192059302521481,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000490723984844845,
      "loss": 5.4308,
      "mean_token_accuracy": 0.18032869696617126,
      "num_tokens": 20252282.0,
      "step": 9075
    },
    {
      "entropy": 5.6138331413269045,
      "epoch": 1.0197675071601056,
      "grad_norm": 1.375,
      "learning_rate": 0.0004907125722617727,
      "loss": 5.4367,
      "mean_token_accuracy": 0.18050792068243027,
      "num_tokens": 20263462.0,
      "step": 9080
    },
    {
      "entropy": 5.6108558177948,
      "epoch": 1.0203290840680632,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004907011528103285,
      "loss": 5.4753,
      "mean_token_accuracy": 0.17466021329164505,
      "num_tokens": 20275062.0,
      "step": 9085
    },
    {
      "entropy": 5.608004331588745,
      "epoch": 1.0208906609760207,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004906897264908759,
      "loss": 5.4529,
      "mean_token_accuracy": 0.1767593130469322,
      "num_tokens": 20286243.0,
      "step": 9090
    },
    {
      "entropy": 5.569535255432129,
      "epoch": 1.0214522378839783,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004906782933037789,
      "loss": 5.4607,
      "mean_token_accuracy": 0.17864596843719482,
      "num_tokens": 20296929.0,
      "step": 9095
    },
    {
      "entropy": 5.589272975921631,
      "epoch": 1.0220138147919358,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004906668532494012,
      "loss": 5.3494,
      "mean_token_accuracy": 0.18342964649200438,
      "num_tokens": 20308354.0,
      "step": 9100
    },
    {
      "entropy": 5.610014581680298,
      "epoch": 1.0225753916998932,
      "grad_norm": 1.75,
      "learning_rate": 0.0004906554063281074,
      "loss": 5.5226,
      "mean_token_accuracy": 0.1788520559668541,
      "num_tokens": 20319487.0,
      "step": 9105
    },
    {
      "entropy": 5.516573715209961,
      "epoch": 1.0231369686078509,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004906439525402618,
      "loss": 5.4296,
      "mean_token_accuracy": 0.18768656700849534,
      "num_tokens": 20329967.0,
      "step": 9110
    },
    {
      "entropy": 5.652112627029419,
      "epoch": 1.0236985455158083,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004906324918862291,
      "loss": 5.5061,
      "mean_token_accuracy": 0.17268284261226655,
      "num_tokens": 20341442.0,
      "step": 9115
    },
    {
      "entropy": 5.684236621856689,
      "epoch": 1.024260122423766,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004906210243663742,
      "loss": 5.6184,
      "mean_token_accuracy": 0.1690230995416641,
      "num_tokens": 20352818.0,
      "step": 9120
    },
    {
      "entropy": 5.731576585769654,
      "epoch": 1.0248216993317234,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004906095499810622,
      "loss": 5.5189,
      "mean_token_accuracy": 0.17407773584127426,
      "num_tokens": 20364787.0,
      "step": 9125
    },
    {
      "entropy": 5.6368835926055905,
      "epoch": 1.025383276239681,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004905980687306586,
      "loss": 5.3863,
      "mean_token_accuracy": 0.193366876244545,
      "num_tokens": 20375491.0,
      "step": 9130
    },
    {
      "entropy": 5.580576801300049,
      "epoch": 1.0259448531476385,
      "grad_norm": 1.5,
      "learning_rate": 0.0004905865806155288,
      "loss": 5.449,
      "mean_token_accuracy": 0.18474821895360946,
      "num_tokens": 20386270.0,
      "step": 9135
    },
    {
      "entropy": 5.697377729415893,
      "epoch": 1.0265064300555962,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004905750856360387,
      "loss": 5.5493,
      "mean_token_accuracy": 0.17359945625066758,
      "num_tokens": 20397953.0,
      "step": 9140
    },
    {
      "entropy": 5.658065843582153,
      "epoch": 1.0270680069635536,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004905635837925543,
      "loss": 5.5395,
      "mean_token_accuracy": 0.17324478775262833,
      "num_tokens": 20409847.0,
      "step": 9145
    },
    {
      "entropy": 5.628514909744263,
      "epoch": 1.0276295838715113,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004905520750854418,
      "loss": 5.5244,
      "mean_token_accuracy": 0.18113597482442856,
      "num_tokens": 20420737.0,
      "step": 9150
    },
    {
      "entropy": 5.598775434494018,
      "epoch": 1.0281911607794687,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004905405595150674,
      "loss": 5.4158,
      "mean_token_accuracy": 0.17999387234449388,
      "num_tokens": 20432395.0,
      "step": 9155
    },
    {
      "entropy": 5.570962142944336,
      "epoch": 1.0287527376874264,
      "grad_norm": 1.375,
      "learning_rate": 0.0004905290370817982,
      "loss": 5.4113,
      "mean_token_accuracy": 0.1848587855696678,
      "num_tokens": 20442964.0,
      "step": 9160
    },
    {
      "entropy": 5.589252853393555,
      "epoch": 1.0293143145953838,
      "grad_norm": 1.46875,
      "learning_rate": 0.000490517507786001,
      "loss": 5.4535,
      "mean_token_accuracy": 0.1752680480480194,
      "num_tokens": 20454050.0,
      "step": 9165
    },
    {
      "entropy": 5.611699819564819,
      "epoch": 1.0298758915033415,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004905059716280425,
      "loss": 5.4671,
      "mean_token_accuracy": 0.17207229137420654,
      "num_tokens": 20465335.0,
      "step": 9170
    },
    {
      "entropy": 5.591488790512085,
      "epoch": 1.030437468411299,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004904944286082904,
      "loss": 5.439,
      "mean_token_accuracy": 0.1793852284550667,
      "num_tokens": 20476781.0,
      "step": 9175
    },
    {
      "entropy": 5.579262685775757,
      "epoch": 1.0309990453192566,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004904828787271121,
      "loss": 5.4773,
      "mean_token_accuracy": 0.17789653241634368,
      "num_tokens": 20488599.0,
      "step": 9180
    },
    {
      "entropy": 5.502214145660401,
      "epoch": 1.031560622227214,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004904713219848753,
      "loss": 5.3622,
      "mean_token_accuracy": 0.18779534250497817,
      "num_tokens": 20498215.0,
      "step": 9185
    },
    {
      "entropy": 5.621550893783569,
      "epoch": 1.0321221991351717,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004904597583819481,
      "loss": 5.3901,
      "mean_token_accuracy": 0.1827726513147354,
      "num_tokens": 20509764.0,
      "step": 9190
    },
    {
      "entropy": 5.771495628356933,
      "epoch": 1.032683776043129,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004904481879186987,
      "loss": 5.6073,
      "mean_token_accuracy": 0.17733301371335983,
      "num_tokens": 20521582.0,
      "step": 9195
    },
    {
      "entropy": 5.563735485076904,
      "epoch": 1.0332453529510865,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004904366105954954,
      "loss": 5.4825,
      "mean_token_accuracy": 0.17996705174446107,
      "num_tokens": 20532050.0,
      "step": 9200
    },
    {
      "entropy": 5.589707326889038,
      "epoch": 1.0338069298590442,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004904250264127067,
      "loss": 5.4864,
      "mean_token_accuracy": 0.17125514298677444,
      "num_tokens": 20543733.0,
      "step": 9205
    },
    {
      "entropy": 5.674970865249634,
      "epoch": 1.0343685067670016,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004904134353707018,
      "loss": 5.511,
      "mean_token_accuracy": 0.17877715229988098,
      "num_tokens": 20554898.0,
      "step": 9210
    },
    {
      "entropy": 5.539908313751221,
      "epoch": 1.0349300836749593,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004904018374698496,
      "loss": 5.3852,
      "mean_token_accuracy": 0.18103183954954147,
      "num_tokens": 20565243.0,
      "step": 9215
    },
    {
      "entropy": 5.526424217224121,
      "epoch": 1.0354916605829168,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004903902327105191,
      "loss": 5.3783,
      "mean_token_accuracy": 0.18287476301193237,
      "num_tokens": 20575321.0,
      "step": 9220
    },
    {
      "entropy": 5.642215061187744,
      "epoch": 1.0360532374908744,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004903786210930803,
      "loss": 5.55,
      "mean_token_accuracy": 0.16923462599515915,
      "num_tokens": 20586450.0,
      "step": 9225
    },
    {
      "entropy": 5.652804899215698,
      "epoch": 1.0366148143988319,
      "grad_norm": 1.375,
      "learning_rate": 0.0004903670026179025,
      "loss": 5.5276,
      "mean_token_accuracy": 0.18359616696834563,
      "num_tokens": 20596996.0,
      "step": 9230
    },
    {
      "entropy": 5.593594217300415,
      "epoch": 1.0371763913067895,
      "grad_norm": 3.09375,
      "learning_rate": 0.000490355377285356,
      "loss": 5.4321,
      "mean_token_accuracy": 0.1745638743042946,
      "num_tokens": 20609136.0,
      "step": 9235
    },
    {
      "entropy": 5.649332761764526,
      "epoch": 1.037737968214747,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004903437450958106,
      "loss": 5.509,
      "mean_token_accuracy": 0.17722001522779465,
      "num_tokens": 20620246.0,
      "step": 9240
    },
    {
      "entropy": 5.5271008014678955,
      "epoch": 1.0382995451227046,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004903321060496369,
      "loss": 5.4402,
      "mean_token_accuracy": 0.17385232001543044,
      "num_tokens": 20631678.0,
      "step": 9245
    },
    {
      "entropy": 5.603906536102295,
      "epoch": 1.038861122030662,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004903204601472054,
      "loss": 5.3884,
      "mean_token_accuracy": 0.18640923500061035,
      "num_tokens": 20643287.0,
      "step": 9250
    },
    {
      "entropy": 5.604735088348389,
      "epoch": 1.0394226989386197,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004903088073888869,
      "loss": 5.4314,
      "mean_token_accuracy": 0.18265591114759444,
      "num_tokens": 20655091.0,
      "step": 9255
    },
    {
      "entropy": 5.589794301986695,
      "epoch": 1.0399842758465772,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004902971477750526,
      "loss": 5.4618,
      "mean_token_accuracy": 0.18000126481056214,
      "num_tokens": 20665043.0,
      "step": 9260
    },
    {
      "entropy": 5.543551874160767,
      "epoch": 1.0405458527545348,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004902854813060736,
      "loss": 5.3904,
      "mean_token_accuracy": 0.18817973136901855,
      "num_tokens": 20675225.0,
      "step": 9265
    },
    {
      "entropy": 5.551910877227783,
      "epoch": 1.0411074296624923,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004902738079823213,
      "loss": 5.4656,
      "mean_token_accuracy": 0.17844348698854445,
      "num_tokens": 20686542.0,
      "step": 9270
    },
    {
      "entropy": 5.541519689559936,
      "epoch": 1.04166900657045,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004902621278041674,
      "loss": 5.3959,
      "mean_token_accuracy": 0.18315559178590773,
      "num_tokens": 20697926.0,
      "step": 9275
    },
    {
      "entropy": 5.589543962478638,
      "epoch": 1.0422305834784074,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004902504407719839,
      "loss": 5.5135,
      "mean_token_accuracy": 0.17791294157505036,
      "num_tokens": 20707890.0,
      "step": 9280
    },
    {
      "entropy": 5.623344087600708,
      "epoch": 1.042792160386365,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004902387468861429,
      "loss": 5.4447,
      "mean_token_accuracy": 0.17739499360322952,
      "num_tokens": 20719554.0,
      "step": 9285
    },
    {
      "entropy": 5.6102501392364506,
      "epoch": 1.0433537372943225,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004902270461470169,
      "loss": 5.4078,
      "mean_token_accuracy": 0.18083870112895967,
      "num_tokens": 20730899.0,
      "step": 9290
    },
    {
      "entropy": 5.58493447303772,
      "epoch": 1.04391531420228,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000490215338554978,
      "loss": 5.522,
      "mean_token_accuracy": 0.17751231491565705,
      "num_tokens": 20741839.0,
      "step": 9295
    },
    {
      "entropy": 5.591024684906006,
      "epoch": 1.0444768911102376,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004902036241103994,
      "loss": 5.4857,
      "mean_token_accuracy": 0.18067652583122254,
      "num_tokens": 20753199.0,
      "step": 9300
    },
    {
      "entropy": 5.661981391906738,
      "epoch": 1.045038468018195,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004901919028136539,
      "loss": 5.5496,
      "mean_token_accuracy": 0.17503437399864197,
      "num_tokens": 20764827.0,
      "step": 9305
    },
    {
      "entropy": 5.610328722000122,
      "epoch": 1.0456000449261527,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004901801746651147,
      "loss": 5.4256,
      "mean_token_accuracy": 0.18545343726873398,
      "num_tokens": 20774953.0,
      "step": 9310
    },
    {
      "entropy": 5.571823215484619,
      "epoch": 1.04616162183411,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004901684396651554,
      "loss": 5.4427,
      "mean_token_accuracy": 0.18092939853668213,
      "num_tokens": 20784428.0,
      "step": 9315
    },
    {
      "entropy": 5.598528337478638,
      "epoch": 1.0467231987420678,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004901566978141494,
      "loss": 5.512,
      "mean_token_accuracy": 0.17941481471061707,
      "num_tokens": 20795672.0,
      "step": 9320
    },
    {
      "entropy": 5.651431894302368,
      "epoch": 1.0472847756500252,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004901449491124708,
      "loss": 5.5068,
      "mean_token_accuracy": 0.1770174965262413,
      "num_tokens": 20806972.0,
      "step": 9325
    },
    {
      "entropy": 5.64578070640564,
      "epoch": 1.0478463525579829,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004901331935604936,
      "loss": 5.5248,
      "mean_token_accuracy": 0.17678644359111786,
      "num_tokens": 20817867.0,
      "step": 9330
    },
    {
      "entropy": 5.498245143890381,
      "epoch": 1.0484079294659403,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000490121431158592,
      "loss": 5.3575,
      "mean_token_accuracy": 0.18257005363702775,
      "num_tokens": 20829205.0,
      "step": 9335
    },
    {
      "entropy": 5.572736358642578,
      "epoch": 1.048969506373898,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004901096619071407,
      "loss": 5.4797,
      "mean_token_accuracy": 0.18204336166381835,
      "num_tokens": 20840478.0,
      "step": 9340
    },
    {
      "entropy": 5.658757066726684,
      "epoch": 1.0495310832818554,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004900978858065142,
      "loss": 5.4306,
      "mean_token_accuracy": 0.18008593916893006,
      "num_tokens": 20850752.0,
      "step": 9345
    },
    {
      "entropy": 5.604446268081665,
      "epoch": 1.050092660189813,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004900861028570877,
      "loss": 5.4501,
      "mean_token_accuracy": 0.17572760283946992,
      "num_tokens": 20862261.0,
      "step": 9350
    },
    {
      "entropy": 5.5822326183319095,
      "epoch": 1.0506542370977705,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004900743130592362,
      "loss": 5.4086,
      "mean_token_accuracy": 0.18594421893358232,
      "num_tokens": 20872604.0,
      "step": 9355
    },
    {
      "entropy": 5.4873597621917725,
      "epoch": 1.0512158140057282,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004900625164133351,
      "loss": 5.3491,
      "mean_token_accuracy": 0.19344383776187896,
      "num_tokens": 20882751.0,
      "step": 9360
    },
    {
      "entropy": 5.581871128082275,
      "epoch": 1.0517773909136856,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004900507129197601,
      "loss": 5.4704,
      "mean_token_accuracy": 0.17706203758716582,
      "num_tokens": 20893809.0,
      "step": 9365
    },
    {
      "entropy": 5.692767477035522,
      "epoch": 1.0523389678216433,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004900389025788871,
      "loss": 5.3872,
      "mean_token_accuracy": 0.18640353083610534,
      "num_tokens": 20904201.0,
      "step": 9370
    },
    {
      "entropy": 5.554631233215332,
      "epoch": 1.0529005447296007,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000490027085391092,
      "loss": 5.4679,
      "mean_token_accuracy": 0.17571618407964706,
      "num_tokens": 20915226.0,
      "step": 9375
    },
    {
      "entropy": 5.554903602600097,
      "epoch": 1.0534621216375584,
      "grad_norm": 1.59375,
      "learning_rate": 0.000490015261356751,
      "loss": 5.4384,
      "mean_token_accuracy": 0.1765777975320816,
      "num_tokens": 20926040.0,
      "step": 9380
    },
    {
      "entropy": 5.656567096710205,
      "epoch": 1.0540236985455158,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004900034304762408,
      "loss": 5.4719,
      "mean_token_accuracy": 0.18317525684833527,
      "num_tokens": 20937051.0,
      "step": 9385
    },
    {
      "entropy": 5.586511945724487,
      "epoch": 1.0545852754534732,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000489991592749938,
      "loss": 5.4956,
      "mean_token_accuracy": 0.17753477841615678,
      "num_tokens": 20947810.0,
      "step": 9390
    },
    {
      "entropy": 5.570503187179566,
      "epoch": 1.055146852361431,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004899797481782193,
      "loss": 5.4455,
      "mean_token_accuracy": 0.17481617033481597,
      "num_tokens": 20958307.0,
      "step": 9395
    },
    {
      "entropy": 5.539948797225952,
      "epoch": 1.0557084292693883,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004899678967614623,
      "loss": 5.3425,
      "mean_token_accuracy": 0.18573132604360582,
      "num_tokens": 20969493.0,
      "step": 9400
    },
    {
      "entropy": 5.6041279315948485,
      "epoch": 1.056270006177346,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004899560385000439,
      "loss": 5.505,
      "mean_token_accuracy": 0.1813835933804512,
      "num_tokens": 20980224.0,
      "step": 9405
    },
    {
      "entropy": 5.611564826965332,
      "epoch": 1.0568315830853034,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000489944173394342,
      "loss": 5.465,
      "mean_token_accuracy": 0.18255140036344528,
      "num_tokens": 20991282.0,
      "step": 9410
    },
    {
      "entropy": 5.660673666000366,
      "epoch": 1.0573931599932611,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004899323014447342,
      "loss": 5.4889,
      "mean_token_accuracy": 0.17426551580429078,
      "num_tokens": 21002395.0,
      "step": 9415
    },
    {
      "entropy": 5.554814577102661,
      "epoch": 1.0579547369012186,
      "grad_norm": 1.75,
      "learning_rate": 0.0004899204226515985,
      "loss": 5.4145,
      "mean_token_accuracy": 0.17696548700332643,
      "num_tokens": 21014136.0,
      "step": 9420
    },
    {
      "entropy": 5.621621894836426,
      "epoch": 1.0585163138091762,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004899085370153133,
      "loss": 5.5344,
      "mean_token_accuracy": 0.1757613092660904,
      "num_tokens": 21025637.0,
      "step": 9425
    },
    {
      "entropy": 5.781061601638794,
      "epoch": 1.0590778907171337,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004898966445362569,
      "loss": 5.5789,
      "mean_token_accuracy": 0.1691136360168457,
      "num_tokens": 21036393.0,
      "step": 9430
    },
    {
      "entropy": 5.683920192718506,
      "epoch": 1.0596394676250913,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004898847452148079,
      "loss": 5.5224,
      "mean_token_accuracy": 0.1836437076330185,
      "num_tokens": 21047339.0,
      "step": 9435
    },
    {
      "entropy": 5.654892826080323,
      "epoch": 1.0602010445330488,
      "grad_norm": 1.75,
      "learning_rate": 0.0004898728390513454,
      "loss": 5.6236,
      "mean_token_accuracy": 0.17097096592187883,
      "num_tokens": 21059427.0,
      "step": 9440
    },
    {
      "entropy": 5.628730916976929,
      "epoch": 1.0607626214410064,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004898609260462483,
      "loss": 5.4506,
      "mean_token_accuracy": 0.17322685420513154,
      "num_tokens": 21068666.0,
      "step": 9445
    },
    {
      "entropy": 5.65942964553833,
      "epoch": 1.0613241983489639,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004898490061998961,
      "loss": 5.4612,
      "mean_token_accuracy": 0.17689654380083084,
      "num_tokens": 21080161.0,
      "step": 9450
    },
    {
      "entropy": 5.6597339630126955,
      "epoch": 1.0618857752569215,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004898370795126681,
      "loss": 5.5456,
      "mean_token_accuracy": 0.17456338703632354,
      "num_tokens": 21092363.0,
      "step": 9455
    },
    {
      "entropy": 5.5144401550292965,
      "epoch": 1.062447352164879,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004898251459849441,
      "loss": 5.4438,
      "mean_token_accuracy": 0.18251396864652633,
      "num_tokens": 21103683.0,
      "step": 9460
    },
    {
      "entropy": 5.56109733581543,
      "epoch": 1.0630089290728366,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004898132056171041,
      "loss": 5.3997,
      "mean_token_accuracy": 0.18426249623298646,
      "num_tokens": 21115110.0,
      "step": 9465
    },
    {
      "entropy": 5.616233873367309,
      "epoch": 1.063570505980794,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004898012584095285,
      "loss": 5.3795,
      "mean_token_accuracy": 0.1885729670524597,
      "num_tokens": 21124327.0,
      "step": 9470
    },
    {
      "entropy": 5.565642499923706,
      "epoch": 1.0641320828887517,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004897893043625975,
      "loss": 5.5125,
      "mean_token_accuracy": 0.17969844341278077,
      "num_tokens": 21135542.0,
      "step": 9475
    },
    {
      "entropy": 5.537214708328247,
      "epoch": 1.0646936597967092,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004897773434766917,
      "loss": 5.4006,
      "mean_token_accuracy": 0.17806485891342164,
      "num_tokens": 21147621.0,
      "step": 9480
    },
    {
      "entropy": 5.565270471572876,
      "epoch": 1.0652552367046666,
      "grad_norm": 1.46875,
      "learning_rate": 0.000489765375752192,
      "loss": 5.4054,
      "mean_token_accuracy": 0.18055632710456848,
      "num_tokens": 21159725.0,
      "step": 9485
    },
    {
      "entropy": 5.556740951538086,
      "epoch": 1.0658168136126243,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004897534011894793,
      "loss": 5.4692,
      "mean_token_accuracy": 0.1807580292224884,
      "num_tokens": 21171303.0,
      "step": 9490
    },
    {
      "entropy": 5.706756448745727,
      "epoch": 1.0663783905205817,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004897414197889352,
      "loss": 5.5675,
      "mean_token_accuracy": 0.17455703616142274,
      "num_tokens": 21183376.0,
      "step": 9495
    },
    {
      "entropy": 5.659519624710083,
      "epoch": 1.0669399674285394,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000489729431550941,
      "loss": 5.4447,
      "mean_token_accuracy": 0.17345247715711593,
      "num_tokens": 21194365.0,
      "step": 9500
    },
    {
      "entropy": 5.444472789764404,
      "epoch": 1.0675015443364968,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004897174364758783,
      "loss": 5.3153,
      "mean_token_accuracy": 0.18512341380119324,
      "num_tokens": 21206440.0,
      "step": 9505
    },
    {
      "entropy": 5.455184984207153,
      "epoch": 1.0680631212444545,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004897054345641291,
      "loss": 5.369,
      "mean_token_accuracy": 0.18379151225090026,
      "num_tokens": 21217188.0,
      "step": 9510
    },
    {
      "entropy": 5.570040655136109,
      "epoch": 1.068624698152412,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004896934258160757,
      "loss": 5.4465,
      "mean_token_accuracy": 0.17723545730113982,
      "num_tokens": 21227716.0,
      "step": 9515
    },
    {
      "entropy": 5.643634176254272,
      "epoch": 1.0691862750603696,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004896814102321004,
      "loss": 5.5694,
      "mean_token_accuracy": 0.17157238572835923,
      "num_tokens": 21241258.0,
      "step": 9520
    },
    {
      "entropy": 5.599542284011841,
      "epoch": 1.069747851968327,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004896693878125857,
      "loss": 5.3809,
      "mean_token_accuracy": 0.1786022424697876,
      "num_tokens": 21252543.0,
      "step": 9525
    },
    {
      "entropy": 5.616940689086914,
      "epoch": 1.0703094288762847,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004896573585579144,
      "loss": 5.4974,
      "mean_token_accuracy": 0.17120815068483353,
      "num_tokens": 21264829.0,
      "step": 9530
    },
    {
      "entropy": 5.66899471282959,
      "epoch": 1.070871005784242,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004896453224684695,
      "loss": 5.4994,
      "mean_token_accuracy": 0.179149766266346,
      "num_tokens": 21276240.0,
      "step": 9535
    },
    {
      "entropy": 5.620451879501343,
      "epoch": 1.0714325826921998,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004896332795446345,
      "loss": 5.5129,
      "mean_token_accuracy": 0.17478324174880983,
      "num_tokens": 21288877.0,
      "step": 9540
    },
    {
      "entropy": 5.644775009155273,
      "epoch": 1.0719941596001572,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004896212297867924,
      "loss": 5.5498,
      "mean_token_accuracy": 0.17083704620599746,
      "num_tokens": 21300147.0,
      "step": 9545
    },
    {
      "entropy": 5.564145803451538,
      "epoch": 1.0725557365081149,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004896091731953271,
      "loss": 5.3379,
      "mean_token_accuracy": 0.1834096446633339,
      "num_tokens": 21309731.0,
      "step": 9550
    },
    {
      "entropy": 5.54395546913147,
      "epoch": 1.0731173134160723,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004895971097706227,
      "loss": 5.413,
      "mean_token_accuracy": 0.17877567708492278,
      "num_tokens": 21320358.0,
      "step": 9555
    },
    {
      "entropy": 5.487132596969604,
      "epoch": 1.07367889032403,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004895850395130629,
      "loss": 5.3176,
      "mean_token_accuracy": 0.19092220366001128,
      "num_tokens": 21331077.0,
      "step": 9560
    },
    {
      "entropy": 5.564539003372192,
      "epoch": 1.0742404672319874,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004895729624230323,
      "loss": 5.5168,
      "mean_token_accuracy": 0.18183142989873885,
      "num_tokens": 21341304.0,
      "step": 9565
    },
    {
      "entropy": 5.6389447212219235,
      "epoch": 1.074802044139945,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004895608785009153,
      "loss": 5.495,
      "mean_token_accuracy": 0.1784301996231079,
      "num_tokens": 21353663.0,
      "step": 9570
    },
    {
      "entropy": 5.57873568534851,
      "epoch": 1.0753636210479025,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004895487877470969,
      "loss": 5.4624,
      "mean_token_accuracy": 0.17929406017065047,
      "num_tokens": 21364987.0,
      "step": 9575
    },
    {
      "entropy": 5.507581281661987,
      "epoch": 1.07592519795586,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004895366901619617,
      "loss": 5.4332,
      "mean_token_accuracy": 0.17917827069759368,
      "num_tokens": 21376169.0,
      "step": 9580
    },
    {
      "entropy": 5.57572569847107,
      "epoch": 1.0764867748638176,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004895245857458951,
      "loss": 5.409,
      "mean_token_accuracy": 0.17978260666131973,
      "num_tokens": 21387141.0,
      "step": 9585
    },
    {
      "entropy": 5.60054030418396,
      "epoch": 1.077048351771775,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004895124744992825,
      "loss": 5.392,
      "mean_token_accuracy": 0.18548154681921006,
      "num_tokens": 21396804.0,
      "step": 9590
    },
    {
      "entropy": 5.509835720062256,
      "epoch": 1.0776099286797327,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004895003564225095,
      "loss": 5.4276,
      "mean_token_accuracy": 0.18453317284584045,
      "num_tokens": 21406948.0,
      "step": 9595
    },
    {
      "entropy": 5.553013610839844,
      "epoch": 1.0781715055876901,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004894882315159621,
      "loss": 5.4084,
      "mean_token_accuracy": 0.18084955364465713,
      "num_tokens": 21417560.0,
      "step": 9600
    },
    {
      "entropy": 5.538203001022339,
      "epoch": 1.0787330824956478,
      "grad_norm": 1.71875,
      "learning_rate": 0.000489476099780026,
      "loss": 5.4176,
      "mean_token_accuracy": 0.18558336198329925,
      "num_tokens": 21428596.0,
      "step": 9605
    },
    {
      "entropy": 5.594052362442016,
      "epoch": 1.0792946594036052,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004894639612150879,
      "loss": 5.5009,
      "mean_token_accuracy": 0.1776292786002159,
      "num_tokens": 21440004.0,
      "step": 9610
    },
    {
      "entropy": 5.603554058074951,
      "epoch": 1.079856236311563,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004894518158215339,
      "loss": 5.437,
      "mean_token_accuracy": 0.18182192593812943,
      "num_tokens": 21451423.0,
      "step": 9615
    },
    {
      "entropy": 5.6315145015716555,
      "epoch": 1.0804178132195204,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000489439663599751,
      "loss": 5.502,
      "mean_token_accuracy": 0.17420097291469575,
      "num_tokens": 21462065.0,
      "step": 9620
    },
    {
      "entropy": 5.565780210494995,
      "epoch": 1.080979390127478,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000489427504550126,
      "loss": 5.4329,
      "mean_token_accuracy": 0.17715296298265457,
      "num_tokens": 21473476.0,
      "step": 9625
    },
    {
      "entropy": 5.676029396057129,
      "epoch": 1.0815409670354355,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004894153386730462,
      "loss": 5.5517,
      "mean_token_accuracy": 0.1776492029428482,
      "num_tokens": 21485591.0,
      "step": 9630
    },
    {
      "entropy": 5.641856098175049,
      "epoch": 1.0821025439433931,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004894031659688988,
      "loss": 5.6286,
      "mean_token_accuracy": 0.17196228802204133,
      "num_tokens": 21497016.0,
      "step": 9635
    },
    {
      "entropy": 5.563565683364868,
      "epoch": 1.0826641208513506,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004893909864380714,
      "loss": 5.4241,
      "mean_token_accuracy": 0.18089785128831865,
      "num_tokens": 21508179.0,
      "step": 9640
    },
    {
      "entropy": 5.661916208267212,
      "epoch": 1.0832256977593082,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000489378800080952,
      "loss": 5.5514,
      "mean_token_accuracy": 0.16957199722528457,
      "num_tokens": 21520053.0,
      "step": 9645
    },
    {
      "entropy": 5.662601566314697,
      "epoch": 1.0837872746672657,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004893666068979283,
      "loss": 5.4513,
      "mean_token_accuracy": 0.18505332469940186,
      "num_tokens": 21530746.0,
      "step": 9650
    },
    {
      "entropy": 5.561122179031372,
      "epoch": 1.0843488515752233,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004893544068893887,
      "loss": 5.4699,
      "mean_token_accuracy": 0.17847533524036407,
      "num_tokens": 21540829.0,
      "step": 9655
    },
    {
      "entropy": 5.55997633934021,
      "epoch": 1.0849104284831808,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004893422000557217,
      "loss": 5.4025,
      "mean_token_accuracy": 0.1824072703719139,
      "num_tokens": 21552337.0,
      "step": 9660
    },
    {
      "entropy": 5.600814437866211,
      "epoch": 1.0854720053911384,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004893299863973159,
      "loss": 5.4138,
      "mean_token_accuracy": 0.18376973569393157,
      "num_tokens": 21563752.0,
      "step": 9665
    },
    {
      "entropy": 5.642785453796387,
      "epoch": 1.0860335822990959,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004893177659145601,
      "loss": 5.5625,
      "mean_token_accuracy": 0.17684261351823807,
      "num_tokens": 21574834.0,
      "step": 9670
    },
    {
      "entropy": 5.6293717384338375,
      "epoch": 1.0865951592070533,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004893055386078438,
      "loss": 5.4672,
      "mean_token_accuracy": 0.17502747774124144,
      "num_tokens": 21586674.0,
      "step": 9675
    },
    {
      "entropy": 5.620609474182129,
      "epoch": 1.087156736115011,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004892933044775558,
      "loss": 5.403,
      "mean_token_accuracy": 0.1848546639084816,
      "num_tokens": 21595998.0,
      "step": 9680
    },
    {
      "entropy": 5.600088262557984,
      "epoch": 1.0877183130229684,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004892810635240859,
      "loss": 5.484,
      "mean_token_accuracy": 0.17975264489650727,
      "num_tokens": 21607122.0,
      "step": 9685
    },
    {
      "entropy": 5.5601122856140135,
      "epoch": 1.088279889930926,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004892688157478238,
      "loss": 5.3974,
      "mean_token_accuracy": 0.18241830170154572,
      "num_tokens": 21618780.0,
      "step": 9690
    },
    {
      "entropy": 5.5932230949401855,
      "epoch": 1.0888414668388835,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004892565611491597,
      "loss": 5.4617,
      "mean_token_accuracy": 0.1802744448184967,
      "num_tokens": 21628868.0,
      "step": 9695
    },
    {
      "entropy": 5.6138755798339846,
      "epoch": 1.0894030437468412,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004892442997284834,
      "loss": 5.3442,
      "mean_token_accuracy": 0.19054587185382843,
      "num_tokens": 21640189.0,
      "step": 9700
    },
    {
      "entropy": 5.5858148574829105,
      "epoch": 1.0899646206547986,
      "grad_norm": 1.25,
      "learning_rate": 0.0004892320314861855,
      "loss": 5.5968,
      "mean_token_accuracy": 0.17798891812562942,
      "num_tokens": 21651452.0,
      "step": 9705
    },
    {
      "entropy": 5.505543899536133,
      "epoch": 1.0905261975627563,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004892197564226568,
      "loss": 5.3238,
      "mean_token_accuracy": 0.18632598519325255,
      "num_tokens": 21662555.0,
      "step": 9710
    },
    {
      "entropy": 5.577691555023193,
      "epoch": 1.0910877744707137,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004892074745382878,
      "loss": 5.3675,
      "mean_token_accuracy": 0.18600122332572938,
      "num_tokens": 21673087.0,
      "step": 9715
    },
    {
      "entropy": 5.559349918365479,
      "epoch": 1.0916493513786714,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004891951858334698,
      "loss": 5.4592,
      "mean_token_accuracy": 0.17866172939538955,
      "num_tokens": 21684084.0,
      "step": 9720
    },
    {
      "entropy": 5.5244140625,
      "epoch": 1.0922109282866288,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004891828903085939,
      "loss": 5.4018,
      "mean_token_accuracy": 0.1834845647215843,
      "num_tokens": 21695765.0,
      "step": 9725
    },
    {
      "entropy": 5.6586473941802975,
      "epoch": 1.0927725051945865,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004891705879640518,
      "loss": 5.4871,
      "mean_token_accuracy": 0.17867584228515626,
      "num_tokens": 21707396.0,
      "step": 9730
    },
    {
      "entropy": 5.589698791503906,
      "epoch": 1.093334082102544,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004891582788002351,
      "loss": 5.4206,
      "mean_token_accuracy": 0.1840369150042534,
      "num_tokens": 21718330.0,
      "step": 9735
    },
    {
      "entropy": 5.627910423278808,
      "epoch": 1.0938956590105016,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004891459628175358,
      "loss": 5.5333,
      "mean_token_accuracy": 0.17498715370893478,
      "num_tokens": 21729841.0,
      "step": 9740
    },
    {
      "entropy": 5.586634826660156,
      "epoch": 1.094457235918459,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000489133640016346,
      "loss": 5.4779,
      "mean_token_accuracy": 0.18365336507558822,
      "num_tokens": 21740403.0,
      "step": 9745
    },
    {
      "entropy": 5.6356080055236815,
      "epoch": 1.0950188128264167,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004891213103970579,
      "loss": 5.4621,
      "mean_token_accuracy": 0.18089024871587753,
      "num_tokens": 21750696.0,
      "step": 9750
    },
    {
      "entropy": 5.52034821510315,
      "epoch": 1.095580389734374,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004891089739600643,
      "loss": 5.361,
      "mean_token_accuracy": 0.18476004898548126,
      "num_tokens": 21761324.0,
      "step": 9755
    },
    {
      "entropy": 5.6163732528686525,
      "epoch": 1.0961419666423318,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004890966307057581,
      "loss": 5.4558,
      "mean_token_accuracy": 0.1767112657427788,
      "num_tokens": 21774046.0,
      "step": 9760
    },
    {
      "entropy": 5.5504951000213625,
      "epoch": 1.0967035435502892,
      "grad_norm": 1.96875,
      "learning_rate": 0.000489084280634532,
      "loss": 5.4554,
      "mean_token_accuracy": 0.1811685934662819,
      "num_tokens": 21784625.0,
      "step": 9765
    },
    {
      "entropy": 5.644229459762573,
      "epoch": 1.0972651204582466,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004890719237467794,
      "loss": 5.556,
      "mean_token_accuracy": 0.17533084750175476,
      "num_tokens": 21795971.0,
      "step": 9770
    },
    {
      "entropy": 5.702903461456299,
      "epoch": 1.0978266973662043,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004890595600428937,
      "loss": 5.517,
      "mean_token_accuracy": 0.17991171777248383,
      "num_tokens": 21806624.0,
      "step": 9775
    },
    {
      "entropy": 5.603931522369384,
      "epoch": 1.0983882742741617,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004890471895232684,
      "loss": 5.4447,
      "mean_token_accuracy": 0.1773332729935646,
      "num_tokens": 21817830.0,
      "step": 9780
    },
    {
      "entropy": 5.577806615829468,
      "epoch": 1.0989498511821194,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004890348121882978,
      "loss": 5.4297,
      "mean_token_accuracy": 0.18548934012651444,
      "num_tokens": 21829540.0,
      "step": 9785
    },
    {
      "entropy": 5.702232933044433,
      "epoch": 1.0995114280900768,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004890224280383758,
      "loss": 5.6124,
      "mean_token_accuracy": 0.17456040531396866,
      "num_tokens": 21840550.0,
      "step": 9790
    },
    {
      "entropy": 5.6826036930084225,
      "epoch": 1.1000730049980345,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004890100370738966,
      "loss": 5.5526,
      "mean_token_accuracy": 0.17297013401985167,
      "num_tokens": 21853366.0,
      "step": 9795
    },
    {
      "entropy": 5.567359399795532,
      "epoch": 1.100634581905992,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004889976392952548,
      "loss": 5.4677,
      "mean_token_accuracy": 0.1834318980574608,
      "num_tokens": 21863522.0,
      "step": 9800
    },
    {
      "entropy": 5.617341470718384,
      "epoch": 1.1011961588139496,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004889852347028452,
      "loss": 5.427,
      "mean_token_accuracy": 0.18069212436676024,
      "num_tokens": 21873988.0,
      "step": 9805
    },
    {
      "entropy": 5.5770001888275145,
      "epoch": 1.101757735721907,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004889728232970626,
      "loss": 5.3645,
      "mean_token_accuracy": 0.18551873415708542,
      "num_tokens": 21884595.0,
      "step": 9810
    },
    {
      "entropy": 5.559672927856445,
      "epoch": 1.1023193126298647,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004889604050783025,
      "loss": 5.4049,
      "mean_token_accuracy": 0.18870704919099807,
      "num_tokens": 21894643.0,
      "step": 9815
    },
    {
      "entropy": 5.608105182647705,
      "epoch": 1.1028808895378222,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048894798004696,
      "loss": 5.4151,
      "mean_token_accuracy": 0.18534329533576965,
      "num_tokens": 21906608.0,
      "step": 9820
    },
    {
      "entropy": 5.72157301902771,
      "epoch": 1.1034424664457798,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004889355482034309,
      "loss": 5.5454,
      "mean_token_accuracy": 0.17869509905576705,
      "num_tokens": 21919386.0,
      "step": 9825
    },
    {
      "entropy": 5.524839878082275,
      "epoch": 1.1040040433537373,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004889231095481111,
      "loss": 5.4214,
      "mean_token_accuracy": 0.18391776829957962,
      "num_tokens": 21930554.0,
      "step": 9830
    },
    {
      "entropy": 5.654389429092407,
      "epoch": 1.104565620261695,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004889106640813964,
      "loss": 5.5247,
      "mean_token_accuracy": 0.17893804013729095,
      "num_tokens": 21941764.0,
      "step": 9835
    },
    {
      "entropy": 5.54154601097107,
      "epoch": 1.1051271971696524,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004888982118036831,
      "loss": 5.351,
      "mean_token_accuracy": 0.18850825279951094,
      "num_tokens": 21951804.0,
      "step": 9840
    },
    {
      "entropy": 5.612653160095215,
      "epoch": 1.10568877407761,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004888857527153679,
      "loss": 5.5164,
      "mean_token_accuracy": 0.1715184286236763,
      "num_tokens": 21963356.0,
      "step": 9845
    },
    {
      "entropy": 5.585508584976196,
      "epoch": 1.1062503509855675,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004888732868168473,
      "loss": 5.4373,
      "mean_token_accuracy": 0.18316725194454192,
      "num_tokens": 21973785.0,
      "step": 9850
    },
    {
      "entropy": 5.632301330566406,
      "epoch": 1.1068119278935251,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004888608141085183,
      "loss": 5.4351,
      "mean_token_accuracy": 0.17661972790956498,
      "num_tokens": 21985057.0,
      "step": 9855
    },
    {
      "entropy": 5.623175954818725,
      "epoch": 1.1073735048014826,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004888483345907781,
      "loss": 5.5038,
      "mean_token_accuracy": 0.18079903870820999,
      "num_tokens": 21996554.0,
      "step": 9860
    },
    {
      "entropy": 5.564160585403442,
      "epoch": 1.10793508170944,
      "grad_norm": 1.390625,
      "learning_rate": 0.000488835848264024,
      "loss": 5.4428,
      "mean_token_accuracy": 0.17722105979919434,
      "num_tokens": 22008284.0,
      "step": 9865
    },
    {
      "entropy": 5.536996221542358,
      "epoch": 1.1084966586173977,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004888233551286534,
      "loss": 5.4887,
      "mean_token_accuracy": 0.1781219720840454,
      "num_tokens": 22019322.0,
      "step": 9870
    },
    {
      "entropy": 5.670701122283935,
      "epoch": 1.109058235525355,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004888108551850644,
      "loss": 5.5697,
      "mean_token_accuracy": 0.16928135007619857,
      "num_tokens": 22031444.0,
      "step": 9875
    },
    {
      "entropy": 5.598286533355713,
      "epoch": 1.1096198124333128,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004887983484336548,
      "loss": 5.3813,
      "mean_token_accuracy": 0.18684137463569642,
      "num_tokens": 22042704.0,
      "step": 9880
    },
    {
      "entropy": 5.579814386367798,
      "epoch": 1.1101813893412702,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004887858348748229,
      "loss": 5.4732,
      "mean_token_accuracy": 0.18037536442279817,
      "num_tokens": 22052721.0,
      "step": 9885
    },
    {
      "entropy": 5.651267671585083,
      "epoch": 1.1107429662492279,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000488773314508967,
      "loss": 5.4894,
      "mean_token_accuracy": 0.17873190343379974,
      "num_tokens": 22063857.0,
      "step": 9890
    },
    {
      "entropy": 5.5537378787994385,
      "epoch": 1.1113045431571853,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004887607873364859,
      "loss": 5.423,
      "mean_token_accuracy": 0.183661812543869,
      "num_tokens": 22074777.0,
      "step": 9895
    },
    {
      "entropy": 5.627150440216065,
      "epoch": 1.111866120065143,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004887482533577786,
      "loss": 5.5092,
      "mean_token_accuracy": 0.17283634841442108,
      "num_tokens": 22085301.0,
      "step": 9900
    },
    {
      "entropy": 5.6233783721923825,
      "epoch": 1.1124276969731004,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004887357125732438,
      "loss": 5.4252,
      "mean_token_accuracy": 0.1815194621682167,
      "num_tokens": 22095635.0,
      "step": 9905
    },
    {
      "entropy": 5.591199684143066,
      "epoch": 1.112989273881058,
      "grad_norm": 1.984375,
      "learning_rate": 0.000488723164983281,
      "loss": 5.4432,
      "mean_token_accuracy": 0.17690271586179734,
      "num_tokens": 22105965.0,
      "step": 9910
    },
    {
      "entropy": 5.499814510345459,
      "epoch": 1.1135508507890155,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004887106105882898,
      "loss": 5.3705,
      "mean_token_accuracy": 0.1817806765437126,
      "num_tokens": 22117122.0,
      "step": 9915
    },
    {
      "entropy": 5.653459739685059,
      "epoch": 1.1141124276969732,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004886980493886699,
      "loss": 5.4242,
      "mean_token_accuracy": 0.18367586880922318,
      "num_tokens": 22128642.0,
      "step": 9920
    },
    {
      "entropy": 5.592040252685547,
      "epoch": 1.1146740046049306,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004886854813848212,
      "loss": 5.4187,
      "mean_token_accuracy": 0.1819361686706543,
      "num_tokens": 22140260.0,
      "step": 9925
    },
    {
      "entropy": 5.504345655441284,
      "epoch": 1.1152355815128883,
      "grad_norm": 1.5,
      "learning_rate": 0.0004886729065771439,
      "loss": 5.5463,
      "mean_token_accuracy": 0.17537332624197005,
      "num_tokens": 22152460.0,
      "step": 9930
    },
    {
      "entropy": 5.600211524963379,
      "epoch": 1.1157971584208457,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004886603249660384,
      "loss": 5.3025,
      "mean_token_accuracy": 0.18542418330907823,
      "num_tokens": 22163134.0,
      "step": 9935
    },
    {
      "entropy": 5.666353702545166,
      "epoch": 1.1163587353288034,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004886477365519052,
      "loss": 5.4622,
      "mean_token_accuracy": 0.17874416559934617,
      "num_tokens": 22173956.0,
      "step": 9940
    },
    {
      "entropy": 5.529520034790039,
      "epoch": 1.1169203122367608,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004886351413351453,
      "loss": 5.4463,
      "mean_token_accuracy": 0.17844821363687516,
      "num_tokens": 22185506.0,
      "step": 9945
    },
    {
      "entropy": 5.517136430740356,
      "epoch": 1.1174818891447185,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004886225393161596,
      "loss": 5.4379,
      "mean_token_accuracy": 0.1817207857966423,
      "num_tokens": 22196934.0,
      "step": 9950
    },
    {
      "entropy": 5.585705089569092,
      "epoch": 1.118043466052676,
      "grad_norm": 1.375,
      "learning_rate": 0.0004886099304953495,
      "loss": 5.4573,
      "mean_token_accuracy": 0.17639363408088685,
      "num_tokens": 22209013.0,
      "step": 9955
    },
    {
      "entropy": 5.688259077072144,
      "epoch": 1.1186050429606333,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004885973148731162,
      "loss": 5.5883,
      "mean_token_accuracy": 0.1751970171928406,
      "num_tokens": 22220890.0,
      "step": 9960
    },
    {
      "entropy": 5.6196691513061525,
      "epoch": 1.119166619868591,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004885846924498617,
      "loss": 5.4045,
      "mean_token_accuracy": 0.18046572655439377,
      "num_tokens": 22232899.0,
      "step": 9965
    },
    {
      "entropy": 5.553362655639648,
      "epoch": 1.1197281967765484,
      "grad_norm": 1.375,
      "learning_rate": 0.0004885720632259876,
      "loss": 5.4285,
      "mean_token_accuracy": 0.17943229377269745,
      "num_tokens": 22243839.0,
      "step": 9970
    },
    {
      "entropy": 5.6693480014801025,
      "epoch": 1.120289773684506,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004885594272018963,
      "loss": 5.4645,
      "mean_token_accuracy": 0.17993124127388,
      "num_tokens": 22253970.0,
      "step": 9975
    },
    {
      "entropy": 5.530833959579468,
      "epoch": 1.1208513505924635,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048854678437799,
      "loss": 5.3467,
      "mean_token_accuracy": 0.18927840441465377,
      "num_tokens": 22263598.0,
      "step": 9980
    },
    {
      "entropy": 5.483437776565552,
      "epoch": 1.1214129275004212,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004885341347546713,
      "loss": 5.5413,
      "mean_token_accuracy": 0.17272351831197738,
      "num_tokens": 22274385.0,
      "step": 9985
    },
    {
      "entropy": 5.579334831237793,
      "epoch": 1.1219745044083786,
      "grad_norm": 2.0,
      "learning_rate": 0.0004885214783323428,
      "loss": 5.4064,
      "mean_token_accuracy": 0.1820865660905838,
      "num_tokens": 22285837.0,
      "step": 9990
    },
    {
      "entropy": 5.6654328346252445,
      "epoch": 1.1225360813163363,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004885088151114076,
      "loss": 5.5008,
      "mean_token_accuracy": 0.17633887529373168,
      "num_tokens": 22296882.0,
      "step": 9995
    },
    {
      "entropy": 5.509787464141846,
      "epoch": 1.1230976582242937,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004884961450922691,
      "loss": 5.3913,
      "mean_token_accuracy": 0.18201883286237716,
      "num_tokens": 22308110.0,
      "step": 10000
    },
    {
      "entropy": 5.510275602340698,
      "epoch": 1.1236592351322514,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004884834682753305,
      "loss": 5.4198,
      "mean_token_accuracy": 0.18542241752147676,
      "num_tokens": 22318541.0,
      "step": 10005
    },
    {
      "entropy": 5.536625671386719,
      "epoch": 1.1242208120402089,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004884707846609953,
      "loss": 5.4128,
      "mean_token_accuracy": 0.18137142658233643,
      "num_tokens": 22328713.0,
      "step": 10010
    },
    {
      "entropy": 5.661680412292481,
      "epoch": 1.1247823889481665,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004884580942496677,
      "loss": 5.4751,
      "mean_token_accuracy": 0.17806675732135774,
      "num_tokens": 22341050.0,
      "step": 10015
    },
    {
      "entropy": 5.58024754524231,
      "epoch": 1.125343965856124,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004884453970417515,
      "loss": 5.3569,
      "mean_token_accuracy": 0.18561971783638,
      "num_tokens": 22351184.0,
      "step": 10020
    },
    {
      "entropy": 5.555691528320312,
      "epoch": 1.1259055427640816,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000488432693037651,
      "loss": 5.4249,
      "mean_token_accuracy": 0.18320182114839553,
      "num_tokens": 22361651.0,
      "step": 10025
    },
    {
      "entropy": 5.648286199569702,
      "epoch": 1.126467119672039,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000488419982237771,
      "loss": 5.532,
      "mean_token_accuracy": 0.17271254062652588,
      "num_tokens": 22373672.0,
      "step": 10030
    },
    {
      "entropy": 5.636139965057373,
      "epoch": 1.1270286965799967,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004884072646425157,
      "loss": 5.5091,
      "mean_token_accuracy": 0.1779921069741249,
      "num_tokens": 22384924.0,
      "step": 10035
    },
    {
      "entropy": 5.6113080978393555,
      "epoch": 1.1275902734879542,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004883945402522907,
      "loss": 5.4102,
      "mean_token_accuracy": 0.18566200882196426,
      "num_tokens": 22396091.0,
      "step": 10040
    },
    {
      "entropy": 5.697112512588501,
      "epoch": 1.1281518503959118,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004883818090675005,
      "loss": 5.5023,
      "mean_token_accuracy": 0.17411854714155198,
      "num_tokens": 22407925.0,
      "step": 10045
    },
    {
      "entropy": 5.580139541625977,
      "epoch": 1.1287134273038693,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004883690710885507,
      "loss": 5.4206,
      "mean_token_accuracy": 0.18167119473218918,
      "num_tokens": 22417917.0,
      "step": 10050
    },
    {
      "entropy": 5.612313365936279,
      "epoch": 1.1292750042118267,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004883563263158471,
      "loss": 5.5153,
      "mean_token_accuracy": 0.17684295028448105,
      "num_tokens": 22429836.0,
      "step": 10055
    },
    {
      "entropy": 5.661643743515015,
      "epoch": 1.1298365811197844,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004883435747497952,
      "loss": 5.585,
      "mean_token_accuracy": 0.17540506571531295,
      "num_tokens": 22440720.0,
      "step": 10060
    },
    {
      "entropy": 5.5146032810211185,
      "epoch": 1.130398158027742,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004883308163908012,
      "loss": 5.3064,
      "mean_token_accuracy": 0.18266170024871825,
      "num_tokens": 22452587.0,
      "step": 10065
    },
    {
      "entropy": 5.582793140411377,
      "epoch": 1.1309597349356995,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004883180512392714,
      "loss": 5.4896,
      "mean_token_accuracy": 0.18124875873327256,
      "num_tokens": 22464063.0,
      "step": 10070
    },
    {
      "entropy": 5.57654972076416,
      "epoch": 1.131521311843657,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004883052792956119,
      "loss": 5.4545,
      "mean_token_accuracy": 0.1777760937809944,
      "num_tokens": 22474956.0,
      "step": 10075
    },
    {
      "entropy": 5.572151613235474,
      "epoch": 1.1320828887516146,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004882925005602296,
      "loss": 5.4247,
      "mean_token_accuracy": 0.17617421299219133,
      "num_tokens": 22486499.0,
      "step": 10080
    },
    {
      "entropy": 5.5781739234924315,
      "epoch": 1.132644465659572,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004882797150335314,
      "loss": 5.3386,
      "mean_token_accuracy": 0.18663488775491716,
      "num_tokens": 22497736.0,
      "step": 10085
    },
    {
      "entropy": 5.540691328048706,
      "epoch": 1.1332060425675297,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048826692271592434,
      "loss": 5.4169,
      "mean_token_accuracy": 0.17964298129081727,
      "num_tokens": 22509997.0,
      "step": 10090
    },
    {
      "entropy": 5.574131822586059,
      "epoch": 1.133767619475487,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004882541236078158,
      "loss": 5.4949,
      "mean_token_accuracy": 0.18007486909627915,
      "num_tokens": 22519721.0,
      "step": 10095
    },
    {
      "entropy": 5.661850929260254,
      "epoch": 1.1343291963834448,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048824131770961316,
      "loss": 5.3899,
      "mean_token_accuracy": 0.1807657703757286,
      "num_tokens": 22530219.0,
      "step": 10100
    },
    {
      "entropy": 5.590708684921265,
      "epoch": 1.1348907732914022,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048822850502172437,
      "loss": 5.4249,
      "mean_token_accuracy": 0.17643967866897584,
      "num_tokens": 22541373.0,
      "step": 10105
    },
    {
      "entropy": 5.462498426437378,
      "epoch": 1.1354523501993599,
      "grad_norm": 1.625,
      "learning_rate": 0.0004882156855445571,
      "loss": 5.3623,
      "mean_token_accuracy": 0.17757754027843475,
      "num_tokens": 22552974.0,
      "step": 10110
    },
    {
      "entropy": 5.580391216278076,
      "epoch": 1.1360139271073173,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004882028592785199,
      "loss": 5.3721,
      "mean_token_accuracy": 0.18000393509864807,
      "num_tokens": 22563468.0,
      "step": 10115
    },
    {
      "entropy": 5.569001054763794,
      "epoch": 1.136575504015275,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048819002622402095,
      "loss": 5.5069,
      "mean_token_accuracy": 0.17577992826700212,
      "num_tokens": 22574386.0,
      "step": 10120
    },
    {
      "entropy": 5.556534337997436,
      "epoch": 1.1371370809232324,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004881771863814688,
      "loss": 5.4123,
      "mean_token_accuracy": 0.18740794956684112,
      "num_tokens": 22585369.0,
      "step": 10125
    },
    {
      "entropy": 5.537963247299194,
      "epoch": 1.13769865783119,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048816433975127244,
      "loss": 5.4334,
      "mean_token_accuracy": 0.18183194398880004,
      "num_tokens": 22595853.0,
      "step": 10130
    },
    {
      "entropy": 5.610812759399414,
      "epoch": 1.1382602347391475,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004881514863338407,
      "loss": 5.5028,
      "mean_token_accuracy": 0.18175214976072313,
      "num_tokens": 22606596.0,
      "step": 10135
    },
    {
      "entropy": 5.613374710083008,
      "epoch": 1.1388218116471052,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048813862612958307,
      "loss": 5.3744,
      "mean_token_accuracy": 0.17816888988018037,
      "num_tokens": 22617196.0,
      "step": 10140
    },
    {
      "entropy": 5.451364135742187,
      "epoch": 1.1393833885550626,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004881257591389089,
      "loss": 5.3344,
      "mean_token_accuracy": 0.18494184762239457,
      "num_tokens": 22627817.0,
      "step": 10145
    },
    {
      "entropy": 5.496274614334107,
      "epoch": 1.13994496546302,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048811288536222794,
      "loss": 5.4362,
      "mean_token_accuracy": 0.1737013503909111,
      "num_tokens": 22637877.0,
      "step": 10150
    },
    {
      "entropy": 5.640788221359253,
      "epoch": 1.1405065423709777,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048810000479995,
      "loss": 5.4677,
      "mean_token_accuracy": 0.17672027200460433,
      "num_tokens": 22648770.0,
      "step": 10155
    },
    {
      "entropy": 5.753425788879395,
      "epoch": 1.1410681192789354,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048808711745248524,
      "loss": 5.5346,
      "mean_token_accuracy": 0.17500863820314408,
      "num_tokens": 22659626.0,
      "step": 10160
    },
    {
      "entropy": 5.588730239868164,
      "epoch": 1.1416296961868928,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048807422332024404,
      "loss": 5.5057,
      "mean_token_accuracy": 0.17713336348533631,
      "num_tokens": 22669035.0,
      "step": 10165
    },
    {
      "entropy": 5.668708276748657,
      "epoch": 1.1421912730948502,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004880613224036369,
      "loss": 5.4893,
      "mean_token_accuracy": 0.18129380345344542,
      "num_tokens": 22680536.0,
      "step": 10170
    },
    {
      "entropy": 5.655089092254639,
      "epoch": 1.142752850002808,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048804841470307466,
      "loss": 5.4386,
      "mean_token_accuracy": 0.1854078873991966,
      "num_tokens": 22692732.0,
      "step": 10175
    },
    {
      "entropy": 5.620869159698486,
      "epoch": 1.1433144269107653,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048803550021896824,
      "loss": 5.563,
      "mean_token_accuracy": 0.17335400730371475,
      "num_tokens": 22703677.0,
      "step": 10180
    },
    {
      "entropy": 5.660672235488891,
      "epoch": 1.143876003818723,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004880225789517289,
      "loss": 5.4861,
      "mean_token_accuracy": 0.17455905973911284,
      "num_tokens": 22714826.0,
      "step": 10185
    },
    {
      "entropy": 5.6227153778076175,
      "epoch": 1.1444375807266804,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000488009650901768,
      "loss": 5.5619,
      "mean_token_accuracy": 0.17745715528726577,
      "num_tokens": 22726013.0,
      "step": 10190
    },
    {
      "entropy": 5.58302526473999,
      "epoch": 1.144999157634638,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048799671606949724,
      "loss": 5.3963,
      "mean_token_accuracy": 0.17846509218215942,
      "num_tokens": 22737494.0,
      "step": 10195
    },
    {
      "entropy": 5.5372984409332275,
      "epoch": 1.1455607345425955,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048798377445532837,
      "loss": 5.3188,
      "mean_token_accuracy": 0.18994600474834442,
      "num_tokens": 22746916.0,
      "step": 10200
    },
    {
      "entropy": 5.630187463760376,
      "epoch": 1.1461223114505532,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004879708260596736,
      "loss": 5.5628,
      "mean_token_accuracy": 0.16555242538452147,
      "num_tokens": 22759197.0,
      "step": 10205
    },
    {
      "entropy": 5.631412601470947,
      "epoch": 1.1466838883585107,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004879578708829451,
      "loss": 5.5195,
      "mean_token_accuracy": 0.17430828064680098,
      "num_tokens": 22770744.0,
      "step": 10210
    },
    {
      "entropy": 5.604301023483276,
      "epoch": 1.1472454652664683,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004879449089255553,
      "loss": 5.446,
      "mean_token_accuracy": 0.17849487960338592,
      "num_tokens": 22781989.0,
      "step": 10215
    },
    {
      "entropy": 5.541107559204102,
      "epoch": 1.1478070421744258,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004879319401879171,
      "loss": 5.3642,
      "mean_token_accuracy": 0.18781604170799254,
      "num_tokens": 22792972.0,
      "step": 10220
    },
    {
      "entropy": 5.562822389602661,
      "epoch": 1.1483686190823834,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048791896467044335,
      "loss": 5.436,
      "mean_token_accuracy": 0.18912334442138673,
      "num_tokens": 22802623.0,
      "step": 10225
    },
    {
      "entropy": 5.459062576293945,
      "epoch": 1.1489301959903409,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048790598237354716,
      "loss": 5.382,
      "mean_token_accuracy": 0.18711169362068175,
      "num_tokens": 22813564.0,
      "step": 10230
    },
    {
      "entropy": 5.596525096893311,
      "epoch": 1.1494917728982985,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004878929932976419,
      "loss": 5.5297,
      "mean_token_accuracy": 0.17973680347204207,
      "num_tokens": 22824681.0,
      "step": 10235
    },
    {
      "entropy": 5.55367226600647,
      "epoch": 1.150053349806256,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004878799974431412,
      "loss": 5.3573,
      "mean_token_accuracy": 0.19110948592424393,
      "num_tokens": 22836030.0,
      "step": 10240
    },
    {
      "entropy": 5.58520393371582,
      "epoch": 1.1506149267142134,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004878669948104588,
      "loss": 5.3781,
      "mean_token_accuracy": 0.1786734029650688,
      "num_tokens": 22847213.0,
      "step": 10245
    },
    {
      "entropy": 5.638460493087768,
      "epoch": 1.151176503622171,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048785398540000867,
      "loss": 5.3846,
      "mean_token_accuracy": 0.18784808963537217,
      "num_tokens": 22857904.0,
      "step": 10250
    },
    {
      "entropy": 5.559061765670776,
      "epoch": 1.1517380805301287,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048784096921220514,
      "loss": 5.526,
      "mean_token_accuracy": 0.17414491176605223,
      "num_tokens": 22868975.0,
      "step": 10255
    },
    {
      "entropy": 5.657341861724854,
      "epoch": 1.1522996574380862,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004878279462474627,
      "loss": 5.4727,
      "mean_token_accuracy": 0.1808004006743431,
      "num_tokens": 22879463.0,
      "step": 10260
    },
    {
      "entropy": 5.637919902801514,
      "epoch": 1.1528612343460436,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048781491650619574,
      "loss": 5.4019,
      "mean_token_accuracy": 0.18174820393323898,
      "num_tokens": 22890382.0,
      "step": 10265
    },
    {
      "entropy": 5.621941804885864,
      "epoch": 1.1534228112540013,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048780187998881944,
      "loss": 5.4951,
      "mean_token_accuracy": 0.1807561531662941,
      "num_tokens": 22901968.0,
      "step": 10270
    },
    {
      "entropy": 5.671910381317138,
      "epoch": 1.1539843881619587,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004877888366957486,
      "loss": 5.5287,
      "mean_token_accuracy": 0.17953866124153137,
      "num_tokens": 22912147.0,
      "step": 10275
    },
    {
      "entropy": 5.588748836517334,
      "epoch": 1.1545459650699164,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004877757866273988,
      "loss": 5.4154,
      "mean_token_accuracy": 0.1782118409872055,
      "num_tokens": 22923744.0,
      "step": 10280
    },
    {
      "entropy": 5.587154865264893,
      "epoch": 1.1551075419778738,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004877627297841853,
      "loss": 5.385,
      "mean_token_accuracy": 0.1853589504957199,
      "num_tokens": 22935487.0,
      "step": 10285
    },
    {
      "entropy": 5.589299440383911,
      "epoch": 1.1556691188858315,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004877496661665241,
      "loss": 5.4237,
      "mean_token_accuracy": 0.18607935905456544,
      "num_tokens": 22946379.0,
      "step": 10290
    },
    {
      "entropy": 5.605984210968018,
      "epoch": 1.156230695793789,
      "grad_norm": 1.75,
      "learning_rate": 0.000487736595774831,
      "loss": 5.4308,
      "mean_token_accuracy": 0.17865341305732726,
      "num_tokens": 22957219.0,
      "step": 10295
    },
    {
      "entropy": 5.564409446716309,
      "epoch": 1.1567922727017466,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048772351860952215,
      "loss": 5.3854,
      "mean_token_accuracy": 0.18352195173501967,
      "num_tokens": 22967749.0,
      "step": 10300
    },
    {
      "entropy": 5.553499841690064,
      "epoch": 1.157353849609704,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004877104346710139,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18055881708860397,
      "num_tokens": 22978317.0,
      "step": 10305
    },
    {
      "entropy": 5.5471031665802,
      "epoch": 1.1579154265176617,
      "grad_norm": 1.6875,
      "learning_rate": 0.000487697343959723,
      "loss": 5.4296,
      "mean_token_accuracy": 0.17836801707744598,
      "num_tokens": 22988793.0,
      "step": 10310
    },
    {
      "entropy": 5.6121556758880615,
      "epoch": 1.158477003425619,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004876842464760661,
      "loss": 5.4918,
      "mean_token_accuracy": 0.17678036242723466,
      "num_tokens": 23000073.0,
      "step": 10315
    },
    {
      "entropy": 5.615738248825073,
      "epoch": 1.1590385803335768,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048767114222046036,
      "loss": 5.4653,
      "mean_token_accuracy": 0.17593652456998826,
      "num_tokens": 23011260.0,
      "step": 10320
    },
    {
      "entropy": 5.559394311904907,
      "epoch": 1.1596001572415342,
      "grad_norm": 1.640625,
      "learning_rate": 0.000487658031193323,
      "loss": 5.4388,
      "mean_token_accuracy": 0.1824623391032219,
      "num_tokens": 23022129.0,
      "step": 10325
    },
    {
      "entropy": 5.504612159729004,
      "epoch": 1.1601617341494919,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004876449133950714,
      "loss": 5.3838,
      "mean_token_accuracy": 0.17638918459415437,
      "num_tokens": 23033666.0,
      "step": 10330
    },
    {
      "entropy": 5.549623918533325,
      "epoch": 1.1607233110574493,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004876317888261233,
      "loss": 5.4085,
      "mean_token_accuracy": 0.18171306997537612,
      "num_tokens": 23044133.0,
      "step": 10335
    },
    {
      "entropy": 5.570271778106689,
      "epoch": 1.1612848879654067,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00048761865748689657,
      "loss": 5.4217,
      "mean_token_accuracy": 0.17687359154224397,
      "num_tokens": 23055877.0,
      "step": 10340
    },
    {
      "entropy": 5.649858665466309,
      "epoch": 1.1618464648733644,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048760551937780927,
      "loss": 5.5343,
      "mean_token_accuracy": 0.1781652018427849,
      "num_tokens": 23066811.0,
      "step": 10345
    },
    {
      "entropy": 5.694931364059448,
      "epoch": 1.162408041781322,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048759237449927986,
      "loss": 5.4794,
      "mean_token_accuracy": 0.17494654953479766,
      "num_tokens": 23078928.0,
      "step": 10350
    },
    {
      "entropy": 5.621661186218262,
      "epoch": 1.1629696186892795,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048757922285172674,
      "loss": 5.4635,
      "mean_token_accuracy": 0.1786074697971344,
      "num_tokens": 23090685.0,
      "step": 10355
    },
    {
      "entropy": 5.539975881576538,
      "epoch": 1.163531195597237,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048756606443556874,
      "loss": 5.3683,
      "mean_token_accuracy": 0.1829329326748848,
      "num_tokens": 23101659.0,
      "step": 10360
    },
    {
      "entropy": 5.580107975006103,
      "epoch": 1.1640927725051946,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048755289925122474,
      "loss": 5.4497,
      "mean_token_accuracy": 0.18081278949975968,
      "num_tokens": 23112970.0,
      "step": 10365
    },
    {
      "entropy": 5.639863586425781,
      "epoch": 1.164654349413152,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004875397272991141,
      "loss": 5.4728,
      "mean_token_accuracy": 0.1762469530105591,
      "num_tokens": 23125391.0,
      "step": 10370
    },
    {
      "entropy": 5.6568304061889645,
      "epoch": 1.1652159263211097,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000487526548579656,
      "loss": 5.5143,
      "mean_token_accuracy": 0.17894838154315948,
      "num_tokens": 23138195.0,
      "step": 10375
    },
    {
      "entropy": 5.670497846603394,
      "epoch": 1.1657775032290671,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048751336309327026,
      "loss": 5.5022,
      "mean_token_accuracy": 0.17685108482837678,
      "num_tokens": 23149378.0,
      "step": 10380
    },
    {
      "entropy": 5.484286737442017,
      "epoch": 1.1663390801370248,
      "grad_norm": 1.5,
      "learning_rate": 0.0004875001708403766,
      "loss": 5.3319,
      "mean_token_accuracy": 0.18941258490085602,
      "num_tokens": 23159394.0,
      "step": 10385
    },
    {
      "entropy": 5.664628553390503,
      "epoch": 1.1669006570449822,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048748697182139503,
      "loss": 5.5844,
      "mean_token_accuracy": 0.17606203109025956,
      "num_tokens": 23171512.0,
      "step": 10390
    },
    {
      "entropy": 5.568928146362305,
      "epoch": 1.16746223395294,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004874737660367459,
      "loss": 5.4147,
      "mean_token_accuracy": 0.18418188095092775,
      "num_tokens": 23181635.0,
      "step": 10395
    },
    {
      "entropy": 5.715925979614258,
      "epoch": 1.1680238108608973,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048746055348684967,
      "loss": 5.6897,
      "mean_token_accuracy": 0.16924521178007126,
      "num_tokens": 23194510.0,
      "step": 10400
    },
    {
      "entropy": 5.7471662044525145,
      "epoch": 1.168585387768855,
      "grad_norm": 1.6875,
      "learning_rate": 0.000487447334172127,
      "loss": 5.5551,
      "mean_token_accuracy": 0.18067686855793,
      "num_tokens": 23205146.0,
      "step": 10405
    },
    {
      "entropy": 5.60310001373291,
      "epoch": 1.1691469646768125,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048743410809299877,
      "loss": 5.463,
      "mean_token_accuracy": 0.17559536695480346,
      "num_tokens": 23216621.0,
      "step": 10410
    },
    {
      "entropy": 5.574869012832641,
      "epoch": 1.1697085415847701,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004874208752498862,
      "loss": 5.4021,
      "mean_token_accuracy": 0.1809071496129036,
      "num_tokens": 23226401.0,
      "step": 10415
    },
    {
      "entropy": 5.549800872802734,
      "epoch": 1.1702701184927276,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004874076356432106,
      "loss": 5.4653,
      "mean_token_accuracy": 0.17664449661970139,
      "num_tokens": 23237921.0,
      "step": 10420
    },
    {
      "entropy": 5.608479022979736,
      "epoch": 1.1708316954006852,
      "grad_norm": 1.5,
      "learning_rate": 0.00048739438927339346,
      "loss": 5.4292,
      "mean_token_accuracy": 0.18261185586452483,
      "num_tokens": 23250015.0,
      "step": 10425
    },
    {
      "entropy": 5.578491735458374,
      "epoch": 1.1713932723086427,
      "grad_norm": 2.1875,
      "learning_rate": 0.00048738113614085656,
      "loss": 5.4575,
      "mean_token_accuracy": 0.17823632657527924,
      "num_tokens": 23261243.0,
      "step": 10430
    },
    {
      "entropy": 5.68017430305481,
      "epoch": 1.1719548492166,
      "grad_norm": 1.640625,
      "learning_rate": 0.000487367876246022,
      "loss": 5.6097,
      "mean_token_accuracy": 0.17084814310073854,
      "num_tokens": 23273264.0,
      "step": 10435
    },
    {
      "entropy": 5.8019805431365965,
      "epoch": 1.1725164261245578,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004873546095893118,
      "loss": 5.6237,
      "mean_token_accuracy": 0.17058354765176773,
      "num_tokens": 23285031.0,
      "step": 10440
    },
    {
      "entropy": 5.589553451538086,
      "epoch": 1.1730780030325154,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004873413361711485,
      "loss": 5.3735,
      "mean_token_accuracy": 0.18483428210020064,
      "num_tokens": 23296387.0,
      "step": 10445
    },
    {
      "entropy": 5.635595417022705,
      "epoch": 1.1736395799404729,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048732805599195465,
      "loss": 5.493,
      "mean_token_accuracy": 0.18177550584077834,
      "num_tokens": 23307950.0,
      "step": 10450
    },
    {
      "entropy": 5.472813606262207,
      "epoch": 1.1742011568484303,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048731476905215326,
      "loss": 5.3059,
      "mean_token_accuracy": 0.18500818759202958,
      "num_tokens": 23318424.0,
      "step": 10455
    },
    {
      "entropy": 5.584354305267334,
      "epoch": 1.174762733756388,
      "grad_norm": 1.625,
      "learning_rate": 0.0004873014753521673,
      "loss": 5.5634,
      "mean_token_accuracy": 0.17263590097427367,
      "num_tokens": 23330186.0,
      "step": 10460
    },
    {
      "entropy": 5.649318361282349,
      "epoch": 1.1753243106643454,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004872881748924199,
      "loss": 5.3496,
      "mean_token_accuracy": 0.18661798983812333,
      "num_tokens": 23341878.0,
      "step": 10465
    },
    {
      "entropy": 5.615511178970337,
      "epoch": 1.175885887572303,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004872748676733347,
      "loss": 5.4777,
      "mean_token_accuracy": 0.17548347413539886,
      "num_tokens": 23352471.0,
      "step": 10470
    },
    {
      "entropy": 5.621794986724853,
      "epoch": 1.1764474644802605,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048726155369533547,
      "loss": 5.4891,
      "mean_token_accuracy": 0.18269820213317872,
      "num_tokens": 23363871.0,
      "step": 10475
    },
    {
      "entropy": 5.656167221069336,
      "epoch": 1.1770090413882182,
      "grad_norm": 1.7890625,
      "learning_rate": 0.000487248232958846,
      "loss": 5.5134,
      "mean_token_accuracy": 0.17936725318431854,
      "num_tokens": 23376072.0,
      "step": 10480
    },
    {
      "entropy": 5.487087917327881,
      "epoch": 1.1775706182961756,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004872349054642904,
      "loss": 5.3773,
      "mean_token_accuracy": 0.18486556112766267,
      "num_tokens": 23386670.0,
      "step": 10485
    },
    {
      "entropy": 5.646738195419312,
      "epoch": 1.1781321952041333,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004872215712120932,
      "loss": 5.4282,
      "mean_token_accuracy": 0.1851399466395378,
      "num_tokens": 23396499.0,
      "step": 10490
    },
    {
      "entropy": 5.627827501296997,
      "epoch": 1.1786937721120907,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004872082302026788,
      "loss": 5.4626,
      "mean_token_accuracy": 0.1846533477306366,
      "num_tokens": 23407898.0,
      "step": 10495
    },
    {
      "entropy": 5.552610301971436,
      "epoch": 1.1792553490200484,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004871948824364721,
      "loss": 5.4426,
      "mean_token_accuracy": 0.1783500924706459,
      "num_tokens": 23418877.0,
      "step": 10500
    },
    {
      "entropy": 5.524981164932251,
      "epoch": 1.1798169259280058,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000487181527913898,
      "loss": 5.4091,
      "mean_token_accuracy": 0.1811755284667015,
      "num_tokens": 23431378.0,
      "step": 10505
    },
    {
      "entropy": 5.616604232788086,
      "epoch": 1.1803785028359635,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048716816663538177,
      "loss": 5.4221,
      "mean_token_accuracy": 0.18377246409654618,
      "num_tokens": 23442590.0,
      "step": 10510
    },
    {
      "entropy": 5.645650339126587,
      "epoch": 1.180940079743921,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004871547986013488,
      "loss": 5.5129,
      "mean_token_accuracy": 0.176759235560894,
      "num_tokens": 23454139.0,
      "step": 10515
    },
    {
      "entropy": 5.634320497512817,
      "epoch": 1.1815016566518786,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004871414238122248,
      "loss": 5.4535,
      "mean_token_accuracy": 0.18437889814376832,
      "num_tokens": 23464908.0,
      "step": 10520
    },
    {
      "entropy": 5.548039531707763,
      "epoch": 1.182063233559836,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004871280422684355,
      "loss": 5.4039,
      "mean_token_accuracy": 0.18599554449319838,
      "num_tokens": 23475436.0,
      "step": 10525
    },
    {
      "entropy": 5.6510828018188475,
      "epoch": 1.1826248104677934,
      "grad_norm": 2.71875,
      "learning_rate": 0.00048711465397040726,
      "loss": 5.5688,
      "mean_token_accuracy": 0.17750822007656097,
      "num_tokens": 23487384.0,
      "step": 10530
    },
    {
      "entropy": 5.718241024017334,
      "epoch": 1.183186387375751,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048710125891856604,
      "loss": 5.5207,
      "mean_token_accuracy": 0.17472441494464874,
      "num_tokens": 23498978.0,
      "step": 10535
    },
    {
      "entropy": 5.529882097244263,
      "epoch": 1.1837479642837088,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048708785711333845,
      "loss": 5.3981,
      "mean_token_accuracy": 0.18602953851222992,
      "num_tokens": 23510825.0,
      "step": 10540
    },
    {
      "entropy": 5.53178858757019,
      "epoch": 1.1843095411916662,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048707444855515124,
      "loss": 5.3904,
      "mean_token_accuracy": 0.18575898855924605,
      "num_tokens": 23522383.0,
      "step": 10545
    },
    {
      "entropy": 5.6463545799255375,
      "epoch": 1.1848711180996236,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004870610332444314,
      "loss": 5.5725,
      "mean_token_accuracy": 0.17957879304885865,
      "num_tokens": 23534255.0,
      "step": 10550
    },
    {
      "entropy": 5.630608129501343,
      "epoch": 1.1854326950075813,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004870476111816059,
      "loss": 5.4772,
      "mean_token_accuracy": 0.17648603469133378,
      "num_tokens": 23545888.0,
      "step": 10555
    },
    {
      "entropy": 5.59116530418396,
      "epoch": 1.1859942719155387,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048703418236710234,
      "loss": 5.479,
      "mean_token_accuracy": 0.17763277441263198,
      "num_tokens": 23556776.0,
      "step": 10560
    },
    {
      "entropy": 5.703448057174683,
      "epoch": 1.1865558488234964,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004870207468013481,
      "loss": 5.4081,
      "mean_token_accuracy": 0.18237634003162384,
      "num_tokens": 23569971.0,
      "step": 10565
    },
    {
      "entropy": 5.602100801467896,
      "epoch": 1.1871174257314538,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004870073044847711,
      "loss": 5.5371,
      "mean_token_accuracy": 0.1801938682794571,
      "num_tokens": 23580751.0,
      "step": 10570
    },
    {
      "entropy": 5.501537895202636,
      "epoch": 1.1876790026394115,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004869938554177993,
      "loss": 5.3992,
      "mean_token_accuracy": 0.18619302660226822,
      "num_tokens": 23591423.0,
      "step": 10575
    },
    {
      "entropy": 5.630936050415039,
      "epoch": 1.188240579547369,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048698039960086084,
      "loss": 5.3665,
      "mean_token_accuracy": 0.17726145088672637,
      "num_tokens": 23601552.0,
      "step": 10580
    },
    {
      "entropy": 5.55292820930481,
      "epoch": 1.1888021564553266,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004869669370343843,
      "loss": 5.466,
      "mean_token_accuracy": 0.18769415318965912,
      "num_tokens": 23613080.0,
      "step": 10585
    },
    {
      "entropy": 5.550881624221802,
      "epoch": 1.189363733363284,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004869534677187983,
      "loss": 5.3534,
      "mean_token_accuracy": 0.18469875305891037,
      "num_tokens": 23623559.0,
      "step": 10590
    },
    {
      "entropy": 5.566106939315796,
      "epoch": 1.1899253102712417,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004869399916545317,
      "loss": 5.3683,
      "mean_token_accuracy": 0.18476739078760146,
      "num_tokens": 23634113.0,
      "step": 10595
    },
    {
      "entropy": 5.513452339172363,
      "epoch": 1.1904868871791991,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004869265088420135,
      "loss": 5.2838,
      "mean_token_accuracy": 0.19047885537147521,
      "num_tokens": 23645203.0,
      "step": 10600
    },
    {
      "entropy": 5.523169946670532,
      "epoch": 1.1910484640871568,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004869130192816731,
      "loss": 5.4231,
      "mean_token_accuracy": 0.18399088382720946,
      "num_tokens": 23654880.0,
      "step": 10605
    },
    {
      "entropy": 5.537129068374634,
      "epoch": 1.1916100409951143,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00048689952297394,
      "loss": 5.3309,
      "mean_token_accuracy": 0.1929388478398323,
      "num_tokens": 23664962.0,
      "step": 10610
    },
    {
      "entropy": 5.602372550964356,
      "epoch": 1.192171617903072,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004868860199192438,
      "loss": 5.3893,
      "mean_token_accuracy": 0.19017872512340545,
      "num_tokens": 23675288.0,
      "step": 10615
    },
    {
      "entropy": 5.52705454826355,
      "epoch": 1.1927331948110294,
      "grad_norm": 2.0,
      "learning_rate": 0.0004868725101180146,
      "loss": 5.3964,
      "mean_token_accuracy": 0.18070517629384994,
      "num_tokens": 23685948.0,
      "step": 10620
    },
    {
      "entropy": 5.641992568969727,
      "epoch": 1.1932947717189868,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004868589935706826,
      "loss": 5.4599,
      "mean_token_accuracy": 0.17490803301334382,
      "num_tokens": 23696980.0,
      "step": 10625
    },
    {
      "entropy": 5.663866281509399,
      "epoch": 1.1938563486269445,
      "grad_norm": 1.84375,
      "learning_rate": 0.00048684547027767793,
      "loss": 5.4986,
      "mean_token_accuracy": 0.17656034231185913,
      "num_tokens": 23708644.0,
      "step": 10630
    },
    {
      "entropy": 5.59591007232666,
      "epoch": 1.1944179255349021,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004868319402394314,
      "loss": 5.4738,
      "mean_token_accuracy": 0.18437945246696472,
      "num_tokens": 23719652.0,
      "step": 10635
    },
    {
      "entropy": 5.6034361839294435,
      "epoch": 1.1949795024428596,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048681840345637375,
      "loss": 5.4253,
      "mean_token_accuracy": 0.18002281337976456,
      "num_tokens": 23731906.0,
      "step": 10640
    },
    {
      "entropy": 5.614962434768676,
      "epoch": 1.195541079350817,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048680485992893597,
      "loss": 5.475,
      "mean_token_accuracy": 0.18151820302009583,
      "num_tokens": 23744184.0,
      "step": 10645
    },
    {
      "entropy": 5.541026449203491,
      "epoch": 1.1961026562587747,
      "grad_norm": 1.890625,
      "learning_rate": 0.00048679130965754926,
      "loss": 5.298,
      "mean_token_accuracy": 0.1837436631321907,
      "num_tokens": 23754807.0,
      "step": 10650
    },
    {
      "entropy": 5.514839363098145,
      "epoch": 1.196664233166732,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004867777526426451,
      "loss": 5.3977,
      "mean_token_accuracy": 0.1806316688656807,
      "num_tokens": 23765907.0,
      "step": 10655
    },
    {
      "entropy": 5.6018781661987305,
      "epoch": 1.1972258100746898,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048676418888465525,
      "loss": 5.4269,
      "mean_token_accuracy": 0.1833296984434128,
      "num_tokens": 23776217.0,
      "step": 10660
    },
    {
      "entropy": 5.576879549026489,
      "epoch": 1.1977873869826472,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004867506183840114,
      "loss": 5.3802,
      "mean_token_accuracy": 0.18599339574575424,
      "num_tokens": 23786170.0,
      "step": 10665
    },
    {
      "entropy": 5.650488471984863,
      "epoch": 1.1983489638906049,
      "grad_norm": 1.875,
      "learning_rate": 0.0004867370411411459,
      "loss": 5.5168,
      "mean_token_accuracy": 0.1732734113931656,
      "num_tokens": 23796491.0,
      "step": 10670
    },
    {
      "entropy": 5.564750051498413,
      "epoch": 1.1989105407985623,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048672345715649066,
      "loss": 5.3899,
      "mean_token_accuracy": 0.1827486753463745,
      "num_tokens": 23807718.0,
      "step": 10675
    },
    {
      "entropy": 5.5354297161102295,
      "epoch": 1.19947211770652,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004867098664304785,
      "loss": 5.5055,
      "mean_token_accuracy": 0.1804208055138588,
      "num_tokens": 23818623.0,
      "step": 10680
    },
    {
      "entropy": 5.555293941497803,
      "epoch": 1.2000336946144774,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004866962689635422,
      "loss": 5.425,
      "mean_token_accuracy": 0.1853599354624748,
      "num_tokens": 23829173.0,
      "step": 10685
    },
    {
      "entropy": 5.594825649261475,
      "epoch": 1.200595271522435,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048668266475611447,
      "loss": 5.3517,
      "mean_token_accuracy": 0.1820319652557373,
      "num_tokens": 23840689.0,
      "step": 10690
    },
    {
      "entropy": 5.675231122970581,
      "epoch": 1.2011568484303925,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004866690538086286,
      "loss": 5.5828,
      "mean_token_accuracy": 0.17370163053274154,
      "num_tokens": 23852858.0,
      "step": 10695
    },
    {
      "entropy": 5.607054805755615,
      "epoch": 1.2017184253383502,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048665543612151803,
      "loss": 5.3814,
      "mean_token_accuracy": 0.1894412964582443,
      "num_tokens": 23864137.0,
      "step": 10700
    },
    {
      "entropy": 5.575736999511719,
      "epoch": 1.2022800022463076,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048664181169521625,
      "loss": 5.4326,
      "mean_token_accuracy": 0.18311810344457627,
      "num_tokens": 23875087.0,
      "step": 10705
    },
    {
      "entropy": 5.549623203277588,
      "epoch": 1.2028415791542653,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004866281805301571,
      "loss": 5.3219,
      "mean_token_accuracy": 0.1916783109307289,
      "num_tokens": 23886521.0,
      "step": 10710
    },
    {
      "entropy": 5.567817354202271,
      "epoch": 1.2034031560622227,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004866145426267746,
      "loss": 5.4351,
      "mean_token_accuracy": 0.17877652496099472,
      "num_tokens": 23897660.0,
      "step": 10715
    },
    {
      "entropy": 5.569176912307739,
      "epoch": 1.2039647329701804,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000486600897985503,
      "loss": 5.4775,
      "mean_token_accuracy": 0.18148595541715623,
      "num_tokens": 23910543.0,
      "step": 10720
    },
    {
      "entropy": 5.651079368591309,
      "epoch": 1.2045263098781378,
      "grad_norm": 2.34375,
      "learning_rate": 0.00048658724660677674,
      "loss": 5.3545,
      "mean_token_accuracy": 0.18516494184732438,
      "num_tokens": 23921687.0,
      "step": 10725
    },
    {
      "entropy": 5.644492149353027,
      "epoch": 1.2050878867860955,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004865735884910305,
      "loss": 5.4527,
      "mean_token_accuracy": 0.17805104255676268,
      "num_tokens": 23933717.0,
      "step": 10730
    },
    {
      "entropy": 5.565663385391235,
      "epoch": 1.205649463694053,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004865599236386991,
      "loss": 5.5131,
      "mean_token_accuracy": 0.17860305160284043,
      "num_tokens": 23947764.0,
      "step": 10735
    },
    {
      "entropy": 5.548420000076294,
      "epoch": 1.2062110406020103,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004865462520502177,
      "loss": 5.3857,
      "mean_token_accuracy": 0.17887697070837022,
      "num_tokens": 23960096.0,
      "step": 10740
    },
    {
      "entropy": 5.641502857208252,
      "epoch": 1.206772617509968,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004865325737260216,
      "loss": 5.4026,
      "mean_token_accuracy": 0.17797549962997436,
      "num_tokens": 23971573.0,
      "step": 10745
    },
    {
      "entropy": 5.594951152801514,
      "epoch": 1.2073341944179254,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048651888866654624,
      "loss": 5.3613,
      "mean_token_accuracy": 0.19261209070682525,
      "num_tokens": 23982710.0,
      "step": 10750
    },
    {
      "entropy": 5.460863971710205,
      "epoch": 1.207895771325883,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004865051968722275,
      "loss": 5.3161,
      "mean_token_accuracy": 0.1874060869216919,
      "num_tokens": 23993663.0,
      "step": 10755
    },
    {
      "entropy": 5.5636594772338865,
      "epoch": 1.2084573482338405,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048649149834350126,
      "loss": 5.561,
      "mean_token_accuracy": 0.1814848840236664,
      "num_tokens": 24005112.0,
      "step": 10760
    },
    {
      "entropy": 5.523722505569458,
      "epoch": 1.2090189251417982,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004864777930808037,
      "loss": 5.3079,
      "mean_token_accuracy": 0.1889803558588028,
      "num_tokens": 24015855.0,
      "step": 10765
    },
    {
      "entropy": 5.533093738555908,
      "epoch": 1.2095805020497556,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004864640810845711,
      "loss": 5.3611,
      "mean_token_accuracy": 0.18602087199687958,
      "num_tokens": 24027232.0,
      "step": 10770
    },
    {
      "entropy": 5.5402055263519285,
      "epoch": 1.2101420789577133,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004864503623552402,
      "loss": 5.3606,
      "mean_token_accuracy": 0.1877458542585373,
      "num_tokens": 24039127.0,
      "step": 10775
    },
    {
      "entropy": 5.627122831344605,
      "epoch": 1.2107036558656707,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004864366368932478,
      "loss": 5.4973,
      "mean_token_accuracy": 0.1808575674891472,
      "num_tokens": 24049821.0,
      "step": 10780
    },
    {
      "entropy": 5.597670602798462,
      "epoch": 1.2112652327736284,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048642290469903083,
      "loss": 5.5212,
      "mean_token_accuracy": 0.17994545847177507,
      "num_tokens": 24060305.0,
      "step": 10785
    },
    {
      "entropy": 5.579167890548706,
      "epoch": 1.2118268096815858,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00048640916577302664,
      "loss": 5.3939,
      "mean_token_accuracy": 0.18764331489801406,
      "num_tokens": 24071193.0,
      "step": 10790
    },
    {
      "entropy": 5.47999062538147,
      "epoch": 1.2123883865895435,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048639542011567243,
      "loss": 5.2623,
      "mean_token_accuracy": 0.2041529268026352,
      "num_tokens": 24082380.0,
      "step": 10795
    },
    {
      "entropy": 5.653561067581177,
      "epoch": 1.212949963497501,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004863816677274062,
      "loss": 5.5384,
      "mean_token_accuracy": 0.179527847468853,
      "num_tokens": 24093168.0,
      "step": 10800
    },
    {
      "entropy": 5.539981079101563,
      "epoch": 1.2135115404054586,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048636790860866576,
      "loss": 5.3951,
      "mean_token_accuracy": 0.18281326144933702,
      "num_tokens": 24105355.0,
      "step": 10805
    },
    {
      "entropy": 5.488740491867065,
      "epoch": 1.214073117313416,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004863541427598889,
      "loss": 5.3561,
      "mean_token_accuracy": 0.18512464463710784,
      "num_tokens": 24117062.0,
      "step": 10810
    },
    {
      "entropy": 5.651952171325684,
      "epoch": 1.2146346942213737,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004863403701815144,
      "loss": 5.5061,
      "mean_token_accuracy": 0.17755191177129745,
      "num_tokens": 24127770.0,
      "step": 10815
    },
    {
      "entropy": 5.588659286499023,
      "epoch": 1.2151962711293312,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004863265908739805,
      "loss": 5.582,
      "mean_token_accuracy": 0.17055042684078217,
      "num_tokens": 24139894.0,
      "step": 10820
    },
    {
      "entropy": 5.603845548629761,
      "epoch": 1.2157578480372888,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004863128048377259,
      "loss": 5.3819,
      "mean_token_accuracy": 0.19205069839954375,
      "num_tokens": 24150799.0,
      "step": 10825
    },
    {
      "entropy": 5.6096703052520756,
      "epoch": 1.2163194249452463,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004862990120731896,
      "loss": 5.4246,
      "mean_token_accuracy": 0.18172770291566848,
      "num_tokens": 24161406.0,
      "step": 10830
    },
    {
      "entropy": 5.434682083129883,
      "epoch": 1.2168810018532037,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048628521258081096,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18648598492145538,
      "num_tokens": 24171884.0,
      "step": 10835
    },
    {
      "entropy": 5.543994188308716,
      "epoch": 1.2174425787611614,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048627140636102906,
      "loss": 5.3909,
      "mean_token_accuracy": 0.18740372359752655,
      "num_tokens": 24183873.0,
      "step": 10840
    },
    {
      "entropy": 5.662725353240967,
      "epoch": 1.2180041556691188,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004862575934142837,
      "loss": 5.4401,
      "mean_token_accuracy": 0.18730383813381196,
      "num_tokens": 24194689.0,
      "step": 10845
    },
    {
      "entropy": 5.545740222930908,
      "epoch": 1.2185657325770765,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004862437737410147,
      "loss": 5.3497,
      "mean_token_accuracy": 0.18136660158634185,
      "num_tokens": 24206675.0,
      "step": 10850
    },
    {
      "entropy": 5.513843154907226,
      "epoch": 1.219127309485034,
      "grad_norm": 2.921875,
      "learning_rate": 0.0004862299473416619,
      "loss": 5.4377,
      "mean_token_accuracy": 0.17925103157758712,
      "num_tokens": 24218845.0,
      "step": 10855
    },
    {
      "entropy": 5.570048236846924,
      "epoch": 1.2196888863929916,
      "grad_norm": 1.84375,
      "learning_rate": 0.00048621611421666573,
      "loss": 5.4684,
      "mean_token_accuracy": 0.17514824569225312,
      "num_tokens": 24229889.0,
      "step": 10860
    },
    {
      "entropy": 5.662149095535279,
      "epoch": 1.220250463300949,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004862022743664665,
      "loss": 5.5148,
      "mean_token_accuracy": 0.1838795378804207,
      "num_tokens": 24240729.0,
      "step": 10865
    },
    {
      "entropy": 5.6571629524230955,
      "epoch": 1.2208120402089067,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048618842779150494,
      "loss": 5.5292,
      "mean_token_accuracy": 0.17597716152668,
      "num_tokens": 24251765.0,
      "step": 10870
    },
    {
      "entropy": 5.558348178863525,
      "epoch": 1.221373617116864,
      "grad_norm": 1.625,
      "learning_rate": 0.00048617457449222205,
      "loss": 5.4365,
      "mean_token_accuracy": 0.1781846523284912,
      "num_tokens": 24262273.0,
      "step": 10875
    },
    {
      "entropy": 5.612859439849854,
      "epoch": 1.2219351940248218,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004861607144690587,
      "loss": 5.4272,
      "mean_token_accuracy": 0.17829763144254684,
      "num_tokens": 24272313.0,
      "step": 10880
    },
    {
      "entropy": 5.64711561203003,
      "epoch": 1.2224967709327792,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004861468477224563,
      "loss": 5.4948,
      "mean_token_accuracy": 0.1757294699549675,
      "num_tokens": 24283971.0,
      "step": 10885
    },
    {
      "entropy": 5.55884313583374,
      "epoch": 1.2230583478407369,
      "grad_norm": 1.5,
      "learning_rate": 0.00048613297425285635,
      "loss": 5.4284,
      "mean_token_accuracy": 0.18071170300245284,
      "num_tokens": 24294357.0,
      "step": 10890
    },
    {
      "entropy": 5.53272123336792,
      "epoch": 1.2236199247486943,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004861190940607007,
      "loss": 5.2968,
      "mean_token_accuracy": 0.1882367879152298,
      "num_tokens": 24305894.0,
      "step": 10895
    },
    {
      "entropy": 5.5826810836792,
      "epoch": 1.224181501656652,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004861052071464312,
      "loss": 5.4099,
      "mean_token_accuracy": 0.1851435512304306,
      "num_tokens": 24318506.0,
      "step": 10900
    },
    {
      "entropy": 5.564007949829102,
      "epoch": 1.2247430785646094,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048609131351048994,
      "loss": 5.4156,
      "mean_token_accuracy": 0.17732575088739394,
      "num_tokens": 24330169.0,
      "step": 10905
    },
    {
      "entropy": 5.518369913101196,
      "epoch": 1.225304655472567,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004860774131533195,
      "loss": 5.3681,
      "mean_token_accuracy": 0.1917150542140007,
      "num_tokens": 24340063.0,
      "step": 10910
    },
    {
      "entropy": 5.599963855743408,
      "epoch": 1.2258662323805245,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048606350607536235,
      "loss": 5.4386,
      "mean_token_accuracy": 0.18412175327539443,
      "num_tokens": 24351506.0,
      "step": 10915
    },
    {
      "entropy": 5.537996482849121,
      "epoch": 1.2264278092884822,
      "grad_norm": 2.9375,
      "learning_rate": 0.0004860495922770612,
      "loss": 5.4124,
      "mean_token_accuracy": 0.17802354991436004,
      "num_tokens": 24362168.0,
      "step": 10920
    },
    {
      "entropy": 5.511146926879883,
      "epoch": 1.2269893861964396,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048603567175885926,
      "loss": 5.3564,
      "mean_token_accuracy": 0.1828833281993866,
      "num_tokens": 24372269.0,
      "step": 10925
    },
    {
      "entropy": 5.5755280494689945,
      "epoch": 1.227550963104397,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004860217445211997,
      "loss": 5.4694,
      "mean_token_accuracy": 0.18339033424854279,
      "num_tokens": 24384368.0,
      "step": 10930
    },
    {
      "entropy": 5.639485788345337,
      "epoch": 1.2281125400123547,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004860078105645259,
      "loss": 5.3821,
      "mean_token_accuracy": 0.18572250306606292,
      "num_tokens": 24395395.0,
      "step": 10935
    },
    {
      "entropy": 5.521928453445435,
      "epoch": 1.2286741169203121,
      "grad_norm": 3.046875,
      "learning_rate": 0.0004859938698892816,
      "loss": 5.34,
      "mean_token_accuracy": 0.1804175704717636,
      "num_tokens": 24406709.0,
      "step": 10940
    },
    {
      "entropy": 5.538202095031738,
      "epoch": 1.2292356938282698,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048597992249591065,
      "loss": 5.4913,
      "mean_token_accuracy": 0.18075333684682846,
      "num_tokens": 24418574.0,
      "step": 10945
    },
    {
      "entropy": 5.719617748260498,
      "epoch": 1.2297972707362272,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048596596838485704,
      "loss": 5.4447,
      "mean_token_accuracy": 0.18116745054721833,
      "num_tokens": 24429418.0,
      "step": 10950
    },
    {
      "entropy": 5.615225076675415,
      "epoch": 1.230358847644185,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004859520075565654,
      "loss": 5.5106,
      "mean_token_accuracy": 0.18256430327892303,
      "num_tokens": 24440301.0,
      "step": 10955
    },
    {
      "entropy": 5.574560832977295,
      "epoch": 1.2309204245521423,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00048593804001147986,
      "loss": 5.5404,
      "mean_token_accuracy": 0.1761474683880806,
      "num_tokens": 24451987.0,
      "step": 10960
    },
    {
      "entropy": 5.59624605178833,
      "epoch": 1.2314820014601,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004859240657500453,
      "loss": 5.3956,
      "mean_token_accuracy": 0.18446308821439744,
      "num_tokens": 24461971.0,
      "step": 10965
    },
    {
      "entropy": 5.531461906433106,
      "epoch": 1.2320435783680574,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048591008477270675,
      "loss": 5.3155,
      "mean_token_accuracy": 0.18719284236431122,
      "num_tokens": 24472193.0,
      "step": 10970
    },
    {
      "entropy": 5.4965959072113035,
      "epoch": 1.232605155276015,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048589609707990925,
      "loss": 5.4151,
      "mean_token_accuracy": 0.1810591474175453,
      "num_tokens": 24483075.0,
      "step": 10975
    },
    {
      "entropy": 5.516279792785644,
      "epoch": 1.2331667321839725,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048588210267209823,
      "loss": 5.3412,
      "mean_token_accuracy": 0.19502355754375458,
      "num_tokens": 24493070.0,
      "step": 10980
    },
    {
      "entropy": 5.492527437210083,
      "epoch": 1.2337283090919302,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048586810154971935,
      "loss": 5.4668,
      "mean_token_accuracy": 0.17707120180130004,
      "num_tokens": 24502866.0,
      "step": 10985
    },
    {
      "entropy": 5.54741096496582,
      "epoch": 1.2342898859998876,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048585409371321826,
      "loss": 5.3397,
      "mean_token_accuracy": 0.19318298250436783,
      "num_tokens": 24514169.0,
      "step": 10990
    },
    {
      "entropy": 5.664519739151001,
      "epoch": 1.2348514629078453,
      "grad_norm": 2.609375,
      "learning_rate": 0.00048584007916304106,
      "loss": 5.5396,
      "mean_token_accuracy": 0.1791799932718277,
      "num_tokens": 24525552.0,
      "step": 10995
    },
    {
      "entropy": 5.563682317733765,
      "epoch": 1.2354130398158027,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048582605789963395,
      "loss": 5.4504,
      "mean_token_accuracy": 0.18420275896787644,
      "num_tokens": 24536370.0,
      "step": 11000
    },
    {
      "entropy": 5.597484731674195,
      "epoch": 1.2359746167237604,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004858120299234434,
      "loss": 5.4714,
      "mean_token_accuracy": 0.1837515950202942,
      "num_tokens": 24546868.0,
      "step": 11005
    },
    {
      "entropy": 5.65571551322937,
      "epoch": 1.2365361936317179,
      "grad_norm": 2.25,
      "learning_rate": 0.00048579799523491607,
      "loss": 5.4293,
      "mean_token_accuracy": 0.17340353578329087,
      "num_tokens": 24558279.0,
      "step": 11010
    },
    {
      "entropy": 5.577965450286865,
      "epoch": 1.2370977705396755,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004857839538344988,
      "loss": 5.3583,
      "mean_token_accuracy": 0.18387570977210999,
      "num_tokens": 24568321.0,
      "step": 11015
    },
    {
      "entropy": 5.52414608001709,
      "epoch": 1.237659347447633,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048576990572263875,
      "loss": 5.417,
      "mean_token_accuracy": 0.17986870557069778,
      "num_tokens": 24579137.0,
      "step": 11020
    },
    {
      "entropy": 5.611784029006958,
      "epoch": 1.2382209243555904,
      "grad_norm": 2.0,
      "learning_rate": 0.0004857558508997831,
      "loss": 5.4729,
      "mean_token_accuracy": 0.1754674106836319,
      "num_tokens": 24591523.0,
      "step": 11025
    },
    {
      "entropy": 5.613817977905273,
      "epoch": 1.238782501263548,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004857417893663794,
      "loss": 5.3944,
      "mean_token_accuracy": 0.18179088830947876,
      "num_tokens": 24601998.0,
      "step": 11030
    },
    {
      "entropy": 5.542374229431152,
      "epoch": 1.2393440781715055,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004857277211228754,
      "loss": 5.3645,
      "mean_token_accuracy": 0.18529855608940124,
      "num_tokens": 24612371.0,
      "step": 11035
    },
    {
      "entropy": 5.59985466003418,
      "epoch": 1.2399056550794632,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004857136461697191,
      "loss": 5.4681,
      "mean_token_accuracy": 0.18155949413776398,
      "num_tokens": 24623350.0,
      "step": 11040
    },
    {
      "entropy": 5.641364288330078,
      "epoch": 1.2404672319874206,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048569956450735853,
      "loss": 5.4495,
      "mean_token_accuracy": 0.18078392893075942,
      "num_tokens": 24634365.0,
      "step": 11045
    },
    {
      "entropy": 5.6573450565338135,
      "epoch": 1.2410288088953783,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048568547613624206,
      "loss": 5.541,
      "mean_token_accuracy": 0.18133490532636642,
      "num_tokens": 24645227.0,
      "step": 11050
    },
    {
      "entropy": 5.669416999816894,
      "epoch": 1.2415903858033357,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048567138105681833,
      "loss": 5.4892,
      "mean_token_accuracy": 0.1802353411912918,
      "num_tokens": 24656807.0,
      "step": 11055
    },
    {
      "entropy": 5.605979776382446,
      "epoch": 1.2421519627112934,
      "grad_norm": 2.015625,
      "learning_rate": 0.00048565727926953617,
      "loss": 5.4928,
      "mean_token_accuracy": 0.18245369344949722,
      "num_tokens": 24669021.0,
      "step": 11060
    },
    {
      "entropy": 5.581379795074463,
      "epoch": 1.2427135396192508,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004856431707748445,
      "loss": 5.393,
      "mean_token_accuracy": 0.18967633247375487,
      "num_tokens": 24678802.0,
      "step": 11065
    },
    {
      "entropy": 5.587186098098755,
      "epoch": 1.2432751165272085,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004856290555731926,
      "loss": 5.3781,
      "mean_token_accuracy": 0.18544357568025588,
      "num_tokens": 24688449.0,
      "step": 11070
    },
    {
      "entropy": 5.535901975631714,
      "epoch": 1.243836693435166,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004856149336650299,
      "loss": 5.4521,
      "mean_token_accuracy": 0.17772786617279052,
      "num_tokens": 24700272.0,
      "step": 11075
    },
    {
      "entropy": 5.649112796783447,
      "epoch": 1.2443982703431236,
      "grad_norm": 1.625,
      "learning_rate": 0.0004856008050508059,
      "loss": 5.5979,
      "mean_token_accuracy": 0.17577450573444367,
      "num_tokens": 24711949.0,
      "step": 11080
    },
    {
      "entropy": 5.661371088027954,
      "epoch": 1.244959847251081,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004855866697309707,
      "loss": 5.4132,
      "mean_token_accuracy": 0.1884407803416252,
      "num_tokens": 24723535.0,
      "step": 11085
    },
    {
      "entropy": 5.493203115463257,
      "epoch": 1.2455214241590387,
      "grad_norm": 1.375,
      "learning_rate": 0.0004855725277059742,
      "loss": 5.4221,
      "mean_token_accuracy": 0.17563299536705018,
      "num_tokens": 24734915.0,
      "step": 11090
    },
    {
      "entropy": 5.537456607818603,
      "epoch": 1.246083001066996,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004855583789762668,
      "loss": 5.3471,
      "mean_token_accuracy": 0.19495553225278855,
      "num_tokens": 24745744.0,
      "step": 11095
    },
    {
      "entropy": 5.653546333312988,
      "epoch": 1.2466445779749538,
      "grad_norm": 1.46875,
      "learning_rate": 0.000485544223542299,
      "loss": 5.5627,
      "mean_token_accuracy": 0.18065301030874253,
      "num_tokens": 24758158.0,
      "step": 11100
    },
    {
      "entropy": 5.655344104766845,
      "epoch": 1.2472061548829112,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004855300614045214,
      "loss": 5.426,
      "mean_token_accuracy": 0.184110064804554,
      "num_tokens": 24768607.0,
      "step": 11105
    },
    {
      "entropy": 5.563532066345215,
      "epoch": 1.2477677317908689,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048551589256338497,
      "loss": 5.4019,
      "mean_token_accuracy": 0.18666046857833862,
      "num_tokens": 24779845.0,
      "step": 11110
    },
    {
      "entropy": 5.534594440460205,
      "epoch": 1.2483293086988263,
      "grad_norm": 2.6875,
      "learning_rate": 0.00048550171701934087,
      "loss": 5.3665,
      "mean_token_accuracy": 0.17801106125116348,
      "num_tokens": 24790049.0,
      "step": 11115
    },
    {
      "entropy": 5.672994184494018,
      "epoch": 1.2488908856067837,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048548753477284053,
      "loss": 5.4315,
      "mean_token_accuracy": 0.18115769624710082,
      "num_tokens": 24802112.0,
      "step": 11120
    },
    {
      "entropy": 5.597347974777222,
      "epoch": 1.2494524625147414,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004854733458243354,
      "loss": 5.3822,
      "mean_token_accuracy": 0.1801334246993065,
      "num_tokens": 24812630.0,
      "step": 11125
    },
    {
      "entropy": 5.534896183013916,
      "epoch": 1.250014039422699,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004854591501742773,
      "loss": 5.3699,
      "mean_token_accuracy": 0.1891779601573944,
      "num_tokens": 24823346.0,
      "step": 11130
    },
    {
      "entropy": 5.553601455688477,
      "epoch": 1.2505756163306565,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048544494782311825,
      "loss": 5.4509,
      "mean_token_accuracy": 0.1782315269112587,
      "num_tokens": 24834374.0,
      "step": 11135
    },
    {
      "entropy": 5.551606559753418,
      "epoch": 1.251137193238614,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048543073877131037,
      "loss": 5.2873,
      "mean_token_accuracy": 0.18320416659116745,
      "num_tokens": 24845474.0,
      "step": 11140
    },
    {
      "entropy": 5.59578857421875,
      "epoch": 1.2516987701465716,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048541652301930627,
      "loss": 5.3773,
      "mean_token_accuracy": 0.1898516371846199,
      "num_tokens": 24857427.0,
      "step": 11145
    },
    {
      "entropy": 5.592479515075683,
      "epoch": 1.252260347054529,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004854023005675585,
      "loss": 5.3868,
      "mean_token_accuracy": 0.18503573685884475,
      "num_tokens": 24868352.0,
      "step": 11150
    },
    {
      "entropy": 5.593089580535889,
      "epoch": 1.2528219239624867,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004853880714165197,
      "loss": 5.4574,
      "mean_token_accuracy": 0.17789413332939147,
      "num_tokens": 24878499.0,
      "step": 11155
    },
    {
      "entropy": 5.557291555404663,
      "epoch": 1.2533835008704441,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048537383556664315,
      "loss": 5.4106,
      "mean_token_accuracy": 0.18766220808029174,
      "num_tokens": 24888290.0,
      "step": 11160
    },
    {
      "entropy": 5.606050968170166,
      "epoch": 1.2539450777784018,
      "grad_norm": 1.5,
      "learning_rate": 0.0004853595930183822,
      "loss": 5.4775,
      "mean_token_accuracy": 0.1763698562979698,
      "num_tokens": 24900201.0,
      "step": 11165
    },
    {
      "entropy": 5.540960454940796,
      "epoch": 1.2545066546863592,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004853453437721901,
      "loss": 5.3913,
      "mean_token_accuracy": 0.1846339926123619,
      "num_tokens": 24910895.0,
      "step": 11170
    },
    {
      "entropy": 5.543153429031372,
      "epoch": 1.255068231594317,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004853310878285208,
      "loss": 5.3216,
      "mean_token_accuracy": 0.19082948118448256,
      "num_tokens": 24922079.0,
      "step": 11175
    },
    {
      "entropy": 5.720382261276245,
      "epoch": 1.2556298085022743,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048531682518782797,
      "loss": 5.6323,
      "mean_token_accuracy": 0.17555736601352692,
      "num_tokens": 24935093.0,
      "step": 11180
    },
    {
      "entropy": 5.695417308807373,
      "epoch": 1.256191385410232,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048530255585056594,
      "loss": 5.5707,
      "mean_token_accuracy": 0.17180080711841583,
      "num_tokens": 24947316.0,
      "step": 11185
    },
    {
      "entropy": 5.511158132553101,
      "epoch": 1.2567529623181894,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004852882798171889,
      "loss": 5.2863,
      "mean_token_accuracy": 0.19156352430582047,
      "num_tokens": 24957255.0,
      "step": 11190
    },
    {
      "entropy": 5.510053825378418,
      "epoch": 1.2573145392261469,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004852739970881515,
      "loss": 5.4074,
      "mean_token_accuracy": 0.18630828112363815,
      "num_tokens": 24967646.0,
      "step": 11195
    },
    {
      "entropy": 5.659108686447143,
      "epoch": 1.2578761161341045,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048525970766390847,
      "loss": 5.5345,
      "mean_token_accuracy": 0.17736627161502838,
      "num_tokens": 24979748.0,
      "step": 11200
    },
    {
      "entropy": 5.711154413223267,
      "epoch": 1.2584376930420622,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004852454115449148,
      "loss": 5.5567,
      "mean_token_accuracy": 0.1786679282784462,
      "num_tokens": 24990727.0,
      "step": 11205
    },
    {
      "entropy": 5.538746738433838,
      "epoch": 1.2589992699500197,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004852311087316257,
      "loss": 5.364,
      "mean_token_accuracy": 0.1842673733830452,
      "num_tokens": 25001912.0,
      "step": 11210
    },
    {
      "entropy": 5.581915807723999,
      "epoch": 1.259560846857977,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004852167992244966,
      "loss": 5.4349,
      "mean_token_accuracy": 0.1812129110097885,
      "num_tokens": 25012952.0,
      "step": 11215
    },
    {
      "entropy": 5.576125526428223,
      "epoch": 1.2601224237659348,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000485202483023983,
      "loss": 5.3562,
      "mean_token_accuracy": 0.18378958404064177,
      "num_tokens": 25022851.0,
      "step": 11220
    },
    {
      "entropy": 5.520572996139526,
      "epoch": 1.2606840006738924,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048518816013054083,
      "loss": 5.3738,
      "mean_token_accuracy": 0.1824027717113495,
      "num_tokens": 25033177.0,
      "step": 11225
    },
    {
      "entropy": 5.541713047027588,
      "epoch": 1.2612455775818499,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048517383054462617,
      "loss": 5.3387,
      "mean_token_accuracy": 0.18704062849283218,
      "num_tokens": 25044140.0,
      "step": 11230
    },
    {
      "entropy": 5.530812120437622,
      "epoch": 1.2618071544898073,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004851594942666952,
      "loss": 5.3636,
      "mean_token_accuracy": 0.18201253265142442,
      "num_tokens": 25054480.0,
      "step": 11235
    },
    {
      "entropy": 5.585394048690796,
      "epoch": 1.262368731397765,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048514515129720445,
      "loss": 5.4568,
      "mean_token_accuracy": 0.17366419583559037,
      "num_tokens": 25065673.0,
      "step": 11240
    },
    {
      "entropy": 5.634736585617065,
      "epoch": 1.2629303083057224,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004851308016366105,
      "loss": 5.4308,
      "mean_token_accuracy": 0.18477717638015748,
      "num_tokens": 25076691.0,
      "step": 11245
    },
    {
      "entropy": 5.619279193878174,
      "epoch": 1.26349188521368,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048511644528537043,
      "loss": 5.4084,
      "mean_token_accuracy": 0.18738510608673095,
      "num_tokens": 25087882.0,
      "step": 11250
    },
    {
      "entropy": 5.62109408378601,
      "epoch": 1.2640534621216375,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004851020822439412,
      "loss": 5.5222,
      "mean_token_accuracy": 0.17479465305805206,
      "num_tokens": 25099325.0,
      "step": 11255
    },
    {
      "entropy": 5.484981060028076,
      "epoch": 1.2646150390295952,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048508771251278015,
      "loss": 5.3035,
      "mean_token_accuracy": 0.1885041892528534,
      "num_tokens": 25110304.0,
      "step": 11260
    },
    {
      "entropy": 5.5678627490997314,
      "epoch": 1.2651766159375526,
      "grad_norm": 1.625,
      "learning_rate": 0.0004850733360923449,
      "loss": 5.4497,
      "mean_token_accuracy": 0.18248049467802047,
      "num_tokens": 25121597.0,
      "step": 11265
    },
    {
      "entropy": 5.579066419601441,
      "epoch": 1.2657381928455103,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004850589529830931,
      "loss": 5.4294,
      "mean_token_accuracy": 0.17716689109802247,
      "num_tokens": 25132817.0,
      "step": 11270
    },
    {
      "entropy": 5.643011140823364,
      "epoch": 1.2662997697534677,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048504456318548285,
      "loss": 5.4824,
      "mean_token_accuracy": 0.18197162598371505,
      "num_tokens": 25143740.0,
      "step": 11275
    },
    {
      "entropy": 5.647823524475098,
      "epoch": 1.2668613466614254,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004850301666999722,
      "loss": 5.4238,
      "mean_token_accuracy": 0.18020653128623962,
      "num_tokens": 25155889.0,
      "step": 11280
    },
    {
      "entropy": 5.519939231872558,
      "epoch": 1.2674229235693828,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004850157635270196,
      "loss": 5.5169,
      "mean_token_accuracy": 0.1792319491505623,
      "num_tokens": 25166043.0,
      "step": 11285
    },
    {
      "entropy": 5.558336400985718,
      "epoch": 1.2679845004773402,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004850013536670836,
      "loss": 5.4013,
      "mean_token_accuracy": 0.18788154423236847,
      "num_tokens": 25176906.0,
      "step": 11290
    },
    {
      "entropy": 5.520482540130615,
      "epoch": 1.268546077385298,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004849869371206231,
      "loss": 5.3926,
      "mean_token_accuracy": 0.18359485119581223,
      "num_tokens": 25187803.0,
      "step": 11295
    },
    {
      "entropy": 5.554761219024658,
      "epoch": 1.2691076542932556,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048497251388809697,
      "loss": 5.344,
      "mean_token_accuracy": 0.1825688436627388,
      "num_tokens": 25199067.0,
      "step": 11300
    },
    {
      "entropy": 5.5986857414245605,
      "epoch": 1.269669231201213,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004849580839699647,
      "loss": 5.4171,
      "mean_token_accuracy": 0.1777142956852913,
      "num_tokens": 25210768.0,
      "step": 11305
    },
    {
      "entropy": 5.627563095092773,
      "epoch": 1.2702308081091704,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048494364736668554,
      "loss": 5.4782,
      "mean_token_accuracy": 0.1803862363100052,
      "num_tokens": 25221602.0,
      "step": 11310
    },
    {
      "entropy": 5.6025909900665285,
      "epoch": 1.270792385017128,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004849292040787192,
      "loss": 5.4338,
      "mean_token_accuracy": 0.18276409357786177,
      "num_tokens": 25232791.0,
      "step": 11315
    },
    {
      "entropy": 5.598787879943847,
      "epoch": 1.2713539619250858,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004849147541065256,
      "loss": 5.3964,
      "mean_token_accuracy": 0.18757463693618776,
      "num_tokens": 25243962.0,
      "step": 11320
    },
    {
      "entropy": 5.591303873062134,
      "epoch": 1.2719155388330432,
      "grad_norm": 1.375,
      "learning_rate": 0.0004849002974505649,
      "loss": 5.3985,
      "mean_token_accuracy": 0.1832546666264534,
      "num_tokens": 25254636.0,
      "step": 11325
    },
    {
      "entropy": 5.515753889083863,
      "epoch": 1.2724771157410006,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048488583411129723,
      "loss": 5.3738,
      "mean_token_accuracy": 0.1880288988351822,
      "num_tokens": 25265255.0,
      "step": 11330
    },
    {
      "entropy": 5.610970878601075,
      "epoch": 1.2730386926489583,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004848713640891832,
      "loss": 5.5106,
      "mean_token_accuracy": 0.18308686316013337,
      "num_tokens": 25276721.0,
      "step": 11335
    },
    {
      "entropy": 5.599207973480224,
      "epoch": 1.2736002695569157,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004848568873846835,
      "loss": 5.3997,
      "mean_token_accuracy": 0.19108064770698546,
      "num_tokens": 25287451.0,
      "step": 11340
    },
    {
      "entropy": 5.557778167724609,
      "epoch": 1.2741618464648734,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048484240399825925,
      "loss": 5.3552,
      "mean_token_accuracy": 0.18809493035078048,
      "num_tokens": 25298609.0,
      "step": 11345
    },
    {
      "entropy": 5.579873132705688,
      "epoch": 1.2747234233728308,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048482791393037134,
      "loss": 5.4116,
      "mean_token_accuracy": 0.184177727997303,
      "num_tokens": 25310058.0,
      "step": 11350
    },
    {
      "entropy": 5.59803032875061,
      "epoch": 1.2752850002807885,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048481341718148137,
      "loss": 5.4147,
      "mean_token_accuracy": 0.18216343820095063,
      "num_tokens": 25321249.0,
      "step": 11355
    },
    {
      "entropy": 5.567466831207275,
      "epoch": 1.275846577188746,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048479891375205074,
      "loss": 5.3734,
      "mean_token_accuracy": 0.1836025670170784,
      "num_tokens": 25332603.0,
      "step": 11360
    },
    {
      "entropy": 5.538301849365235,
      "epoch": 1.2764081540967036,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048478440364254136,
      "loss": 5.3926,
      "mean_token_accuracy": 0.19262938499450682,
      "num_tokens": 25344955.0,
      "step": 11365
    },
    {
      "entropy": 5.593557691574096,
      "epoch": 1.276969731004661,
      "grad_norm": 2.125,
      "learning_rate": 0.0004847698868534152,
      "loss": 5.522,
      "mean_token_accuracy": 0.17380810678005218,
      "num_tokens": 25356125.0,
      "step": 11370
    },
    {
      "entropy": 5.587371444702148,
      "epoch": 1.2775313079126187,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004847553633851345,
      "loss": 5.3496,
      "mean_token_accuracy": 0.1864945501089096,
      "num_tokens": 25366969.0,
      "step": 11375
    },
    {
      "entropy": 5.659316301345825,
      "epoch": 1.2780928848205761,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004847408332381616,
      "loss": 5.4353,
      "mean_token_accuracy": 0.1855549156665802,
      "num_tokens": 25378815.0,
      "step": 11380
    },
    {
      "entropy": 5.514390516281128,
      "epoch": 1.2786544617285336,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048472629641295933,
      "loss": 5.3468,
      "mean_token_accuracy": 0.18857791572809218,
      "num_tokens": 25389969.0,
      "step": 11385
    },
    {
      "entropy": 5.539934206008911,
      "epoch": 1.2792160386364912,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004847117529099904,
      "loss": 5.4205,
      "mean_token_accuracy": 0.17841923385858535,
      "num_tokens": 25401443.0,
      "step": 11390
    },
    {
      "entropy": 5.61718053817749,
      "epoch": 1.279777615544449,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048469720272971797,
      "loss": 5.4406,
      "mean_token_accuracy": 0.18052399754524232,
      "num_tokens": 25412857.0,
      "step": 11395
    },
    {
      "entropy": 5.517834424972534,
      "epoch": 1.2803391924524063,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004846826458726052,
      "loss": 5.3263,
      "mean_token_accuracy": 0.1841700106859207,
      "num_tokens": 25423804.0,
      "step": 11400
    },
    {
      "entropy": 5.569157838821411,
      "epoch": 1.2809007693603638,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048466808233911567,
      "loss": 5.4815,
      "mean_token_accuracy": 0.17656506299972535,
      "num_tokens": 25435686.0,
      "step": 11405
    },
    {
      "entropy": 5.654522323608399,
      "epoch": 1.2814623462683215,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048465351212971313,
      "loss": 5.4205,
      "mean_token_accuracy": 0.1811635062098503,
      "num_tokens": 25446012.0,
      "step": 11410
    },
    {
      "entropy": 5.654188346862793,
      "epoch": 1.2820239231762791,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004846389352448614,
      "loss": 5.4819,
      "mean_token_accuracy": 0.18275005370378494,
      "num_tokens": 25457768.0,
      "step": 11415
    },
    {
      "entropy": 5.60784878730774,
      "epoch": 1.2825855000842366,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004846243516850247,
      "loss": 5.4008,
      "mean_token_accuracy": 0.1821306124329567,
      "num_tokens": 25468706.0,
      "step": 11420
    },
    {
      "entropy": 5.647635555267334,
      "epoch": 1.283147076992194,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004846097614506673,
      "loss": 5.4781,
      "mean_token_accuracy": 0.17850955724716186,
      "num_tokens": 25479446.0,
      "step": 11425
    },
    {
      "entropy": 5.589556455612183,
      "epoch": 1.2837086539001517,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048459516454225384,
      "loss": 5.4035,
      "mean_token_accuracy": 0.18292578905820847,
      "num_tokens": 25491090.0,
      "step": 11430
    },
    {
      "entropy": 5.669900703430176,
      "epoch": 1.284270230808109,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000484580560960249,
      "loss": 5.4949,
      "mean_token_accuracy": 0.18713346719741822,
      "num_tokens": 25502105.0,
      "step": 11435
    },
    {
      "entropy": 5.716572904586792,
      "epoch": 1.2848318077160668,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004845659507051178,
      "loss": 5.5657,
      "mean_token_accuracy": 0.17665825933218002,
      "num_tokens": 25513970.0,
      "step": 11440
    },
    {
      "entropy": 5.581077337265015,
      "epoch": 1.2853933846240242,
      "grad_norm": 2.65625,
      "learning_rate": 0.0004845513337773255,
      "loss": 5.3877,
      "mean_token_accuracy": 0.18205983340740203,
      "num_tokens": 25524508.0,
      "step": 11445
    },
    {
      "entropy": 5.575494241714478,
      "epoch": 1.2859549615319819,
      "grad_norm": 2.3125,
      "learning_rate": 0.00048453671017733755,
      "loss": 5.3989,
      "mean_token_accuracy": 0.17941714376211165,
      "num_tokens": 25536006.0,
      "step": 11450
    },
    {
      "entropy": 5.538037443161011,
      "epoch": 1.2865165384399393,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004845220799056194,
      "loss": 5.4167,
      "mean_token_accuracy": 0.17841051369905472,
      "num_tokens": 25545726.0,
      "step": 11455
    },
    {
      "entropy": 5.547820711135865,
      "epoch": 1.287078115347897,
      "grad_norm": 1.75,
      "learning_rate": 0.00048450744296263685,
      "loss": 5.3506,
      "mean_token_accuracy": 0.1820966675877571,
      "num_tokens": 25555914.0,
      "step": 11460
    },
    {
      "entropy": 5.636511039733887,
      "epoch": 1.2876396922558544,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048449279934885625,
      "loss": 5.5396,
      "mean_token_accuracy": 0.17541532963514328,
      "num_tokens": 25568021.0,
      "step": 11465
    },
    {
      "entropy": 5.657505226135254,
      "epoch": 1.288201269163812,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004844781490647435,
      "loss": 5.434,
      "mean_token_accuracy": 0.18049876689910888,
      "num_tokens": 25578788.0,
      "step": 11470
    },
    {
      "entropy": 5.582758998870849,
      "epoch": 1.2887628460717695,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004844634921107654,
      "loss": 5.4571,
      "mean_token_accuracy": 0.18426980525255204,
      "num_tokens": 25590997.0,
      "step": 11475
    },
    {
      "entropy": 5.624828338623047,
      "epoch": 1.289324422979727,
      "grad_norm": 1.921875,
      "learning_rate": 0.00048444882848738835,
      "loss": 5.5133,
      "mean_token_accuracy": 0.17030210494995118,
      "num_tokens": 25602810.0,
      "step": 11480
    },
    {
      "entropy": 5.700673151016235,
      "epoch": 1.2898859998876846,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048443415819507944,
      "loss": 5.4931,
      "mean_token_accuracy": 0.18021892607212067,
      "num_tokens": 25614251.0,
      "step": 11485
    },
    {
      "entropy": 5.600652503967285,
      "epoch": 1.2904475767956423,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004844194812343057,
      "loss": 5.4952,
      "mean_token_accuracy": 0.18398517966270447,
      "num_tokens": 25626162.0,
      "step": 11490
    },
    {
      "entropy": 5.621218585968018,
      "epoch": 1.2910091537035997,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004844047976055344,
      "loss": 5.473,
      "mean_token_accuracy": 0.1821221113204956,
      "num_tokens": 25637598.0,
      "step": 11495
    },
    {
      "entropy": 5.6588897705078125,
      "epoch": 1.2915707306115571,
      "grad_norm": 1.75,
      "learning_rate": 0.00048439010730923313,
      "loss": 5.4176,
      "mean_token_accuracy": 0.18844881355762483,
      "num_tokens": 25648297.0,
      "step": 11500
    },
    {
      "entropy": 5.634639263153076,
      "epoch": 1.2921323075195148,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048437541034586964,
      "loss": 5.4856,
      "mean_token_accuracy": 0.17767189294099808,
      "num_tokens": 25660139.0,
      "step": 11505
    },
    {
      "entropy": 5.563700008392334,
      "epoch": 1.2926938844274725,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048436070671591195,
      "loss": 5.4008,
      "mean_token_accuracy": 0.18382530957460402,
      "num_tokens": 25671947.0,
      "step": 11510
    },
    {
      "entropy": 5.540437698364258,
      "epoch": 1.29325546133543,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048434599641982806,
      "loss": 5.3415,
      "mean_token_accuracy": 0.18799777030944825,
      "num_tokens": 25682059.0,
      "step": 11515
    },
    {
      "entropy": 5.595335960388184,
      "epoch": 1.2938170382433873,
      "grad_norm": 2.234375,
      "learning_rate": 0.00048433127945808645,
      "loss": 5.4388,
      "mean_token_accuracy": 0.1800031155347824,
      "num_tokens": 25692792.0,
      "step": 11520
    },
    {
      "entropy": 5.577079010009766,
      "epoch": 1.294378615151345,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004843165558311558,
      "loss": 5.4565,
      "mean_token_accuracy": 0.17733431309461595,
      "num_tokens": 25703253.0,
      "step": 11525
    },
    {
      "entropy": 5.642816686630249,
      "epoch": 1.2949401920593024,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004843018255395048,
      "loss": 5.3965,
      "mean_token_accuracy": 0.1866176277399063,
      "num_tokens": 25714169.0,
      "step": 11530
    },
    {
      "entropy": 5.609458255767822,
      "epoch": 1.29550176896726,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048428708858360244,
      "loss": 5.4945,
      "mean_token_accuracy": 0.17917241752147675,
      "num_tokens": 25725827.0,
      "step": 11535
    },
    {
      "entropy": 5.56046314239502,
      "epoch": 1.2960633458752175,
      "grad_norm": 3.4375,
      "learning_rate": 0.00048427234496391816,
      "loss": 5.4327,
      "mean_token_accuracy": 0.18428800702095033,
      "num_tokens": 25737951.0,
      "step": 11540
    },
    {
      "entropy": 5.581019639968872,
      "epoch": 1.2966249227831752,
      "grad_norm": 2.5,
      "learning_rate": 0.00048425759468092113,
      "loss": 5.4824,
      "mean_token_accuracy": 0.17608369290828704,
      "num_tokens": 25749676.0,
      "step": 11545
    },
    {
      "entropy": 5.661048555374146,
      "epoch": 1.2971864996911326,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004842428377350811,
      "loss": 5.4637,
      "mean_token_accuracy": 0.18972608149051667,
      "num_tokens": 25760005.0,
      "step": 11550
    },
    {
      "entropy": 5.671072864532471,
      "epoch": 1.2977480765990903,
      "grad_norm": 2.6875,
      "learning_rate": 0.00048422807412686796,
      "loss": 5.5169,
      "mean_token_accuracy": 0.17256719172000884,
      "num_tokens": 25770595.0,
      "step": 11555
    },
    {
      "entropy": 5.661506319046021,
      "epoch": 1.2983096535070477,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004842133038567519,
      "loss": 5.5335,
      "mean_token_accuracy": 0.18202261179685592,
      "num_tokens": 25782301.0,
      "step": 11560
    },
    {
      "entropy": 5.594887542724609,
      "epoch": 1.2988712304150054,
      "grad_norm": 2.765625,
      "learning_rate": 0.000484198526925203,
      "loss": 5.3719,
      "mean_token_accuracy": 0.18578586727380753,
      "num_tokens": 25792904.0,
      "step": 11565
    },
    {
      "entropy": 5.600375318527222,
      "epoch": 1.2994328073229628,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004841837433326919,
      "loss": 5.4983,
      "mean_token_accuracy": 0.17811461091041564,
      "num_tokens": 25804594.0,
      "step": 11570
    },
    {
      "entropy": 5.605722141265869,
      "epoch": 1.2999943842309205,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004841689530796893,
      "loss": 5.3978,
      "mean_token_accuracy": 0.18363674581050873,
      "num_tokens": 25816496.0,
      "step": 11575
    },
    {
      "entropy": 5.576470279693604,
      "epoch": 1.300555961138878,
      "grad_norm": 3.25,
      "learning_rate": 0.0004841541561666661,
      "loss": 5.3506,
      "mean_token_accuracy": 0.18314846009016036,
      "num_tokens": 25827271.0,
      "step": 11580
    },
    {
      "entropy": 5.49079384803772,
      "epoch": 1.3011175380468356,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048413935259409343,
      "loss": 5.4112,
      "mean_token_accuracy": 0.17746203988790513,
      "num_tokens": 25837869.0,
      "step": 11585
    },
    {
      "entropy": 5.539881849288941,
      "epoch": 1.301679114954793,
      "grad_norm": 2.203125,
      "learning_rate": 0.00048412454236244267,
      "loss": 5.3612,
      "mean_token_accuracy": 0.1847861185669899,
      "num_tokens": 25850244.0,
      "step": 11590
    },
    {
      "entropy": 5.626516580581665,
      "epoch": 1.3022406918627505,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00048410972547218545,
      "loss": 5.451,
      "mean_token_accuracy": 0.18317297101020813,
      "num_tokens": 25859981.0,
      "step": 11595
    },
    {
      "entropy": 5.539654350280761,
      "epoch": 1.3028022687707081,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048409490192379337,
      "loss": 5.3361,
      "mean_token_accuracy": 0.19252536147832872,
      "num_tokens": 25870393.0,
      "step": 11600
    },
    {
      "entropy": 5.587570095062256,
      "epoch": 1.3033638456786658,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004840800717177386,
      "loss": 5.4834,
      "mean_token_accuracy": 0.18229357302188873,
      "num_tokens": 25882517.0,
      "step": 11605
    },
    {
      "entropy": 5.586652374267578,
      "epoch": 1.3039254225866233,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004840652348544931,
      "loss": 5.4173,
      "mean_token_accuracy": 0.1851235270500183,
      "num_tokens": 25893628.0,
      "step": 11610
    },
    {
      "entropy": 5.658670234680176,
      "epoch": 1.3044869994945807,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004840503913345296,
      "loss": 5.4557,
      "mean_token_accuracy": 0.17808904200792314,
      "num_tokens": 25904891.0,
      "step": 11615
    },
    {
      "entropy": 5.6350267887115475,
      "epoch": 1.3050485764025384,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004840355411583205,
      "loss": 5.4281,
      "mean_token_accuracy": 0.17956867814064026,
      "num_tokens": 25914598.0,
      "step": 11620
    },
    {
      "entropy": 5.635906028747558,
      "epoch": 1.3056101533104958,
      "grad_norm": 2.484375,
      "learning_rate": 0.00048402068432633864,
      "loss": 5.4167,
      "mean_token_accuracy": 0.18526556342840195,
      "num_tokens": 25924412.0,
      "step": 11625
    },
    {
      "entropy": 5.537234210968018,
      "epoch": 1.3061717302184535,
      "grad_norm": 3.171875,
      "learning_rate": 0.0004840058208390572,
      "loss": 5.4194,
      "mean_token_accuracy": 0.17871478348970413,
      "num_tokens": 25935583.0,
      "step": 11630
    },
    {
      "entropy": 5.512251567840576,
      "epoch": 1.306733307126411,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004839909506969494,
      "loss": 5.338,
      "mean_token_accuracy": 0.18482319563627242,
      "num_tokens": 25946949.0,
      "step": 11635
    },
    {
      "entropy": 5.515132904052734,
      "epoch": 1.3072948840343686,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004839760739004886,
      "loss": 5.3429,
      "mean_token_accuracy": 0.18653517663478852,
      "num_tokens": 25957070.0,
      "step": 11640
    },
    {
      "entropy": 5.604286241531372,
      "epoch": 1.307856460942326,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048396119045014853,
      "loss": 5.4657,
      "mean_token_accuracy": 0.1766536921262741,
      "num_tokens": 25968375.0,
      "step": 11645
    },
    {
      "entropy": 5.687704467773438,
      "epoch": 1.3084180378502837,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004839463003464032,
      "loss": 5.4541,
      "mean_token_accuracy": 0.18244671523571016,
      "num_tokens": 25979690.0,
      "step": 11650
    },
    {
      "entropy": 5.653723621368409,
      "epoch": 1.308979614758241,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004839314035897265,
      "loss": 5.5593,
      "mean_token_accuracy": 0.1719047576189041,
      "num_tokens": 25990660.0,
      "step": 11655
    },
    {
      "entropy": 5.43707971572876,
      "epoch": 1.3095411916661988,
      "grad_norm": 1.828125,
      "learning_rate": 0.000483916500180593,
      "loss": 5.3169,
      "mean_token_accuracy": 0.18557359874248505,
      "num_tokens": 26002158.0,
      "step": 11660
    },
    {
      "entropy": 5.6041358470916744,
      "epoch": 1.3101027685741562,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048390159011947706,
      "loss": 5.3403,
      "mean_token_accuracy": 0.18800112903118132,
      "num_tokens": 26011890.0,
      "step": 11665
    },
    {
      "entropy": 5.591357707977295,
      "epoch": 1.3106643454821139,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004838866734068535,
      "loss": 5.4562,
      "mean_token_accuracy": 0.17782177180051803,
      "num_tokens": 26023548.0,
      "step": 11670
    },
    {
      "entropy": 5.581407451629639,
      "epoch": 1.3112259223900713,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048387175004319723,
      "loss": 5.4621,
      "mean_token_accuracy": 0.18198118060827256,
      "num_tokens": 26034178.0,
      "step": 11675
    },
    {
      "entropy": 5.562361097335815,
      "epoch": 1.311787499298029,
      "grad_norm": 3.0,
      "learning_rate": 0.0004838568200289834,
      "loss": 5.3856,
      "mean_token_accuracy": 0.18595320582389832,
      "num_tokens": 26045068.0,
      "step": 11680
    },
    {
      "entropy": 5.599232959747314,
      "epoch": 1.3123490762059864,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004838418833646875,
      "loss": 5.5081,
      "mean_token_accuracy": 0.17779320776462554,
      "num_tokens": 26057400.0,
      "step": 11685
    },
    {
      "entropy": 5.562916946411133,
      "epoch": 1.3129106531139438,
      "grad_norm": 1.921875,
      "learning_rate": 0.00048382694005078485,
      "loss": 5.4286,
      "mean_token_accuracy": 0.1821630820631981,
      "num_tokens": 26068981.0,
      "step": 11690
    },
    {
      "entropy": 5.64361310005188,
      "epoch": 1.3134722300219015,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048381199008775165,
      "loss": 5.3865,
      "mean_token_accuracy": 0.18470406383275986,
      "num_tokens": 26078902.0,
      "step": 11695
    },
    {
      "entropy": 5.46088490486145,
      "epoch": 1.3140338069298592,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004837970334760635,
      "loss": 5.2692,
      "mean_token_accuracy": 0.19096807092428209,
      "num_tokens": 26089891.0,
      "step": 11700
    },
    {
      "entropy": 5.485475730895996,
      "epoch": 1.3145953838378166,
      "grad_norm": 2.265625,
      "learning_rate": 0.000483782070216197,
      "loss": 5.3797,
      "mean_token_accuracy": 0.18987158834934234,
      "num_tokens": 26100701.0,
      "step": 11705
    },
    {
      "entropy": 5.52687087059021,
      "epoch": 1.315156960745774,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004837671003086283,
      "loss": 5.3908,
      "mean_token_accuracy": 0.18701540678739548,
      "num_tokens": 26110496.0,
      "step": 11710
    },
    {
      "entropy": 5.523316240310669,
      "epoch": 1.3157185376537317,
      "grad_norm": 2.0,
      "learning_rate": 0.0004837521237538342,
      "loss": 5.3529,
      "mean_token_accuracy": 0.18877830654382705,
      "num_tokens": 26120580.0,
      "step": 11715
    },
    {
      "entropy": 5.506276321411133,
      "epoch": 1.3162801145616891,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004837371405522915,
      "loss": 5.4204,
      "mean_token_accuracy": 0.18576560616493226,
      "num_tokens": 26130936.0,
      "step": 11720
    },
    {
      "entropy": 5.65242748260498,
      "epoch": 1.3168416914696468,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048372215070447734,
      "loss": 5.4132,
      "mean_token_accuracy": 0.18514294624328614,
      "num_tokens": 26143027.0,
      "step": 11725
    },
    {
      "entropy": 5.596934413909912,
      "epoch": 1.3174032683776042,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048370715421086884,
      "loss": 5.3053,
      "mean_token_accuracy": 0.18550607711076736,
      "num_tokens": 26152874.0,
      "step": 11730
    },
    {
      "entropy": 5.498037242889405,
      "epoch": 1.317964845285562,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004836921510719438,
      "loss": 5.2872,
      "mean_token_accuracy": 0.18218962997198104,
      "num_tokens": 26162781.0,
      "step": 11735
    },
    {
      "entropy": 5.490464353561402,
      "epoch": 1.3185264221935193,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004836771412881795,
      "loss": 5.3809,
      "mean_token_accuracy": 0.187151300907135,
      "num_tokens": 26173498.0,
      "step": 11740
    },
    {
      "entropy": 5.562977123260498,
      "epoch": 1.319087999101477,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004836621248600542,
      "loss": 5.3842,
      "mean_token_accuracy": 0.18796051293611526,
      "num_tokens": 26185090.0,
      "step": 11745
    },
    {
      "entropy": 5.658957481384277,
      "epoch": 1.3196495760094344,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004836471017880459,
      "loss": 5.6025,
      "mean_token_accuracy": 0.1729616180062294,
      "num_tokens": 26197048.0,
      "step": 11750
    },
    {
      "entropy": 5.568052053451538,
      "epoch": 1.320211152917392,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004836320720726329,
      "loss": 5.4367,
      "mean_token_accuracy": 0.18527210354804993,
      "num_tokens": 26208037.0,
      "step": 11755
    },
    {
      "entropy": 5.6516584873199465,
      "epoch": 1.3207727298253495,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004836170357142939,
      "loss": 5.4554,
      "mean_token_accuracy": 0.18128653913736342,
      "num_tokens": 26219345.0,
      "step": 11760
    },
    {
      "entropy": 5.66205587387085,
      "epoch": 1.3213343067333072,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004836019927135076,
      "loss": 5.4583,
      "mean_token_accuracy": 0.1837382659316063,
      "num_tokens": 26230299.0,
      "step": 11765
    },
    {
      "entropy": 5.693551540374756,
      "epoch": 1.3218958836412646,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004835869430707529,
      "loss": 5.5218,
      "mean_token_accuracy": 0.17959502041339875,
      "num_tokens": 26241427.0,
      "step": 11770
    },
    {
      "entropy": 5.5806210994720455,
      "epoch": 1.3224574605492223,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048357188678650905,
      "loss": 5.4826,
      "mean_token_accuracy": 0.18851787298917771,
      "num_tokens": 26252051.0,
      "step": 11775
    },
    {
      "entropy": 5.569334602355957,
      "epoch": 1.3230190374571797,
      "grad_norm": 4.59375,
      "learning_rate": 0.0004835568238612553,
      "loss": 5.4019,
      "mean_token_accuracy": 0.17965350449085235,
      "num_tokens": 26263640.0,
      "step": 11780
    },
    {
      "entropy": 5.71147723197937,
      "epoch": 1.3235806143651372,
      "grad_norm": 2.875,
      "learning_rate": 0.00048354175429547155,
      "loss": 5.5939,
      "mean_token_accuracy": 0.1769303485751152,
      "num_tokens": 26274996.0,
      "step": 11785
    },
    {
      "entropy": 5.760419845581055,
      "epoch": 1.3241421912730948,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004835266780896374,
      "loss": 5.5576,
      "mean_token_accuracy": 0.17181482911109924,
      "num_tokens": 26287859.0,
      "step": 11790
    },
    {
      "entropy": 5.569895267486572,
      "epoch": 1.3247037681810525,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004835115952442329,
      "loss": 5.3491,
      "mean_token_accuracy": 0.18307339251041413,
      "num_tokens": 26298302.0,
      "step": 11795
    },
    {
      "entropy": 5.614945650100708,
      "epoch": 1.32526534508901,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048349650575973843,
      "loss": 5.4486,
      "mean_token_accuracy": 0.1809273272752762,
      "num_tokens": 26310755.0,
      "step": 11800
    },
    {
      "entropy": 5.631687355041504,
      "epoch": 1.3258269219969674,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048348140963663436,
      "loss": 5.542,
      "mean_token_accuracy": 0.17422901540994645,
      "num_tokens": 26322041.0,
      "step": 11805
    },
    {
      "entropy": 5.630207347869873,
      "epoch": 1.326388498904925,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004834663068754012,
      "loss": 5.4791,
      "mean_token_accuracy": 0.18391888439655305,
      "num_tokens": 26334028.0,
      "step": 11810
    },
    {
      "entropy": 5.544382095336914,
      "epoch": 1.3269500758128827,
      "grad_norm": 2.5,
      "learning_rate": 0.0004834511974765202,
      "loss": 5.3836,
      "mean_token_accuracy": 0.18592217117547988,
      "num_tokens": 26344834.0,
      "step": 11815
    },
    {
      "entropy": 5.559587526321411,
      "epoch": 1.3275116527208402,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00048343608144047194,
      "loss": 5.394,
      "mean_token_accuracy": 0.18360603153705596,
      "num_tokens": 26354896.0,
      "step": 11820
    },
    {
      "entropy": 5.644568014144897,
      "epoch": 1.3280732296287976,
      "grad_norm": 1.6875,
      "learning_rate": 0.00048342095876773825,
      "loss": 5.4762,
      "mean_token_accuracy": 0.18239444643259048,
      "num_tokens": 26367487.0,
      "step": 11825
    },
    {
      "entropy": 5.630744314193725,
      "epoch": 1.3286348065367553,
      "grad_norm": 2.125,
      "learning_rate": 0.0004834058294588003,
      "loss": 5.479,
      "mean_token_accuracy": 0.17969143390655518,
      "num_tokens": 26379384.0,
      "step": 11830
    },
    {
      "entropy": 5.6427594184875485,
      "epoch": 1.3291963834447127,
      "grad_norm": 1.921875,
      "learning_rate": 0.00048339069351413993,
      "loss": 5.429,
      "mean_token_accuracy": 0.18330038636922835,
      "num_tokens": 26390268.0,
      "step": 11835
    },
    {
      "entropy": 5.535543394088745,
      "epoch": 1.3297579603526704,
      "grad_norm": 1.546875,
      "learning_rate": 0.000483375550934239,
      "loss": 5.5305,
      "mean_token_accuracy": 0.17828609347343444,
      "num_tokens": 26401348.0,
      "step": 11840
    },
    {
      "entropy": 5.5631701946258545,
      "epoch": 1.3303195372606278,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004833604017195798,
      "loss": 5.3935,
      "mean_token_accuracy": 0.1797025293111801,
      "num_tokens": 26411955.0,
      "step": 11845
    },
    {
      "entropy": 5.5863142013549805,
      "epoch": 1.3308811141685855,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004833452458706445,
      "loss": 5.3328,
      "mean_token_accuracy": 0.1938882663846016,
      "num_tokens": 26422200.0,
      "step": 11850
    },
    {
      "entropy": 5.592746877670288,
      "epoch": 1.331442691076543,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004833300833879159,
      "loss": 5.3499,
      "mean_token_accuracy": 0.18699274361133575,
      "num_tokens": 26433292.0,
      "step": 11855
    },
    {
      "entropy": 5.554074811935425,
      "epoch": 1.3320042679845006,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004833149142718765,
      "loss": 5.3882,
      "mean_token_accuracy": 0.1879195973277092,
      "num_tokens": 26444003.0,
      "step": 11860
    },
    {
      "entropy": 5.555953025817871,
      "epoch": 1.332565844892458,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004832997385230096,
      "loss": 5.3987,
      "mean_token_accuracy": 0.18996921330690383,
      "num_tokens": 26455743.0,
      "step": 11865
    },
    {
      "entropy": 5.543155670166016,
      "epoch": 1.3331274218004157,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004832845561417981,
      "loss": 5.3767,
      "mean_token_accuracy": 0.18165364116430283,
      "num_tokens": 26466032.0,
      "step": 11870
    },
    {
      "entropy": 5.601006317138672,
      "epoch": 1.333688998708373,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004832693671287256,
      "loss": 5.401,
      "mean_token_accuracy": 0.1820191591978073,
      "num_tokens": 26477660.0,
      "step": 11875
    },
    {
      "entropy": 5.540877723693848,
      "epoch": 1.3342505756163305,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004832541714842757,
      "loss": 5.3848,
      "mean_token_accuracy": 0.18291471302509307,
      "num_tokens": 26488778.0,
      "step": 11880
    },
    {
      "entropy": 5.655148315429687,
      "epoch": 1.3348121525242882,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004832389692089321,
      "loss": 5.4454,
      "mean_token_accuracy": 0.1753410890698433,
      "num_tokens": 26500771.0,
      "step": 11885
    },
    {
      "entropy": 5.571166133880615,
      "epoch": 1.3353737294322459,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048322376030317905,
      "loss": 5.3715,
      "mean_token_accuracy": 0.1887003391981125,
      "num_tokens": 26511786.0,
      "step": 11890
    },
    {
      "entropy": 5.556343078613281,
      "epoch": 1.3359353063402033,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004832085447675007,
      "loss": 5.4463,
      "mean_token_accuracy": 0.18483859449625015,
      "num_tokens": 26522273.0,
      "step": 11895
    },
    {
      "entropy": 5.75804123878479,
      "epoch": 1.3364968832481607,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048319332260238153,
      "loss": 5.7205,
      "mean_token_accuracy": 0.16999229341745375,
      "num_tokens": 26536619.0,
      "step": 11900
    },
    {
      "entropy": 5.735089111328125,
      "epoch": 1.3370584601561184,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048317809380830624,
      "loss": 5.4625,
      "mean_token_accuracy": 0.17418597191572188,
      "num_tokens": 26547807.0,
      "step": 11905
    },
    {
      "entropy": 5.627963924407959,
      "epoch": 1.337620037064076,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004831628583857597,
      "loss": 5.4546,
      "mean_token_accuracy": 0.1782565787434578,
      "num_tokens": 26558924.0,
      "step": 11910
    },
    {
      "entropy": 5.571965122222901,
      "epoch": 1.3381816139720335,
      "grad_norm": 1.90625,
      "learning_rate": 0.00048314761633522696,
      "loss": 5.4068,
      "mean_token_accuracy": 0.18269326090812682,
      "num_tokens": 26570123.0,
      "step": 11915
    },
    {
      "entropy": 5.65542631149292,
      "epoch": 1.338743190879991,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004831323676571934,
      "loss": 5.4918,
      "mean_token_accuracy": 0.18208651542663573,
      "num_tokens": 26582841.0,
      "step": 11920
    },
    {
      "entropy": 5.607376909255981,
      "epoch": 1.3393047677879486,
      "grad_norm": 3.15625,
      "learning_rate": 0.0004831171123521446,
      "loss": 5.3852,
      "mean_token_accuracy": 0.18670095056295394,
      "num_tokens": 26594397.0,
      "step": 11925
    },
    {
      "entropy": 5.5590019702911375,
      "epoch": 1.339866344695906,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004831018504205662,
      "loss": 5.4673,
      "mean_token_accuracy": 0.17675997614860534,
      "num_tokens": 26605055.0,
      "step": 11930
    },
    {
      "entropy": 5.593759393692016,
      "epoch": 1.3404279216038637,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048308658186294414,
      "loss": 5.514,
      "mean_token_accuracy": 0.1753700017929077,
      "num_tokens": 26616110.0,
      "step": 11935
    },
    {
      "entropy": 5.627068376541137,
      "epoch": 1.3409894985118211,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004830713066797646,
      "loss": 5.4105,
      "mean_token_accuracy": 0.18769517093896865,
      "num_tokens": 26627196.0,
      "step": 11940
    },
    {
      "entropy": 5.584965181350708,
      "epoch": 1.3415510754197788,
      "grad_norm": 1.6875,
      "learning_rate": 0.00048305602487151396,
      "loss": 5.4821,
      "mean_token_accuracy": 0.18012309074401855,
      "num_tokens": 26637962.0,
      "step": 11945
    },
    {
      "entropy": 5.51606125831604,
      "epoch": 1.3421126523277362,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004830407364386789,
      "loss": 5.3846,
      "mean_token_accuracy": 0.18352729231119155,
      "num_tokens": 26649366.0,
      "step": 11950
    },
    {
      "entropy": 5.768883609771729,
      "epoch": 1.342674229235694,
      "grad_norm": 1.8515625,
      "learning_rate": 0.000483025441381746,
      "loss": 5.6188,
      "mean_token_accuracy": 0.17198943495750427,
      "num_tokens": 26662455.0,
      "step": 11955
    },
    {
      "entropy": 5.583873987197876,
      "epoch": 1.3432358061436513,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048301013970120235,
      "loss": 5.4889,
      "mean_token_accuracy": 0.18145816326141356,
      "num_tokens": 26673467.0,
      "step": 11960
    },
    {
      "entropy": 5.538484048843384,
      "epoch": 1.343797383051609,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048299483139753517,
      "loss": 5.3531,
      "mean_token_accuracy": 0.1864481821656227,
      "num_tokens": 26684803.0,
      "step": 11965
    },
    {
      "entropy": 5.667122411727905,
      "epoch": 1.3443589599595664,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004829795164712319,
      "loss": 5.4569,
      "mean_token_accuracy": 0.1781364396214485,
      "num_tokens": 26696886.0,
      "step": 11970
    },
    {
      "entropy": 5.429471158981324,
      "epoch": 1.3449205368675239,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048296419492278023,
      "loss": 5.2782,
      "mean_token_accuracy": 0.1929266095161438,
      "num_tokens": 26708489.0,
      "step": 11975
    },
    {
      "entropy": 5.556291913986206,
      "epoch": 1.3454821137754815,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004829488667526679,
      "loss": 5.4411,
      "mean_token_accuracy": 0.18234337568283082,
      "num_tokens": 26719599.0,
      "step": 11980
    },
    {
      "entropy": 5.543146467208862,
      "epoch": 1.3460436906834392,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048293353196138294,
      "loss": 5.3606,
      "mean_token_accuracy": 0.18849279433488847,
      "num_tokens": 26730359.0,
      "step": 11985
    },
    {
      "entropy": 5.640635585784912,
      "epoch": 1.3466052675913966,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048291819054941375,
      "loss": 5.5639,
      "mean_token_accuracy": 0.17377166748046874,
      "num_tokens": 26742363.0,
      "step": 11990
    },
    {
      "entropy": 5.627382469177246,
      "epoch": 1.347166844499354,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004829028425172487,
      "loss": 5.5331,
      "mean_token_accuracy": 0.17799247652292252,
      "num_tokens": 26754058.0,
      "step": 11995
    },
    {
      "entropy": 5.582808637619019,
      "epoch": 1.3477284214073117,
      "grad_norm": 2.0,
      "learning_rate": 0.00048288748786537656,
      "loss": 5.4443,
      "mean_token_accuracy": 0.17711522728204726,
      "num_tokens": 26765602.0,
      "step": 12000
    },
    {
      "epoch": 1.3477284214073117,
      "eval_entropy": 5.474635892901987,
      "eval_loss": 5.5089006423950195,
      "eval_mean_token_accuracy": 0.18679746414583032,
      "eval_num_tokens": 26765602.0,
      "eval_runtime": 23.9588,
      "eval_samples_per_second": 1294.432,
      "eval_steps_per_second": 161.82,
      "step": 12000
    },
    {
      "entropy": 5.619168281555176,
      "epoch": 1.3482899983152694,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004828721265942861,
      "loss": 5.3781,
      "mean_token_accuracy": 0.18612894862890245,
      "num_tokens": 26776867.0,
      "step": 12005
    },
    {
      "entropy": 5.525385332107544,
      "epoch": 1.3488515752232269,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048285675870446653,
      "loss": 5.3686,
      "mean_token_accuracy": 0.18679242134094237,
      "num_tokens": 26786383.0,
      "step": 12010
    },
    {
      "entropy": 5.659783935546875,
      "epoch": 1.3494131521311843,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048284138419640724,
      "loss": 5.5115,
      "mean_token_accuracy": 0.1743268057703972,
      "num_tokens": 26797647.0,
      "step": 12015
    },
    {
      "entropy": 5.650413465499878,
      "epoch": 1.349974729039142,
      "grad_norm": 1.5,
      "learning_rate": 0.0004828260030705976,
      "loss": 5.364,
      "mean_token_accuracy": 0.1815338224172592,
      "num_tokens": 26808674.0,
      "step": 12020
    },
    {
      "entropy": 5.514647388458252,
      "epoch": 1.3505363059470994,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048281061532752734,
      "loss": 5.2714,
      "mean_token_accuracy": 0.1890656590461731,
      "num_tokens": 26820858.0,
      "step": 12025
    },
    {
      "entropy": 5.549510478973389,
      "epoch": 1.351097882855057,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004827952209676866,
      "loss": 5.3452,
      "mean_token_accuracy": 0.18801091462373734,
      "num_tokens": 26831432.0,
      "step": 12030
    },
    {
      "entropy": 5.600547218322754,
      "epoch": 1.3516594597630145,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048277981999156534,
      "loss": 5.4652,
      "mean_token_accuracy": 0.18412834256887436,
      "num_tokens": 26843240.0,
      "step": 12035
    },
    {
      "entropy": 5.650083303451538,
      "epoch": 1.3522210366709722,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048276441239965414,
      "loss": 5.4694,
      "mean_token_accuracy": 0.18023325204849244,
      "num_tokens": 26853556.0,
      "step": 12040
    },
    {
      "entropy": 5.595720195770264,
      "epoch": 1.3527826135789296,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004827489981924434,
      "loss": 5.4097,
      "mean_token_accuracy": 0.18325205147266388,
      "num_tokens": 26866040.0,
      "step": 12045
    },
    {
      "entropy": 5.494285726547242,
      "epoch": 1.3533441904868873,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048273357737042393,
      "loss": 5.263,
      "mean_token_accuracy": 0.18558333963155746,
      "num_tokens": 26876796.0,
      "step": 12050
    },
    {
      "entropy": 5.495880937576294,
      "epoch": 1.3539057673948447,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048271814993408694,
      "loss": 5.4558,
      "mean_token_accuracy": 0.17964325249195098,
      "num_tokens": 26888081.0,
      "step": 12055
    },
    {
      "entropy": 5.584234189987183,
      "epoch": 1.3544673443028024,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004827027158839233,
      "loss": 5.422,
      "mean_token_accuracy": 0.17885801792144776,
      "num_tokens": 26898511.0,
      "step": 12060
    },
    {
      "entropy": 5.638433265686035,
      "epoch": 1.3550289212107598,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004826872752204248,
      "loss": 5.4271,
      "mean_token_accuracy": 0.18308894336223602,
      "num_tokens": 26909525.0,
      "step": 12065
    },
    {
      "entropy": 5.599750947952271,
      "epoch": 1.3555904981187172,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048267182794408276,
      "loss": 5.4611,
      "mean_token_accuracy": 0.18938700258731841,
      "num_tokens": 26920267.0,
      "step": 12070
    },
    {
      "entropy": 5.596344470977783,
      "epoch": 1.356152075026675,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004826563740553893,
      "loss": 5.4296,
      "mean_token_accuracy": 0.18646478950977324,
      "num_tokens": 26931351.0,
      "step": 12075
    },
    {
      "entropy": 5.497289943695068,
      "epoch": 1.3567136519346326,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004826409135548363,
      "loss": 5.3661,
      "mean_token_accuracy": 0.17962656319141387,
      "num_tokens": 26942235.0,
      "step": 12080
    },
    {
      "entropy": 5.645137453079224,
      "epoch": 1.35727522884259,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004826254464429161,
      "loss": 5.5326,
      "mean_token_accuracy": 0.1759307160973549,
      "num_tokens": 26953375.0,
      "step": 12085
    },
    {
      "entropy": 5.621385192871093,
      "epoch": 1.3578368057505474,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048260997272012114,
      "loss": 5.4295,
      "mean_token_accuracy": 0.18265892565250397,
      "num_tokens": 26965081.0,
      "step": 12090
    },
    {
      "entropy": 5.610290002822876,
      "epoch": 1.358398382658505,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004825944923869441,
      "loss": 5.4589,
      "mean_token_accuracy": 0.1742696687579155,
      "num_tokens": 26975428.0,
      "step": 12095
    },
    {
      "entropy": 5.584657144546509,
      "epoch": 1.3589599595664628,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048257900544387795,
      "loss": 5.3793,
      "mean_token_accuracy": 0.18407012820243834,
      "num_tokens": 26987040.0,
      "step": 12100
    },
    {
      "entropy": 5.59127950668335,
      "epoch": 1.3595215364744202,
      "grad_norm": 1.375,
      "learning_rate": 0.00048256351189141574,
      "loss": 5.4826,
      "mean_token_accuracy": 0.17611093670129777,
      "num_tokens": 26997931.0,
      "step": 12105
    },
    {
      "entropy": 5.63390531539917,
      "epoch": 1.3600831133823776,
      "grad_norm": 1.625,
      "learning_rate": 0.0004825480117300508,
      "loss": 5.3812,
      "mean_token_accuracy": 0.1833555981516838,
      "num_tokens": 27009528.0,
      "step": 12110
    },
    {
      "entropy": 5.441861200332641,
      "epoch": 1.3606446902903353,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048253250496027675,
      "loss": 5.2788,
      "mean_token_accuracy": 0.19329649358987808,
      "num_tokens": 27020161.0,
      "step": 12115
    },
    {
      "entropy": 5.574842214584351,
      "epoch": 1.3612062671982927,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048251699158258714,
      "loss": 5.4026,
      "mean_token_accuracy": 0.18437140434980392,
      "num_tokens": 27032246.0,
      "step": 12120
    },
    {
      "entropy": 5.6189673900604244,
      "epoch": 1.3617678441062504,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004825014715974761,
      "loss": 5.4676,
      "mean_token_accuracy": 0.18131349235773087,
      "num_tokens": 27042903.0,
      "step": 12125
    },
    {
      "entropy": 5.668713092803955,
      "epoch": 1.3623294210142078,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048248594500543767,
      "loss": 5.5788,
      "mean_token_accuracy": 0.17454876899719238,
      "num_tokens": 27054854.0,
      "step": 12130
    },
    {
      "entropy": 5.61062593460083,
      "epoch": 1.3628909979221655,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048247041180696633,
      "loss": 5.401,
      "mean_token_accuracy": 0.19103506803512574,
      "num_tokens": 27065021.0,
      "step": 12135
    },
    {
      "entropy": 5.56108980178833,
      "epoch": 1.363452574830123,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004824548720025566,
      "loss": 5.4468,
      "mean_token_accuracy": 0.17915666699409485,
      "num_tokens": 27075717.0,
      "step": 12140
    },
    {
      "entropy": 5.608147001266479,
      "epoch": 1.3640141517380806,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004824393255927033,
      "loss": 5.5546,
      "mean_token_accuracy": 0.1782521888613701,
      "num_tokens": 27087310.0,
      "step": 12145
    },
    {
      "entropy": 5.69928469657898,
      "epoch": 1.364575728646038,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048242377257790145,
      "loss": 5.5615,
      "mean_token_accuracy": 0.1710447758436203,
      "num_tokens": 27100218.0,
      "step": 12150
    },
    {
      "entropy": 5.532822942733764,
      "epoch": 1.3651373055539957,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048240821295864614,
      "loss": 5.3169,
      "mean_token_accuracy": 0.18881547152996064,
      "num_tokens": 27111548.0,
      "step": 12155
    },
    {
      "entropy": 5.564854431152344,
      "epoch": 1.3656988824619531,
      "grad_norm": 1.8828125,
      "learning_rate": 0.000482392646735433,
      "loss": 5.4586,
      "mean_token_accuracy": 0.18205681890249253,
      "num_tokens": 27123434.0,
      "step": 12160
    },
    {
      "entropy": 5.627358484268188,
      "epoch": 1.3662604593699106,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004823770739087574,
      "loss": 5.4814,
      "mean_token_accuracy": 0.18017132431268693,
      "num_tokens": 27135936.0,
      "step": 12165
    },
    {
      "entropy": 5.766518306732178,
      "epoch": 1.3668220362778682,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004823614944791154,
      "loss": 5.6394,
      "mean_token_accuracy": 0.17236140221357346,
      "num_tokens": 27148240.0,
      "step": 12170
    },
    {
      "entropy": 5.628880310058594,
      "epoch": 1.367383613185826,
      "grad_norm": 1.625,
      "learning_rate": 0.0004823459084470031,
      "loss": 5.3949,
      "mean_token_accuracy": 0.18684293627738952,
      "num_tokens": 27159057.0,
      "step": 12175
    },
    {
      "entropy": 5.5123375415802,
      "epoch": 1.3679451900937833,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004823303158129165,
      "loss": 5.4063,
      "mean_token_accuracy": 0.18343780785799027,
      "num_tokens": 27169189.0,
      "step": 12180
    },
    {
      "entropy": 5.538809442520142,
      "epoch": 1.3685067670017408,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004823147165773523,
      "loss": 5.3756,
      "mean_token_accuracy": 0.18663349747657776,
      "num_tokens": 27180777.0,
      "step": 12185
    },
    {
      "entropy": 5.608754634857178,
      "epoch": 1.3690683439096984,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048229911074080715,
      "loss": 5.4179,
      "mean_token_accuracy": 0.1894705057144165,
      "num_tokens": 27190991.0,
      "step": 12190
    },
    {
      "entropy": 5.571418428421021,
      "epoch": 1.369629920817656,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048228349830377787,
      "loss": 5.4526,
      "mean_token_accuracy": 0.17986156195402145,
      "num_tokens": 27203776.0,
      "step": 12195
    },
    {
      "entropy": 5.492271375656128,
      "epoch": 1.3701914977256135,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048226787926676165,
      "loss": 5.3669,
      "mean_token_accuracy": 0.18599619418382646,
      "num_tokens": 27213533.0,
      "step": 12200
    },
    {
      "entropy": 5.615678501129151,
      "epoch": 1.370753074633571,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048225225363025576,
      "loss": 5.4857,
      "mean_token_accuracy": 0.17912229895591736,
      "num_tokens": 27224832.0,
      "step": 12205
    },
    {
      "entropy": 5.665140295028687,
      "epoch": 1.3713146515415287,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048223662139475773,
      "loss": 5.5298,
      "mean_token_accuracy": 0.17312545627355574,
      "num_tokens": 27237313.0,
      "step": 12210
    },
    {
      "entropy": 5.6187337875366214,
      "epoch": 1.371876228449486,
      "grad_norm": 1.5,
      "learning_rate": 0.00048222098256076523,
      "loss": 5.3418,
      "mean_token_accuracy": 0.18424533009529115,
      "num_tokens": 27248159.0,
      "step": 12215
    },
    {
      "entropy": 5.623969841003418,
      "epoch": 1.3724378053574438,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004822053371287763,
      "loss": 5.4661,
      "mean_token_accuracy": 0.18050478100776673,
      "num_tokens": 27259723.0,
      "step": 12220
    },
    {
      "entropy": 5.574584579467773,
      "epoch": 1.3729993822654012,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004821896850992891,
      "loss": 5.4104,
      "mean_token_accuracy": 0.17844849526882173,
      "num_tokens": 27270207.0,
      "step": 12225
    },
    {
      "entropy": 5.568454647064209,
      "epoch": 1.3735609591733589,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000482174026472802,
      "loss": 5.4621,
      "mean_token_accuracy": 0.188611501455307,
      "num_tokens": 27281243.0,
      "step": 12230
    },
    {
      "entropy": 5.4445079326629635,
      "epoch": 1.3741225360813163,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004821583612498135,
      "loss": 5.2634,
      "mean_token_accuracy": 0.18975935131311417,
      "num_tokens": 27292816.0,
      "step": 12235
    },
    {
      "entropy": 5.51634955406189,
      "epoch": 1.374684112989274,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048214268943082246,
      "loss": 5.4411,
      "mean_token_accuracy": 0.18239236921072005,
      "num_tokens": 27304254.0,
      "step": 12240
    },
    {
      "entropy": 5.529271507263184,
      "epoch": 1.3752456898972314,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004821270110163278,
      "loss": 5.3758,
      "mean_token_accuracy": 0.17939476817846298,
      "num_tokens": 27314974.0,
      "step": 12245
    },
    {
      "entropy": 5.587137651443482,
      "epoch": 1.375807266805189,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004821113260068288,
      "loss": 5.4711,
      "mean_token_accuracy": 0.17995233684778214,
      "num_tokens": 27326051.0,
      "step": 12250
    },
    {
      "entropy": 5.519277620315552,
      "epoch": 1.3763688437131465,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004820956344028249,
      "loss": 5.3203,
      "mean_token_accuracy": 0.18622014671564102,
      "num_tokens": 27337589.0,
      "step": 12255
    },
    {
      "entropy": 5.579274892807007,
      "epoch": 1.376930420621104,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048207993620481566,
      "loss": 5.4029,
      "mean_token_accuracy": 0.18406821340322493,
      "num_tokens": 27348581.0,
      "step": 12260
    },
    {
      "entropy": 5.6482140064239506,
      "epoch": 1.3774919975290616,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004820642314133009,
      "loss": 5.5,
      "mean_token_accuracy": 0.1824349656701088,
      "num_tokens": 27359963.0,
      "step": 12265
    },
    {
      "entropy": 5.583898305892944,
      "epoch": 1.3780535744370193,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048204852002878073,
      "loss": 5.3464,
      "mean_token_accuracy": 0.18646378070116043,
      "num_tokens": 27370149.0,
      "step": 12270
    },
    {
      "entropy": 5.462600898742676,
      "epoch": 1.3786151513449767,
      "grad_norm": 2.0,
      "learning_rate": 0.00048203280205175544,
      "loss": 5.3477,
      "mean_token_accuracy": 0.1911865547299385,
      "num_tokens": 27381657.0,
      "step": 12275
    },
    {
      "entropy": 5.587764310836792,
      "epoch": 1.3791767282529341,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004820170774827253,
      "loss": 5.4043,
      "mean_token_accuracy": 0.17355224043130874,
      "num_tokens": 27392679.0,
      "step": 12280
    },
    {
      "entropy": 5.503405570983887,
      "epoch": 1.3797383051608918,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004820013463221912,
      "loss": 5.3837,
      "mean_token_accuracy": 0.18209337592124938,
      "num_tokens": 27404003.0,
      "step": 12285
    },
    {
      "entropy": 5.5964930057525635,
      "epoch": 1.3802998820688495,
      "grad_norm": 1.75,
      "learning_rate": 0.000481985608570654,
      "loss": 5.4163,
      "mean_token_accuracy": 0.18666171431541442,
      "num_tokens": 27414593.0,
      "step": 12290
    },
    {
      "entropy": 5.658124732971191,
      "epoch": 1.380861458976807,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004819698642286147,
      "loss": 5.5594,
      "mean_token_accuracy": 0.17700323164463044,
      "num_tokens": 27426549.0,
      "step": 12295
    },
    {
      "entropy": 5.61198239326477,
      "epoch": 1.3814230358847643,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048195411329657464,
      "loss": 5.5078,
      "mean_token_accuracy": 0.1794409230351448,
      "num_tokens": 27439451.0,
      "step": 12300
    },
    {
      "entropy": 5.658680152893067,
      "epoch": 1.381984612792722,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004819383557750353,
      "loss": 5.4367,
      "mean_token_accuracy": 0.18478858917951585,
      "num_tokens": 27451153.0,
      "step": 12305
    },
    {
      "entropy": 5.594944858551026,
      "epoch": 1.3825461897006794,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048192259166449845,
      "loss": 5.3848,
      "mean_token_accuracy": 0.18378443419933319,
      "num_tokens": 27462229.0,
      "step": 12310
    },
    {
      "entropy": 5.531543779373169,
      "epoch": 1.383107766608637,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048190682096546606,
      "loss": 5.3906,
      "mean_token_accuracy": 0.18350973427295686,
      "num_tokens": 27473901.0,
      "step": 12315
    },
    {
      "entropy": 5.493347024917602,
      "epoch": 1.3836693435165945,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048189104367844034,
      "loss": 5.3864,
      "mean_token_accuracy": 0.1844016879796982,
      "num_tokens": 27485981.0,
      "step": 12320
    },
    {
      "entropy": 5.548513746261596,
      "epoch": 1.3842309204245522,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048187525980392343,
      "loss": 5.3151,
      "mean_token_accuracy": 0.1873913511633873,
      "num_tokens": 27495764.0,
      "step": 12325
    },
    {
      "entropy": 5.561328363418579,
      "epoch": 1.3847924973325096,
      "grad_norm": 2.28125,
      "learning_rate": 0.000481859469342418,
      "loss": 5.4035,
      "mean_token_accuracy": 0.185747367143631,
      "num_tokens": 27506602.0,
      "step": 12330
    },
    {
      "entropy": 5.430418491363525,
      "epoch": 1.3853540742404673,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004818436722944269,
      "loss": 5.3569,
      "mean_token_accuracy": 0.184988309442997,
      "num_tokens": 27517482.0,
      "step": 12335
    },
    {
      "entropy": 5.559353876113891,
      "epoch": 1.3859156511484247,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004818278686604531,
      "loss": 5.4023,
      "mean_token_accuracy": 0.18987395912408828,
      "num_tokens": 27528560.0,
      "step": 12340
    },
    {
      "entropy": 5.590506649017334,
      "epoch": 1.3864772280563824,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004818120584409996,
      "loss": 5.4171,
      "mean_token_accuracy": 0.18152203857898713,
      "num_tokens": 27540019.0,
      "step": 12345
    },
    {
      "entropy": 5.635791635513305,
      "epoch": 1.3870388049643398,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004817962416365701,
      "loss": 5.4703,
      "mean_token_accuracy": 0.1841760665178299,
      "num_tokens": 27550553.0,
      "step": 12350
    },
    {
      "entropy": 5.652179098129272,
      "epoch": 1.3876003818722973,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048178041824766796,
      "loss": 5.3921,
      "mean_token_accuracy": 0.18463724851608276,
      "num_tokens": 27562169.0,
      "step": 12355
    },
    {
      "entropy": 5.573112106323242,
      "epoch": 1.388161958780255,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004817645882747971,
      "loss": 5.3501,
      "mean_token_accuracy": 0.18608878701925277,
      "num_tokens": 27571732.0,
      "step": 12360
    },
    {
      "entropy": 5.548039531707763,
      "epoch": 1.3887235356882126,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004817487517184616,
      "loss": 5.3879,
      "mean_token_accuracy": 0.18541265428066253,
      "num_tokens": 27584394.0,
      "step": 12365
    },
    {
      "entropy": 5.508686971664429,
      "epoch": 1.38928511259617,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004817329085791657,
      "loss": 5.3564,
      "mean_token_accuracy": 0.19151072800159455,
      "num_tokens": 27594813.0,
      "step": 12370
    },
    {
      "entropy": 5.497928047180176,
      "epoch": 1.3898466895041275,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004817170588574138,
      "loss": 5.3126,
      "mean_token_accuracy": 0.19099864661693572,
      "num_tokens": 27605375.0,
      "step": 12375
    },
    {
      "entropy": 5.497993659973145,
      "epoch": 1.3904082664120851,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004817012025537105,
      "loss": 5.3271,
      "mean_token_accuracy": 0.18947082310914992,
      "num_tokens": 27617066.0,
      "step": 12380
    },
    {
      "entropy": 5.597756481170654,
      "epoch": 1.3909698433200428,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004816853396685608,
      "loss": 5.4817,
      "mean_token_accuracy": 0.18175812810659409,
      "num_tokens": 27628667.0,
      "step": 12385
    },
    {
      "entropy": 5.560773801803589,
      "epoch": 1.3915314202280002,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048166947020246973,
      "loss": 5.4373,
      "mean_token_accuracy": 0.18439250141382219,
      "num_tokens": 27638858.0,
      "step": 12390
    },
    {
      "entropy": 5.639008903503418,
      "epoch": 1.3920929971359577,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048165359415594245,
      "loss": 5.5106,
      "mean_token_accuracy": 0.18048693388700485,
      "num_tokens": 27650518.0,
      "step": 12395
    },
    {
      "entropy": 5.587079381942749,
      "epoch": 1.3926545740439153,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048163771152948473,
      "loss": 5.4689,
      "mean_token_accuracy": 0.18535237908363342,
      "num_tokens": 27661781.0,
      "step": 12400
    },
    {
      "entropy": 5.548201274871826,
      "epoch": 1.3932161509518728,
      "grad_norm": 1.296875,
      "learning_rate": 0.000481621822323602,
      "loss": 5.4116,
      "mean_token_accuracy": 0.1857258066534996,
      "num_tokens": 27673040.0,
      "step": 12405
    },
    {
      "entropy": 5.57417516708374,
      "epoch": 1.3937777278598305,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004816059265388004,
      "loss": 5.4514,
      "mean_token_accuracy": 0.1865248292684555,
      "num_tokens": 27683541.0,
      "step": 12410
    },
    {
      "entropy": 5.655316543579102,
      "epoch": 1.394339304767788,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048159002417558596,
      "loss": 5.4949,
      "mean_token_accuracy": 0.1792720377445221,
      "num_tokens": 27693648.0,
      "step": 12415
    },
    {
      "entropy": 5.596675729751587,
      "epoch": 1.3949008816757456,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004815741152344651,
      "loss": 5.322,
      "mean_token_accuracy": 0.18869377970695494,
      "num_tokens": 27703920.0,
      "step": 12420
    },
    {
      "entropy": 5.652033138275146,
      "epoch": 1.395462458583703,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004815581997159442,
      "loss": 5.4572,
      "mean_token_accuracy": 0.17444158643484114,
      "num_tokens": 27715869.0,
      "step": 12425
    },
    {
      "entropy": 5.5294694900512695,
      "epoch": 1.3960240354916607,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048154227762053003,
      "loss": 5.402,
      "mean_token_accuracy": 0.18923304080963135,
      "num_tokens": 27726727.0,
      "step": 12430
    },
    {
      "entropy": 5.565631341934204,
      "epoch": 1.396585612399618,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004815263489487298,
      "loss": 5.3971,
      "mean_token_accuracy": 0.18441684991121293,
      "num_tokens": 27738217.0,
      "step": 12435
    },
    {
      "entropy": 5.597028398513794,
      "epoch": 1.3971471893075758,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004815104137010504,
      "loss": 5.4008,
      "mean_token_accuracy": 0.183449324965477,
      "num_tokens": 27749022.0,
      "step": 12440
    },
    {
      "entropy": 5.541413402557373,
      "epoch": 1.3977087662155332,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004814944718779993,
      "loss": 5.3621,
      "mean_token_accuracy": 0.18739566802978516,
      "num_tokens": 27760589.0,
      "step": 12445
    },
    {
      "entropy": 5.665905904769898,
      "epoch": 1.3982703431234906,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004814785234800843,
      "loss": 5.4442,
      "mean_token_accuracy": 0.18137060031294822,
      "num_tokens": 27772460.0,
      "step": 12450
    },
    {
      "entropy": 5.499793148040771,
      "epoch": 1.3988319200314483,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048146256850781287,
      "loss": 5.3198,
      "mean_token_accuracy": 0.1835328683257103,
      "num_tokens": 27783921.0,
      "step": 12455
    },
    {
      "entropy": 5.597337436676026,
      "epoch": 1.399393496939406,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048144660696169327,
      "loss": 5.4682,
      "mean_token_accuracy": 0.1830869808793068,
      "num_tokens": 27794843.0,
      "step": 12460
    },
    {
      "entropy": 5.643671464920044,
      "epoch": 1.3999550738473634,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048143063884223365,
      "loss": 5.364,
      "mean_token_accuracy": 0.1830323278903961,
      "num_tokens": 27805973.0,
      "step": 12465
    },
    {
      "entropy": 5.555152940750122,
      "epoch": 1.4005166507553208,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004814146641499424,
      "loss": 5.4186,
      "mean_token_accuracy": 0.18221746683120726,
      "num_tokens": 27816916.0,
      "step": 12470
    },
    {
      "entropy": 5.5165797710418705,
      "epoch": 1.4010782276632785,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048139868288532816,
      "loss": 5.4158,
      "mean_token_accuracy": 0.18309104293584824,
      "num_tokens": 27828077.0,
      "step": 12475
    },
    {
      "entropy": 5.526023244857788,
      "epoch": 1.4016398045712362,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004813826950488998,
      "loss": 5.3147,
      "mean_token_accuracy": 0.18483598977327348,
      "num_tokens": 27838502.0,
      "step": 12480
    },
    {
      "entropy": 5.578309106826782,
      "epoch": 1.4022013814791936,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00048136670064116645,
      "loss": 5.4476,
      "mean_token_accuracy": 0.17935547083616257,
      "num_tokens": 27850125.0,
      "step": 12485
    },
    {
      "entropy": 5.602385520935059,
      "epoch": 1.402762958387151,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004813506996626373,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18731053471565245,
      "num_tokens": 27861099.0,
      "step": 12490
    },
    {
      "entropy": 5.608717012405395,
      "epoch": 1.4033245352951087,
      "grad_norm": 2.6875,
      "learning_rate": 0.0004813346921138218,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18824754655361176,
      "num_tokens": 27872769.0,
      "step": 12495
    },
    {
      "entropy": 5.632760953903198,
      "epoch": 1.4038861122030661,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004813186779952297,
      "loss": 5.4829,
      "mean_token_accuracy": 0.1809535726904869,
      "num_tokens": 27884448.0,
      "step": 12500
    },
    {
      "entropy": 5.465057134628296,
      "epoch": 1.4044476891110238,
      "grad_norm": 1.5,
      "learning_rate": 0.0004813026573073708,
      "loss": 5.3272,
      "mean_token_accuracy": 0.17773733884096146,
      "num_tokens": 27894074.0,
      "step": 12505
    },
    {
      "entropy": 5.678692483901978,
      "epoch": 1.4050092660189812,
      "grad_norm": 2.125,
      "learning_rate": 0.0004812866300507554,
      "loss": 5.5052,
      "mean_token_accuracy": 0.18612149655818938,
      "num_tokens": 27905150.0,
      "step": 12510
    },
    {
      "entropy": 5.655053567886353,
      "epoch": 1.405570842926939,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004812705962258936,
      "loss": 5.4147,
      "mean_token_accuracy": 0.1827722519636154,
      "num_tokens": 27916471.0,
      "step": 12515
    },
    {
      "entropy": 5.563557434082031,
      "epoch": 1.4061324198348963,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004812545558332961,
      "loss": 5.3875,
      "mean_token_accuracy": 0.18819346278905869,
      "num_tokens": 27928276.0,
      "step": 12520
    },
    {
      "entropy": 5.621595621109009,
      "epoch": 1.406693996742854,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048123850887347346,
      "loss": 5.4148,
      "mean_token_accuracy": 0.17566152215003966,
      "num_tokens": 27940417.0,
      "step": 12525
    },
    {
      "entropy": 5.474638271331787,
      "epoch": 1.4072555736508114,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048122245534693676,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19022467583417893,
      "num_tokens": 27950490.0,
      "step": 12530
    },
    {
      "entropy": 5.6266542911529545,
      "epoch": 1.407817150558769,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004812063952541971,
      "loss": 5.4814,
      "mean_token_accuracy": 0.17631009668111802,
      "num_tokens": 27962593.0,
      "step": 12535
    },
    {
      "entropy": 5.678593349456787,
      "epoch": 1.4083787274667265,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048119032859576577,
      "loss": 5.5351,
      "mean_token_accuracy": 0.1745996281504631,
      "num_tokens": 27973567.0,
      "step": 12540
    },
    {
      "entropy": 5.501290893554687,
      "epoch": 1.408940304374684,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004811742553721545,
      "loss": 5.3323,
      "mean_token_accuracy": 0.1903274819254875,
      "num_tokens": 27984383.0,
      "step": 12545
    },
    {
      "entropy": 5.646518278121948,
      "epoch": 1.4095018812826416,
      "grad_norm": 2.0,
      "learning_rate": 0.00048115817558387494,
      "loss": 5.4665,
      "mean_token_accuracy": 0.17970054000616073,
      "num_tokens": 27995735.0,
      "step": 12550
    },
    {
      "entropy": 5.6405247211456295,
      "epoch": 1.4100634581905993,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004811420892314391,
      "loss": 5.5232,
      "mean_token_accuracy": 0.17082713842391967,
      "num_tokens": 28008261.0,
      "step": 12555
    },
    {
      "entropy": 5.518653488159179,
      "epoch": 1.4106250350985567,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004811259963153592,
      "loss": 5.3617,
      "mean_token_accuracy": 0.18512748330831527,
      "num_tokens": 28019424.0,
      "step": 12560
    },
    {
      "entropy": 5.602254915237427,
      "epoch": 1.4111866120065142,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004811098968361476,
      "loss": 5.4736,
      "mean_token_accuracy": 0.18438162207603453,
      "num_tokens": 28030961.0,
      "step": 12565
    },
    {
      "entropy": 5.580341243743897,
      "epoch": 1.4117481889144718,
      "grad_norm": 1.8125,
      "learning_rate": 0.000481093790794317,
      "loss": 5.4445,
      "mean_token_accuracy": 0.17799729555845262,
      "num_tokens": 28042580.0,
      "step": 12570
    },
    {
      "entropy": 5.760495996475219,
      "epoch": 1.4123097658224295,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004810776781903801,
      "loss": 5.5708,
      "mean_token_accuracy": 0.1800316646695137,
      "num_tokens": 28053435.0,
      "step": 12575
    },
    {
      "entropy": 5.72199559211731,
      "epoch": 1.412871342730387,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048106155902485003,
      "loss": 5.543,
      "mean_token_accuracy": 0.17407634854316711,
      "num_tokens": 28065195.0,
      "step": 12580
    },
    {
      "entropy": 5.580473899841309,
      "epoch": 1.4134329196383444,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004810454332982401,
      "loss": 5.4714,
      "mean_token_accuracy": 0.18441999405622483,
      "num_tokens": 28077457.0,
      "step": 12585
    },
    {
      "entropy": 5.530069398880005,
      "epoch": 1.413994496546302,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004810293010110635,
      "loss": 5.374,
      "mean_token_accuracy": 0.18203333616256714,
      "num_tokens": 28087598.0,
      "step": 12590
    },
    {
      "entropy": 5.564488935470581,
      "epoch": 1.4145560734542595,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004810131621638342,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18823318481445311,
      "num_tokens": 28097232.0,
      "step": 12595
    },
    {
      "entropy": 5.5692249774932865,
      "epoch": 1.4151176503622171,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048099701675706583,
      "loss": 5.4612,
      "mean_token_accuracy": 0.18515441864728927,
      "num_tokens": 28107333.0,
      "step": 12600
    },
    {
      "entropy": 5.607000064849854,
      "epoch": 1.4156792272701746,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004809808647912725,
      "loss": 5.4363,
      "mean_token_accuracy": 0.17578344345092772,
      "num_tokens": 28117886.0,
      "step": 12605
    },
    {
      "entropy": 5.625807571411133,
      "epoch": 1.4162408041781323,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004809647062669686,
      "loss": 5.4257,
      "mean_token_accuracy": 0.18431251794099807,
      "num_tokens": 28128884.0,
      "step": 12610
    },
    {
      "entropy": 5.611122894287109,
      "epoch": 1.4168023810860897,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048094854118466856,
      "loss": 5.5125,
      "mean_token_accuracy": 0.17216638326644898,
      "num_tokens": 28139859.0,
      "step": 12615
    },
    {
      "entropy": 5.722675848007202,
      "epoch": 1.4173639579940474,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048093236954488715,
      "loss": 5.4578,
      "mean_token_accuracy": 0.18167351186275482,
      "num_tokens": 28150398.0,
      "step": 12620
    },
    {
      "entropy": 5.565816211700439,
      "epoch": 1.4179255349020048,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004809161913481392,
      "loss": 5.3456,
      "mean_token_accuracy": 0.19193344712257385,
      "num_tokens": 28160766.0,
      "step": 12625
    },
    {
      "entropy": 5.537726974487304,
      "epoch": 1.4184871118099625,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048090000659493985,
      "loss": 5.4534,
      "mean_token_accuracy": 0.1803365468978882,
      "num_tokens": 28171583.0,
      "step": 12630
    },
    {
      "entropy": 5.526329898834229,
      "epoch": 1.41904868871792,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048088381528580445,
      "loss": 5.2915,
      "mean_token_accuracy": 0.1874199703335762,
      "num_tokens": 28183598.0,
      "step": 12635
    },
    {
      "entropy": 5.602271223068238,
      "epoch": 1.4196102656258773,
      "grad_norm": 1.96875,
      "learning_rate": 0.00048086761742124846,
      "loss": 5.4352,
      "mean_token_accuracy": 0.17657459378242493,
      "num_tokens": 28193377.0,
      "step": 12640
    },
    {
      "entropy": 5.519383335113526,
      "epoch": 1.420171842533835,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048085141300178774,
      "loss": 5.3145,
      "mean_token_accuracy": 0.19562322795391082,
      "num_tokens": 28204536.0,
      "step": 12645
    },
    {
      "entropy": 5.523086929321289,
      "epoch": 1.4207334194417927,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048083520202793815,
      "loss": 5.4666,
      "mean_token_accuracy": 0.18537127673625947,
      "num_tokens": 28215806.0,
      "step": 12650
    },
    {
      "entropy": 5.608553647994995,
      "epoch": 1.42129499634975,
      "grad_norm": 1.5,
      "learning_rate": 0.0004808189845002159,
      "loss": 5.4048,
      "mean_token_accuracy": 0.1814317598938942,
      "num_tokens": 28226765.0,
      "step": 12655
    },
    {
      "entropy": 5.74329023361206,
      "epoch": 1.4218565732577075,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004808027604191374,
      "loss": 5.5899,
      "mean_token_accuracy": 0.17849552035331726,
      "num_tokens": 28238077.0,
      "step": 12660
    },
    {
      "entropy": 5.542377138137818,
      "epoch": 1.4224181501656652,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048078652978521914,
      "loss": 5.4332,
      "mean_token_accuracy": 0.1853796973824501,
      "num_tokens": 28248992.0,
      "step": 12665
    },
    {
      "entropy": 5.575919246673584,
      "epoch": 1.4229797270736229,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048077029259897797,
      "loss": 5.403,
      "mean_token_accuracy": 0.18549903780221938,
      "num_tokens": 28259818.0,
      "step": 12670
    },
    {
      "entropy": 5.549159860610962,
      "epoch": 1.4235413039815803,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048075404886093094,
      "loss": 5.3757,
      "mean_token_accuracy": 0.1832862064242363,
      "num_tokens": 28271139.0,
      "step": 12675
    },
    {
      "entropy": 5.616514015197754,
      "epoch": 1.4241028808895377,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048073779857159504,
      "loss": 5.4163,
      "mean_token_accuracy": 0.18590553998947143,
      "num_tokens": 28282093.0,
      "step": 12680
    },
    {
      "entropy": 5.550446653366089,
      "epoch": 1.4246644577974954,
      "grad_norm": 1.53125,
      "learning_rate": 0.000480721541731488,
      "loss": 5.3172,
      "mean_token_accuracy": 0.19124546498060227,
      "num_tokens": 28292701.0,
      "step": 12685
    },
    {
      "entropy": 5.481839513778686,
      "epoch": 1.4252260347054528,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048070527834112727,
      "loss": 5.365,
      "mean_token_accuracy": 0.18471651673316955,
      "num_tokens": 28303585.0,
      "step": 12690
    },
    {
      "entropy": 5.536582088470459,
      "epoch": 1.4257876116134105,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004806890084010306,
      "loss": 5.3138,
      "mean_token_accuracy": 0.1883379191160202,
      "num_tokens": 28314583.0,
      "step": 12695
    },
    {
      "entropy": 5.511456155776978,
      "epoch": 1.426349188521368,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048067273191171615,
      "loss": 5.3755,
      "mean_token_accuracy": 0.1849355146288872,
      "num_tokens": 28325895.0,
      "step": 12700
    },
    {
      "entropy": 5.630955028533935,
      "epoch": 1.4269107654293256,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004806564488737022,
      "loss": 5.4393,
      "mean_token_accuracy": 0.17738497406244277,
      "num_tokens": 28337011.0,
      "step": 12705
    },
    {
      "entropy": 5.73394136428833,
      "epoch": 1.427472342337283,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004806401592875071,
      "loss": 5.4884,
      "mean_token_accuracy": 0.17780508249998092,
      "num_tokens": 28347468.0,
      "step": 12710
    },
    {
      "entropy": 5.582927083969116,
      "epoch": 1.4280339192452407,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048062386315364964,
      "loss": 5.409,
      "mean_token_accuracy": 0.1885148122906685,
      "num_tokens": 28357287.0,
      "step": 12715
    },
    {
      "entropy": 5.632099342346192,
      "epoch": 1.4285954961531981,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004806075604726485,
      "loss": 5.4255,
      "mean_token_accuracy": 0.19006743729114534,
      "num_tokens": 28368508.0,
      "step": 12720
    },
    {
      "entropy": 5.629025793075561,
      "epoch": 1.4291570730611558,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048059125124502304,
      "loss": 5.3773,
      "mean_token_accuracy": 0.18756729513406753,
      "num_tokens": 28378211.0,
      "step": 12725
    },
    {
      "entropy": 5.47719259262085,
      "epoch": 1.4297186499691132,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004805749354712924,
      "loss": 5.3437,
      "mean_token_accuracy": 0.1933674082159996,
      "num_tokens": 28389083.0,
      "step": 12730
    },
    {
      "entropy": 5.546667385101318,
      "epoch": 1.4302802268770707,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000480558613151976,
      "loss": 5.3948,
      "mean_token_accuracy": 0.18546858429908752,
      "num_tokens": 28399786.0,
      "step": 12735
    },
    {
      "entropy": 5.515288591384888,
      "epoch": 1.4308418037850283,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004805422842875936,
      "loss": 5.4403,
      "mean_token_accuracy": 0.17986469864845275,
      "num_tokens": 28410650.0,
      "step": 12740
    },
    {
      "entropy": 5.658247184753418,
      "epoch": 1.431403380692986,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048052594887866524,
      "loss": 5.4536,
      "mean_token_accuracy": 0.18518347442150115,
      "num_tokens": 28421485.0,
      "step": 12745
    },
    {
      "entropy": 5.559001684188843,
      "epoch": 1.4319649576009434,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048050960692571086,
      "loss": 5.4078,
      "mean_token_accuracy": 0.17889249175786973,
      "num_tokens": 28432469.0,
      "step": 12750
    },
    {
      "entropy": 5.473956680297851,
      "epoch": 1.4325265345089009,
      "grad_norm": 1.453125,
      "learning_rate": 0.000480493258429251,
      "loss": 5.3408,
      "mean_token_accuracy": 0.18645906299352646,
      "num_tokens": 28443628.0,
      "step": 12755
    },
    {
      "entropy": 5.631206893920899,
      "epoch": 1.4330881114168585,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000480476903389806,
      "loss": 5.4609,
      "mean_token_accuracy": 0.17793752402067184,
      "num_tokens": 28455106.0,
      "step": 12760
    },
    {
      "entropy": 5.617642450332641,
      "epoch": 1.4336496883248162,
      "grad_norm": 1.625,
      "learning_rate": 0.0004804605418078968,
      "loss": 5.4179,
      "mean_token_accuracy": 0.18299987465143203,
      "num_tokens": 28466326.0,
      "step": 12765
    },
    {
      "entropy": 5.567136287689209,
      "epoch": 1.4342112652327736,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048044417368404426,
      "loss": 5.5462,
      "mean_token_accuracy": 0.18090612143278123,
      "num_tokens": 28477490.0,
      "step": 12770
    },
    {
      "entropy": 5.567657136917115,
      "epoch": 1.434772842140731,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048042779901876946,
      "loss": 5.3405,
      "mean_token_accuracy": 0.18477673530578614,
      "num_tokens": 28489780.0,
      "step": 12775
    },
    {
      "entropy": 5.6495922088623045,
      "epoch": 1.4353344190486887,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004804114178125939,
      "loss": 5.4825,
      "mean_token_accuracy": 0.18110819458961486,
      "num_tokens": 28500335.0,
      "step": 12780
    },
    {
      "entropy": 5.478008508682251,
      "epoch": 1.4358959959566462,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004803950300660391,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1903608411550522,
      "num_tokens": 28510680.0,
      "step": 12785
    },
    {
      "entropy": 5.486847877502441,
      "epoch": 1.4364575728646038,
      "grad_norm": 1.5,
      "learning_rate": 0.0004803786357796269,
      "loss": 5.3116,
      "mean_token_accuracy": 0.18844760060310364,
      "num_tokens": 28521719.0,
      "step": 12790
    },
    {
      "entropy": 5.55893497467041,
      "epoch": 1.4370191497725613,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048036223495387934,
      "loss": 5.4082,
      "mean_token_accuracy": 0.1853305444121361,
      "num_tokens": 28531935.0,
      "step": 12795
    },
    {
      "entropy": 5.608813142776489,
      "epoch": 1.437580726680519,
      "grad_norm": 1.375,
      "learning_rate": 0.0004803458275893185,
      "loss": 5.3988,
      "mean_token_accuracy": 0.18346945196390152,
      "num_tokens": 28542529.0,
      "step": 12800
    },
    {
      "entropy": 5.550368213653565,
      "epoch": 1.4381423035884764,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004803294136864669,
      "loss": 5.4149,
      "mean_token_accuracy": 0.18546790927648543,
      "num_tokens": 28554724.0,
      "step": 12805
    },
    {
      "entropy": 5.576000595092774,
      "epoch": 1.438703880496434,
      "grad_norm": 2.203125,
      "learning_rate": 0.00048031299324584716,
      "loss": 5.3648,
      "mean_token_accuracy": 0.18881458044052124,
      "num_tokens": 28565781.0,
      "step": 12810
    },
    {
      "entropy": 5.5950792789459225,
      "epoch": 1.4392654574043915,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000480296566267982,
      "loss": 5.3744,
      "mean_token_accuracy": 0.18683136254549026,
      "num_tokens": 28576186.0,
      "step": 12815
    },
    {
      "entropy": 5.525846147537232,
      "epoch": 1.4398270343123492,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004802801327533947,
      "loss": 5.3999,
      "mean_token_accuracy": 0.18060891479253768,
      "num_tokens": 28588240.0,
      "step": 12820
    },
    {
      "entropy": 5.628109312057495,
      "epoch": 1.4403886112203066,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048026369270260814,
      "loss": 5.4325,
      "mean_token_accuracy": 0.18242243230342864,
      "num_tokens": 28598240.0,
      "step": 12825
    },
    {
      "entropy": 5.609430503845215,
      "epoch": 1.440950188128264,
      "grad_norm": 2.734375,
      "learning_rate": 0.0004802472461161462,
      "loss": 5.4613,
      "mean_token_accuracy": 0.18128970116376877,
      "num_tokens": 28610091.0,
      "step": 12830
    },
    {
      "entropy": 5.628044462203979,
      "epoch": 1.4415117650362217,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004802307929945322,
      "loss": 5.4469,
      "mean_token_accuracy": 0.18347975760698318,
      "num_tokens": 28621060.0,
      "step": 12835
    },
    {
      "entropy": 5.635926198959351,
      "epoch": 1.4420733419441794,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048021433333829015,
      "loss": 5.4789,
      "mean_token_accuracy": 0.18282741606235503,
      "num_tokens": 28632027.0,
      "step": 12840
    },
    {
      "entropy": 5.671726369857788,
      "epoch": 1.4426349188521368,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048019786714794424,
      "loss": 5.5139,
      "mean_token_accuracy": 0.183630333840847,
      "num_tokens": 28643959.0,
      "step": 12845
    },
    {
      "entropy": 5.591601991653443,
      "epoch": 1.4431964957600942,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004801813944240186,
      "loss": 5.4171,
      "mean_token_accuracy": 0.1879538491368294,
      "num_tokens": 28654166.0,
      "step": 12850
    },
    {
      "entropy": 5.6133091926574705,
      "epoch": 1.443758072668052,
      "grad_norm": 1.375,
      "learning_rate": 0.0004801649151670378,
      "loss": 5.4046,
      "mean_token_accuracy": 0.18230132460594178,
      "num_tokens": 28667975.0,
      "step": 12855
    },
    {
      "entropy": 5.492213249206543,
      "epoch": 1.4443196495760096,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004801484293775265,
      "loss": 5.3787,
      "mean_token_accuracy": 0.18387535512447356,
      "num_tokens": 28679333.0,
      "step": 12860
    },
    {
      "entropy": 5.522174453735351,
      "epoch": 1.444881226483967,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004801319370560097,
      "loss": 5.2824,
      "mean_token_accuracy": 0.18480635285377503,
      "num_tokens": 28689333.0,
      "step": 12865
    },
    {
      "entropy": 5.563249206542968,
      "epoch": 1.4454428033919244,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004801154382030124,
      "loss": 5.4743,
      "mean_token_accuracy": 0.16807381361722945,
      "num_tokens": 28700625.0,
      "step": 12870
    },
    {
      "entropy": 5.6666676044464115,
      "epoch": 1.446004380299882,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048009893281906004,
      "loss": 5.4905,
      "mean_token_accuracy": 0.1777285158634186,
      "num_tokens": 28711768.0,
      "step": 12875
    },
    {
      "entropy": 5.589005374908448,
      "epoch": 1.4465659572078395,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048008242090467803,
      "loss": 5.4779,
      "mean_token_accuracy": 0.1786105677485466,
      "num_tokens": 28722521.0,
      "step": 12880
    },
    {
      "entropy": 5.579559087753296,
      "epoch": 1.4471275341157972,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004800659024603923,
      "loss": 5.422,
      "mean_token_accuracy": 0.18361022770404817,
      "num_tokens": 28733293.0,
      "step": 12885
    },
    {
      "entropy": 5.630140066146851,
      "epoch": 1.4476891110237546,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048004937748672864,
      "loss": 5.4078,
      "mean_token_accuracy": 0.18235597312450408,
      "num_tokens": 28744351.0,
      "step": 12890
    },
    {
      "entropy": 5.592684125900268,
      "epoch": 1.4482506879317123,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048003284598421327,
      "loss": 5.347,
      "mean_token_accuracy": 0.17893218547105788,
      "num_tokens": 28755140.0,
      "step": 12895
    },
    {
      "entropy": 5.5412681102752686,
      "epoch": 1.4488122648396697,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004800163079533726,
      "loss": 5.4643,
      "mean_token_accuracy": 0.1798962414264679,
      "num_tokens": 28767161.0,
      "step": 12900
    },
    {
      "entropy": 5.583915328979492,
      "epoch": 1.4493738417476274,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004799997633947332,
      "loss": 5.4035,
      "mean_token_accuracy": 0.18006214052438735,
      "num_tokens": 28779501.0,
      "step": 12905
    },
    {
      "entropy": 5.656120872497558,
      "epoch": 1.4499354186555848,
      "grad_norm": 2.21875,
      "learning_rate": 0.00047998321230882174,
      "loss": 5.5706,
      "mean_token_accuracy": 0.17843035608530045,
      "num_tokens": 28792021.0,
      "step": 12910
    },
    {
      "entropy": 5.5608038902282715,
      "epoch": 1.4504969955635425,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047996665469616536,
      "loss": 5.3407,
      "mean_token_accuracy": 0.18791882395744325,
      "num_tokens": 28802445.0,
      "step": 12915
    },
    {
      "entropy": 5.532950925827026,
      "epoch": 1.4510585724715,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047995009055729114,
      "loss": 5.3515,
      "mean_token_accuracy": 0.17671438902616501,
      "num_tokens": 28813821.0,
      "step": 12920
    },
    {
      "entropy": 5.575395822525024,
      "epoch": 1.4516201493794574,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004799335198927266,
      "loss": 5.4052,
      "mean_token_accuracy": 0.18110319525003432,
      "num_tokens": 28824764.0,
      "step": 12925
    },
    {
      "entropy": 5.645011758804321,
      "epoch": 1.452181726287415,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004799169427029993,
      "loss": 5.4385,
      "mean_token_accuracy": 0.18372306376695632,
      "num_tokens": 28836720.0,
      "step": 12930
    },
    {
      "entropy": 5.567357015609741,
      "epoch": 1.4527433031953727,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047990035898863706,
      "loss": 5.3383,
      "mean_token_accuracy": 0.18835611790418624,
      "num_tokens": 28848846.0,
      "step": 12935
    },
    {
      "entropy": 5.512576341629028,
      "epoch": 1.4533048801033301,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004798837687501679,
      "loss": 5.368,
      "mean_token_accuracy": 0.1866194039583206,
      "num_tokens": 28859453.0,
      "step": 12940
    },
    {
      "entropy": 5.541036367416382,
      "epoch": 1.4538664570112876,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004798671719881201,
      "loss": 5.4521,
      "mean_token_accuracy": 0.1791348472237587,
      "num_tokens": 28869447.0,
      "step": 12945
    },
    {
      "entropy": 5.652652931213379,
      "epoch": 1.4544280339192452,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004798505687030221,
      "loss": 5.4959,
      "mean_token_accuracy": 0.18058837205171585,
      "num_tokens": 28881660.0,
      "step": 12950
    },
    {
      "entropy": 5.6639848232269285,
      "epoch": 1.454989610827203,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004798339588954025,
      "loss": 5.4952,
      "mean_token_accuracy": 0.1759917840361595,
      "num_tokens": 28893128.0,
      "step": 12955
    },
    {
      "entropy": 5.5727091312408445,
      "epoch": 1.4555511877351603,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004798173425657903,
      "loss": 5.4054,
      "mean_token_accuracy": 0.1861330583691597,
      "num_tokens": 28904323.0,
      "step": 12960
    },
    {
      "entropy": 5.6069197177886965,
      "epoch": 1.4561127646431178,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004798007197147144,
      "loss": 5.4407,
      "mean_token_accuracy": 0.18523162007331848,
      "num_tokens": 28915672.0,
      "step": 12965
    },
    {
      "entropy": 5.73742413520813,
      "epoch": 1.4566743415510754,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004797840903427042,
      "loss": 5.5245,
      "mean_token_accuracy": 0.1795831501483917,
      "num_tokens": 28926657.0,
      "step": 12970
    },
    {
      "entropy": 5.563154983520508,
      "epoch": 1.4572359184590329,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047976745445028906,
      "loss": 5.3409,
      "mean_token_accuracy": 0.18444264233112334,
      "num_tokens": 28938787.0,
      "step": 12975
    },
    {
      "entropy": 5.519552993774414,
      "epoch": 1.4577974953669905,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047975081203799883,
      "loss": 5.4254,
      "mean_token_accuracy": 0.18172381669282914,
      "num_tokens": 28949836.0,
      "step": 12980
    },
    {
      "entropy": 5.532623434066773,
      "epoch": 1.458359072274948,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047973416310636327,
      "loss": 5.3584,
      "mean_token_accuracy": 0.18170606344938278,
      "num_tokens": 28960053.0,
      "step": 12985
    },
    {
      "entropy": 5.6591089248657225,
      "epoch": 1.4589206491829056,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004797175076559126,
      "loss": 5.5321,
      "mean_token_accuracy": 0.18378119617700578,
      "num_tokens": 28972698.0,
      "step": 12990
    },
    {
      "entropy": 5.645683240890503,
      "epoch": 1.459482226090863,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047970084568717703,
      "loss": 5.5249,
      "mean_token_accuracy": 0.18232111632823944,
      "num_tokens": 28984756.0,
      "step": 12995
    },
    {
      "entropy": 5.6243682384490965,
      "epoch": 1.4600438029988208,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004796841772006872,
      "loss": 5.4459,
      "mean_token_accuracy": 0.17344191670417786,
      "num_tokens": 28996915.0,
      "step": 13000
    },
    {
      "entropy": 5.540755319595337,
      "epoch": 1.4606053799067782,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00047966750219697376,
      "loss": 5.3666,
      "mean_token_accuracy": 0.18296120762825013,
      "num_tokens": 29007462.0,
      "step": 13005
    },
    {
      "entropy": 5.591592836380005,
      "epoch": 1.4611669568147359,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004796508206765677,
      "loss": 5.4545,
      "mean_token_accuracy": 0.18116957396268846,
      "num_tokens": 29017790.0,
      "step": 13010
    },
    {
      "entropy": 5.630393075942993,
      "epoch": 1.4617285337226933,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047963413264000005,
      "loss": 5.5117,
      "mean_token_accuracy": 0.18219025880098344,
      "num_tokens": 29029536.0,
      "step": 13015
    },
    {
      "entropy": 5.560514307022094,
      "epoch": 1.4622901106306507,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004796174380878023,
      "loss": 5.3121,
      "mean_token_accuracy": 0.18617115467786788,
      "num_tokens": 29039840.0,
      "step": 13020
    },
    {
      "entropy": 5.667528057098389,
      "epoch": 1.4628516875386084,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000479600737020506,
      "loss": 5.4887,
      "mean_token_accuracy": 0.17511072009801865,
      "num_tokens": 29052769.0,
      "step": 13025
    },
    {
      "entropy": 5.5814313888549805,
      "epoch": 1.463413264446566,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004795840294386429,
      "loss": 5.3535,
      "mean_token_accuracy": 0.1846114620566368,
      "num_tokens": 29063645.0,
      "step": 13030
    },
    {
      "entropy": 5.602754878997803,
      "epoch": 1.4639748413545235,
      "grad_norm": 1.765625,
      "learning_rate": 0.00047956731534274483,
      "loss": 5.4582,
      "mean_token_accuracy": 0.17893723547458648,
      "num_tokens": 29075241.0,
      "step": 13035
    },
    {
      "entropy": 5.6240301609039305,
      "epoch": 1.464536418262481,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047955059473334425,
      "loss": 5.4224,
      "mean_token_accuracy": 0.1843517988920212,
      "num_tokens": 29087592.0,
      "step": 13040
    },
    {
      "entropy": 5.545098924636841,
      "epoch": 1.4650979951704386,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004795338676109733,
      "loss": 5.391,
      "mean_token_accuracy": 0.18238293528556823,
      "num_tokens": 29099314.0,
      "step": 13045
    },
    {
      "entropy": 5.586019706726074,
      "epoch": 1.4656595720783963,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004795171339761647,
      "loss": 5.4493,
      "mean_token_accuracy": 0.1814487487077713,
      "num_tokens": 29111916.0,
      "step": 13050
    },
    {
      "entropy": 5.6290851593017575,
      "epoch": 1.4662211489863537,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047950039382945126,
      "loss": 5.4251,
      "mean_token_accuracy": 0.18521298468112946,
      "num_tokens": 29122871.0,
      "step": 13055
    },
    {
      "entropy": 5.560302591323852,
      "epoch": 1.4667827258943111,
      "grad_norm": 2.15625,
      "learning_rate": 0.00047948364717136587,
      "loss": 5.4137,
      "mean_token_accuracy": 0.18072257339954376,
      "num_tokens": 29133873.0,
      "step": 13060
    },
    {
      "entropy": 5.498769521713257,
      "epoch": 1.4673443028022688,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047946689400244196,
      "loss": 5.3212,
      "mean_token_accuracy": 0.18523789048194886,
      "num_tokens": 29144995.0,
      "step": 13065
    },
    {
      "entropy": 5.644399929046631,
      "epoch": 1.4679058797102262,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047945013432321284,
      "loss": 5.4945,
      "mean_token_accuracy": 0.17930782586336136,
      "num_tokens": 29156801.0,
      "step": 13070
    },
    {
      "entropy": 5.604148054122925,
      "epoch": 1.468467456618184,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047943336813421205,
      "loss": 5.5131,
      "mean_token_accuracy": 0.17270220518112184,
      "num_tokens": 29168072.0,
      "step": 13075
    },
    {
      "entropy": 5.550663137435913,
      "epoch": 1.4690290335261413,
      "grad_norm": 1.625,
      "learning_rate": 0.0004794165954359736,
      "loss": 5.3695,
      "mean_token_accuracy": 0.18234278857707978,
      "num_tokens": 29179314.0,
      "step": 13080
    },
    {
      "entropy": 5.630499362945557,
      "epoch": 1.469590610434099,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004793998162290315,
      "loss": 5.4712,
      "mean_token_accuracy": 0.17594666928052902,
      "num_tokens": 29190808.0,
      "step": 13085
    },
    {
      "entropy": 5.7291069507598875,
      "epoch": 1.4701521873420564,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004793830305139199,
      "loss": 5.5508,
      "mean_token_accuracy": 0.17875656634569168,
      "num_tokens": 29202557.0,
      "step": 13090
    },
    {
      "entropy": 5.5481040477752686,
      "epoch": 1.470713764250014,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004793662382911734,
      "loss": 5.3192,
      "mean_token_accuracy": 0.19619740098714827,
      "num_tokens": 29213959.0,
      "step": 13095
    },
    {
      "entropy": 5.595314407348633,
      "epoch": 1.4712753411579715,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004793494395613265,
      "loss": 5.4464,
      "mean_token_accuracy": 0.1871610775589943,
      "num_tokens": 29226069.0,
      "step": 13100
    },
    {
      "entropy": 5.649613761901856,
      "epoch": 1.4718369180659292,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004793326343249144,
      "loss": 5.4764,
      "mean_token_accuracy": 0.17758978456258773,
      "num_tokens": 29237882.0,
      "step": 13105
    },
    {
      "entropy": 5.53978533744812,
      "epoch": 1.4723984949738866,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004793158225824718,
      "loss": 5.4077,
      "mean_token_accuracy": 0.19019132256507873,
      "num_tokens": 29248642.0,
      "step": 13110
    },
    {
      "entropy": 5.496613502502441,
      "epoch": 1.4729600718818443,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047929900433453423,
      "loss": 5.3945,
      "mean_token_accuracy": 0.18171194493770598,
      "num_tokens": 29260632.0,
      "step": 13115
    },
    {
      "entropy": 5.694778394699097,
      "epoch": 1.4735216487898017,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047928217958163703,
      "loss": 5.5104,
      "mean_token_accuracy": 0.18265950977802276,
      "num_tokens": 29272563.0,
      "step": 13120
    },
    {
      "entropy": 5.673144721984864,
      "epoch": 1.4740832256977594,
      "grad_norm": 1.375,
      "learning_rate": 0.00047926534832431615,
      "loss": 5.477,
      "mean_token_accuracy": 0.18252675831317902,
      "num_tokens": 29284307.0,
      "step": 13125
    },
    {
      "entropy": 5.536691761016845,
      "epoch": 1.4746448026057168,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004792485105631073,
      "loss": 5.3236,
      "mean_token_accuracy": 0.18838772028684617,
      "num_tokens": 29295413.0,
      "step": 13130
    },
    {
      "entropy": 5.6502501487731935,
      "epoch": 1.4752063795136743,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004792316662985467,
      "loss": 5.4561,
      "mean_token_accuracy": 0.18947436511516572,
      "num_tokens": 29306612.0,
      "step": 13135
    },
    {
      "entropy": 5.562028217315674,
      "epoch": 1.475767956421632,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047921481553117053,
      "loss": 5.3896,
      "mean_token_accuracy": 0.18733631670475007,
      "num_tokens": 29316168.0,
      "step": 13140
    },
    {
      "entropy": 5.562285995483398,
      "epoch": 1.4763295333295896,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047919795826151544,
      "loss": 5.5198,
      "mean_token_accuracy": 0.1820887580513954,
      "num_tokens": 29326949.0,
      "step": 13145
    },
    {
      "entropy": 5.575417566299438,
      "epoch": 1.476891110237547,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004791810944901183,
      "loss": 5.3574,
      "mean_token_accuracy": 0.17930267602205277,
      "num_tokens": 29336744.0,
      "step": 13150
    },
    {
      "entropy": 5.6073588848114015,
      "epoch": 1.4774526871455045,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004791642242175158,
      "loss": 5.384,
      "mean_token_accuracy": 0.18314199000597,
      "num_tokens": 29346964.0,
      "step": 13155
    },
    {
      "entropy": 5.590050077438354,
      "epoch": 1.4780142640534621,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047914734744424517,
      "loss": 5.4227,
      "mean_token_accuracy": 0.18171590864658355,
      "num_tokens": 29358689.0,
      "step": 13160
    },
    {
      "entropy": 5.576000499725342,
      "epoch": 1.4785758409614196,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047913046417084393,
      "loss": 5.4339,
      "mean_token_accuracy": 0.1828928828239441,
      "num_tokens": 29371243.0,
      "step": 13165
    },
    {
      "entropy": 5.574913835525512,
      "epoch": 1.4791374178693772,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047911357439784947,
      "loss": 5.434,
      "mean_token_accuracy": 0.18351355493068694,
      "num_tokens": 29383185.0,
      "step": 13170
    },
    {
      "entropy": 5.481876945495605,
      "epoch": 1.4796989947773347,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047909667812579964,
      "loss": 5.2203,
      "mean_token_accuracy": 0.1933044970035553,
      "num_tokens": 29393986.0,
      "step": 13175
    },
    {
      "entropy": 5.500658226013184,
      "epoch": 1.4802605716852923,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047907977535523253,
      "loss": 5.4469,
      "mean_token_accuracy": 0.18550995141267776,
      "num_tokens": 29405407.0,
      "step": 13180
    },
    {
      "entropy": 5.555845069885254,
      "epoch": 1.4808221485932498,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004790628660866861,
      "loss": 5.3541,
      "mean_token_accuracy": 0.18297166675329207,
      "num_tokens": 29416352.0,
      "step": 13185
    },
    {
      "entropy": 5.556106805801392,
      "epoch": 1.4813837255012074,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004790459503206989,
      "loss": 5.2981,
      "mean_token_accuracy": 0.18662861436605455,
      "num_tokens": 29428369.0,
      "step": 13190
    },
    {
      "entropy": 5.486964416503906,
      "epoch": 1.4819453024091649,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004790290280578095,
      "loss": 5.3957,
      "mean_token_accuracy": 0.18780310004949569,
      "num_tokens": 29439276.0,
      "step": 13195
    },
    {
      "entropy": 5.645100069046021,
      "epoch": 1.4825068793171226,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047901209929855675,
      "loss": 5.4164,
      "mean_token_accuracy": 0.18187098652124406,
      "num_tokens": 29451361.0,
      "step": 13200
    },
    {
      "entropy": 5.531702375411987,
      "epoch": 1.48306845622508,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004789951640434796,
      "loss": 5.4417,
      "mean_token_accuracy": 0.18351416289806366,
      "num_tokens": 29462977.0,
      "step": 13205
    },
    {
      "entropy": 5.537817907333374,
      "epoch": 1.4836300331330377,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004789782222931173,
      "loss": 5.3729,
      "mean_token_accuracy": 0.1877996265888214,
      "num_tokens": 29474641.0,
      "step": 13210
    },
    {
      "entropy": 5.55263385772705,
      "epoch": 1.484191610040995,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047896127404800935,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1885400354862213,
      "num_tokens": 29485560.0,
      "step": 13215
    },
    {
      "entropy": 5.521211004257202,
      "epoch": 1.4847531869489528,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004789443193086953,
      "loss": 5.3831,
      "mean_token_accuracy": 0.18604848980903627,
      "num_tokens": 29495872.0,
      "step": 13220
    },
    {
      "entropy": 5.511582851409912,
      "epoch": 1.4853147638569102,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047892735807571505,
      "loss": 5.3327,
      "mean_token_accuracy": 0.1902692899107933,
      "num_tokens": 29506704.0,
      "step": 13225
    },
    {
      "entropy": 5.564581727981567,
      "epoch": 1.4858763407648676,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004789103903496086,
      "loss": 5.3683,
      "mean_token_accuracy": 0.18327164202928542,
      "num_tokens": 29518226.0,
      "step": 13230
    },
    {
      "entropy": 5.5849768161773685,
      "epoch": 1.4864379176728253,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004788934161309161,
      "loss": 5.3296,
      "mean_token_accuracy": 0.19128019660711287,
      "num_tokens": 29529176.0,
      "step": 13235
    },
    {
      "entropy": 5.470476198196411,
      "epoch": 1.486999494580783,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047887643542017836,
      "loss": 5.2218,
      "mean_token_accuracy": 0.18971727937459945,
      "num_tokens": 29539268.0,
      "step": 13240
    },
    {
      "entropy": 5.569934368133545,
      "epoch": 1.4875610714887404,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047885944821793574,
      "loss": 5.5137,
      "mean_token_accuracy": 0.1754368558526039,
      "num_tokens": 29550491.0,
      "step": 13245
    },
    {
      "entropy": 5.50352201461792,
      "epoch": 1.4881226483966978,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047884245452472926,
      "loss": 5.3284,
      "mean_token_accuracy": 0.1866819903254509,
      "num_tokens": 29561520.0,
      "step": 13250
    },
    {
      "entropy": 5.517250204086304,
      "epoch": 1.4886842253046555,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004788254543410999,
      "loss": 5.3886,
      "mean_token_accuracy": 0.18395980447530746,
      "num_tokens": 29573974.0,
      "step": 13255
    },
    {
      "entropy": 5.493844413757325,
      "epoch": 1.489245802212613,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047880844766758914,
      "loss": 5.2992,
      "mean_token_accuracy": 0.19390214681625367,
      "num_tokens": 29584059.0,
      "step": 13260
    },
    {
      "entropy": 5.647783708572388,
      "epoch": 1.4898073791205706,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047879143450473824,
      "loss": 5.5231,
      "mean_token_accuracy": 0.1793433427810669,
      "num_tokens": 29595726.0,
      "step": 13265
    },
    {
      "entropy": 5.555963134765625,
      "epoch": 1.490368956028528,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047877441485308917,
      "loss": 5.4138,
      "mean_token_accuracy": 0.18182573169469834,
      "num_tokens": 29607354.0,
      "step": 13270
    },
    {
      "entropy": 5.549566316604614,
      "epoch": 1.4909305329364857,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047875738871318356,
      "loss": 5.3485,
      "mean_token_accuracy": 0.18683600872755052,
      "num_tokens": 29618698.0,
      "step": 13275
    },
    {
      "entropy": 5.464523887634277,
      "epoch": 1.4914921098444431,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004787403560855637,
      "loss": 5.3188,
      "mean_token_accuracy": 0.18717088401317597,
      "num_tokens": 29629578.0,
      "step": 13280
    },
    {
      "entropy": 5.492139101028442,
      "epoch": 1.4920536867524008,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004787233169707719,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19385868310928345,
      "num_tokens": 29640277.0,
      "step": 13285
    },
    {
      "entropy": 5.497411632537842,
      "epoch": 1.4926152636603582,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004787062713693507,
      "loss": 5.4392,
      "mean_token_accuracy": 0.18391389548778533,
      "num_tokens": 29652029.0,
      "step": 13290
    },
    {
      "entropy": 5.581491851806641,
      "epoch": 1.493176840568316,
      "grad_norm": 1.5,
      "learning_rate": 0.00047868921928184276,
      "loss": 5.4319,
      "mean_token_accuracy": 0.18335914015769958,
      "num_tokens": 29663191.0,
      "step": 13295
    },
    {
      "entropy": 5.638195371627807,
      "epoch": 1.4937384174762733,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004786721607087911,
      "loss": 5.4582,
      "mean_token_accuracy": 0.17531873881816865,
      "num_tokens": 29674238.0,
      "step": 13300
    },
    {
      "entropy": 5.566009187698365,
      "epoch": 1.494299994384231,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004786550956507389,
      "loss": 5.3684,
      "mean_token_accuracy": 0.18569211661815643,
      "num_tokens": 29687255.0,
      "step": 13305
    },
    {
      "entropy": 5.474252653121948,
      "epoch": 1.4948615712921884,
      "grad_norm": 1.5,
      "learning_rate": 0.0004786380241082294,
      "loss": 5.3085,
      "mean_token_accuracy": 0.18408209681510926,
      "num_tokens": 29699769.0,
      "step": 13310
    },
    {
      "entropy": 5.601528692245483,
      "epoch": 1.495423148200146,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047862094608180623,
      "loss": 5.4829,
      "mean_token_accuracy": 0.17950135916471482,
      "num_tokens": 29710281.0,
      "step": 13315
    },
    {
      "entropy": 5.551890897750854,
      "epoch": 1.4959847251081035,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004786038615720132,
      "loss": 5.2745,
      "mean_token_accuracy": 0.1837810605764389,
      "num_tokens": 29720647.0,
      "step": 13320
    },
    {
      "entropy": 5.549169540405273,
      "epoch": 1.496546302016061,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047858677057939427,
      "loss": 5.4406,
      "mean_token_accuracy": 0.1845221921801567,
      "num_tokens": 29731565.0,
      "step": 13325
    },
    {
      "entropy": 5.552046060562134,
      "epoch": 1.4971078789240186,
      "grad_norm": 1.796875,
      "learning_rate": 0.00047856967310449356,
      "loss": 5.3408,
      "mean_token_accuracy": 0.18324533998966216,
      "num_tokens": 29743361.0,
      "step": 13330
    },
    {
      "entropy": 5.5653197288513185,
      "epoch": 1.4976694558319763,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047855256914785556,
      "loss": 5.3985,
      "mean_token_accuracy": 0.18504771292209626,
      "num_tokens": 29754799.0,
      "step": 13335
    },
    {
      "entropy": 5.57679591178894,
      "epoch": 1.4982310327399337,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047853545871002476,
      "loss": 5.5172,
      "mean_token_accuracy": 0.18125257045030593,
      "num_tokens": 29766401.0,
      "step": 13340
    },
    {
      "entropy": 5.487023830413818,
      "epoch": 1.4987926096478912,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047851834179154595,
      "loss": 5.2762,
      "mean_token_accuracy": 0.19227621853351592,
      "num_tokens": 29778268.0,
      "step": 13345
    },
    {
      "entropy": 5.4932716369628904,
      "epoch": 1.4993541865558488,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004785012183929643,
      "loss": 5.3409,
      "mean_token_accuracy": 0.1875157192349434,
      "num_tokens": 29788551.0,
      "step": 13350
    },
    {
      "entropy": 5.5828437328338625,
      "epoch": 1.4999157634638065,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004784840885148249,
      "loss": 5.3323,
      "mean_token_accuracy": 0.1854691818356514,
      "num_tokens": 29799726.0,
      "step": 13355
    },
    {
      "entropy": 5.564011764526367,
      "epoch": 1.500477340371764,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004784669521576732,
      "loss": 5.4717,
      "mean_token_accuracy": 0.1789710581302643,
      "num_tokens": 29810615.0,
      "step": 13360
    },
    {
      "entropy": 5.5941304683685305,
      "epoch": 1.5010389172797214,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004784498093220549,
      "loss": 5.3797,
      "mean_token_accuracy": 0.18280550539493562,
      "num_tokens": 29821766.0,
      "step": 13365
    },
    {
      "entropy": 5.513896894454956,
      "epoch": 1.501600494187679,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004784326600085157,
      "loss": 5.4186,
      "mean_token_accuracy": 0.18026091903448105,
      "num_tokens": 29833724.0,
      "step": 13370
    },
    {
      "entropy": 5.505526065826416,
      "epoch": 1.5021620710956367,
      "grad_norm": 1.875,
      "learning_rate": 0.00047841550421760164,
      "loss": 5.3227,
      "mean_token_accuracy": 0.18708880245685577,
      "num_tokens": 29845057.0,
      "step": 13375
    },
    {
      "entropy": 5.651104593276978,
      "epoch": 1.502723648003594,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047839834194985916,
      "loss": 5.4529,
      "mean_token_accuracy": 0.18251316100358964,
      "num_tokens": 29855649.0,
      "step": 13380
    },
    {
      "entropy": 5.469983243942261,
      "epoch": 1.5032852249115516,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004783811732058345,
      "loss": 5.274,
      "mean_token_accuracy": 0.18951803594827651,
      "num_tokens": 29866873.0,
      "step": 13385
    },
    {
      "entropy": 5.519393491744995,
      "epoch": 1.5038468018195092,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047836399798607446,
      "loss": 5.331,
      "mean_token_accuracy": 0.18913227319717407,
      "num_tokens": 29877690.0,
      "step": 13390
    },
    {
      "entropy": 5.528921747207642,
      "epoch": 1.5044083787274667,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047834681629112583,
      "loss": 5.3901,
      "mean_token_accuracy": 0.19536549896001815,
      "num_tokens": 29888461.0,
      "step": 13395
    },
    {
      "entropy": 5.482246589660645,
      "epoch": 1.5049699556354241,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004783296281215357,
      "loss": 5.3671,
      "mean_token_accuracy": 0.1873343139886856,
      "num_tokens": 29900559.0,
      "step": 13400
    },
    {
      "entropy": 5.457982730865479,
      "epoch": 1.5055315325433818,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004783124334778513,
      "loss": 5.2906,
      "mean_token_accuracy": 0.19076751917600632,
      "num_tokens": 29910815.0,
      "step": 13405
    },
    {
      "entropy": 5.6054411888122555,
      "epoch": 1.5060931094513395,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004782952323606202,
      "loss": 5.532,
      "mean_token_accuracy": 0.18430778980255128,
      "num_tokens": 29923262.0,
      "step": 13410
    },
    {
      "entropy": 5.541047430038452,
      "epoch": 1.506654686359297,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004782780247703901,
      "loss": 5.3602,
      "mean_token_accuracy": 0.18784934133291245,
      "num_tokens": 29934744.0,
      "step": 13415
    },
    {
      "entropy": 5.644116735458374,
      "epoch": 1.5072162632672543,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004782608107077088,
      "loss": 5.5336,
      "mean_token_accuracy": 0.17931553572416306,
      "num_tokens": 29945828.0,
      "step": 13420
    },
    {
      "entropy": 5.485628843307495,
      "epoch": 1.507777840175212,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047824359017312445,
      "loss": 5.3171,
      "mean_token_accuracy": 0.1859288841485977,
      "num_tokens": 29956964.0,
      "step": 13425
    },
    {
      "entropy": 5.5612828731536865,
      "epoch": 1.5083394170831697,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004782263631671854,
      "loss": 5.3824,
      "mean_token_accuracy": 0.1855710193514824,
      "num_tokens": 29968004.0,
      "step": 13430
    },
    {
      "entropy": 5.571560430526733,
      "epoch": 1.508900993991127,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047820912969044014,
      "loss": 5.3444,
      "mean_token_accuracy": 0.189272677898407,
      "num_tokens": 29978237.0,
      "step": 13435
    },
    {
      "entropy": 5.493322277069092,
      "epoch": 1.5094625708990845,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047819188974343734,
      "loss": 5.4205,
      "mean_token_accuracy": 0.17970777153968812,
      "num_tokens": 29988379.0,
      "step": 13440
    },
    {
      "entropy": 5.5735115051269535,
      "epoch": 1.5100241478070422,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004781746433267259,
      "loss": 5.4585,
      "mean_token_accuracy": 0.18298395723104477,
      "num_tokens": 29999024.0,
      "step": 13445
    },
    {
      "entropy": 5.618076181411743,
      "epoch": 1.5105857247149999,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047815739044085515,
      "loss": 5.3546,
      "mean_token_accuracy": 0.18803201615810394,
      "num_tokens": 30009411.0,
      "step": 13450
    },
    {
      "entropy": 5.50294942855835,
      "epoch": 1.5111473016229573,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047814013108637423,
      "loss": 5.3741,
      "mean_token_accuracy": 0.18963730782270433,
      "num_tokens": 30020746.0,
      "step": 13455
    },
    {
      "entropy": 5.566942977905273,
      "epoch": 1.5117088785309147,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047812286526383274,
      "loss": 5.3481,
      "mean_token_accuracy": 0.18417930454015732,
      "num_tokens": 30032568.0,
      "step": 13460
    },
    {
      "entropy": 5.58520917892456,
      "epoch": 1.5122704554388724,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004781055929737804,
      "loss": 5.4001,
      "mean_token_accuracy": 0.18209205716848373,
      "num_tokens": 30042772.0,
      "step": 13465
    },
    {
      "entropy": 5.557912015914917,
      "epoch": 1.51283203234683,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047808831421676724,
      "loss": 5.4403,
      "mean_token_accuracy": 0.1883595496416092,
      "num_tokens": 30053842.0,
      "step": 13470
    },
    {
      "entropy": 5.63917121887207,
      "epoch": 1.5133936092547875,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004780710289933435,
      "loss": 5.4059,
      "mean_token_accuracy": 0.1867174282670021,
      "num_tokens": 30064744.0,
      "step": 13475
    },
    {
      "entropy": 5.541257667541504,
      "epoch": 1.513955186162745,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004780537373040593,
      "loss": 5.3928,
      "mean_token_accuracy": 0.1826452910900116,
      "num_tokens": 30075475.0,
      "step": 13480
    },
    {
      "entropy": 5.565896034240723,
      "epoch": 1.5145167630707026,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047803643914946537,
      "loss": 5.3805,
      "mean_token_accuracy": 0.18455331176519393,
      "num_tokens": 30087189.0,
      "step": 13485
    },
    {
      "entropy": 5.576182174682617,
      "epoch": 1.51507833997866,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004780191345301125,
      "loss": 5.4967,
      "mean_token_accuracy": 0.17630792707204818,
      "num_tokens": 30099263.0,
      "step": 13490
    },
    {
      "entropy": 5.564552450180054,
      "epoch": 1.5156399168866175,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047800182344655165,
      "loss": 5.417,
      "mean_token_accuracy": 0.18011668920516968,
      "num_tokens": 30110925.0,
      "step": 13495
    },
    {
      "entropy": 5.598606777191162,
      "epoch": 1.5162014937945751,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000477984505899334,
      "loss": 5.4129,
      "mean_token_accuracy": 0.17879739701747893,
      "num_tokens": 30122708.0,
      "step": 13500
    },
    {
      "entropy": 5.53363790512085,
      "epoch": 1.5167630707025328,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047796718188901095,
      "loss": 5.2992,
      "mean_token_accuracy": 0.18720662593841553,
      "num_tokens": 30133624.0,
      "step": 13505
    },
    {
      "entropy": 5.4703856945037845,
      "epoch": 1.5173246476104902,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047794985141613413,
      "loss": 5.3796,
      "mean_token_accuracy": 0.18744681477546693,
      "num_tokens": 30144501.0,
      "step": 13510
    },
    {
      "entropy": 5.5106078624725345,
      "epoch": 1.5178862245184477,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047793251448125526,
      "loss": 5.3317,
      "mean_token_accuracy": 0.1898783937096596,
      "num_tokens": 30156351.0,
      "step": 13515
    },
    {
      "entropy": 5.570866870880127,
      "epoch": 1.5184478014264053,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004779151710849265,
      "loss": 5.4284,
      "mean_token_accuracy": 0.1814264938235283,
      "num_tokens": 30167445.0,
      "step": 13520
    },
    {
      "entropy": 5.620449876785278,
      "epoch": 1.519009378334363,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004778978212276999,
      "loss": 5.49,
      "mean_token_accuracy": 0.17341291755437852,
      "num_tokens": 30179054.0,
      "step": 13525
    },
    {
      "entropy": 5.637458610534668,
      "epoch": 1.5195709552423204,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004778804649101281,
      "loss": 5.4138,
      "mean_token_accuracy": 0.18240256160497664,
      "num_tokens": 30189412.0,
      "step": 13530
    },
    {
      "entropy": 5.5557867050170895,
      "epoch": 1.5201325321502779,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004778631021327636,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18172027468681334,
      "num_tokens": 30200775.0,
      "step": 13535
    },
    {
      "entropy": 5.553967905044556,
      "epoch": 1.5206941090582355,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004778457328961592,
      "loss": 5.3454,
      "mean_token_accuracy": 0.18539057523012162,
      "num_tokens": 30212121.0,
      "step": 13540
    },
    {
      "entropy": 5.494845914840698,
      "epoch": 1.5212556859661932,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047782835720086797,
      "loss": 5.3007,
      "mean_token_accuracy": 0.18967946916818618,
      "num_tokens": 30222814.0,
      "step": 13545
    },
    {
      "entropy": 5.5532997131347654,
      "epoch": 1.5218172628741506,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004778109750474432,
      "loss": 5.4067,
      "mean_token_accuracy": 0.18181711733341216,
      "num_tokens": 30232414.0,
      "step": 13550
    },
    {
      "entropy": 5.634558773040771,
      "epoch": 1.522378839782108,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004777935864364383,
      "loss": 5.5068,
      "mean_token_accuracy": 0.1819259151816368,
      "num_tokens": 30243479.0,
      "step": 13555
    },
    {
      "entropy": 5.6132118701934814,
      "epoch": 1.5229404166900657,
      "grad_norm": 1.5,
      "learning_rate": 0.000477776191368407,
      "loss": 5.4383,
      "mean_token_accuracy": 0.1828339710831642,
      "num_tokens": 30254721.0,
      "step": 13560
    },
    {
      "entropy": 5.567163944244385,
      "epoch": 1.5235019935980234,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004777587898439031,
      "loss": 5.3443,
      "mean_token_accuracy": 0.18657080829143524,
      "num_tokens": 30266962.0,
      "step": 13565
    },
    {
      "entropy": 5.559728813171387,
      "epoch": 1.5240635705059808,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047774138186348073,
      "loss": 5.3792,
      "mean_token_accuracy": 0.18521272838115693,
      "num_tokens": 30278773.0,
      "step": 13570
    },
    {
      "entropy": 5.475287580490113,
      "epoch": 1.5246251474139383,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004777239674276941,
      "loss": 5.3051,
      "mean_token_accuracy": 0.1914447709918022,
      "num_tokens": 30289631.0,
      "step": 13575
    },
    {
      "entropy": 5.525226402282715,
      "epoch": 1.525186724321896,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004777065465370977,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1838540181517601,
      "num_tokens": 30301433.0,
      "step": 13580
    },
    {
      "entropy": 5.593246507644653,
      "epoch": 1.5257483012298534,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004776891191922463,
      "loss": 5.3571,
      "mean_token_accuracy": 0.18819249272346497,
      "num_tokens": 30312695.0,
      "step": 13585
    },
    {
      "entropy": 5.51133131980896,
      "epoch": 1.5263098781378108,
      "grad_norm": 1.5,
      "learning_rate": 0.0004776716853936947,
      "loss": 5.4017,
      "mean_token_accuracy": 0.18176380246877671,
      "num_tokens": 30324254.0,
      "step": 13590
    },
    {
      "entropy": 5.558914995193481,
      "epoch": 1.5268714550457685,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000477654245141998,
      "loss": 5.4717,
      "mean_token_accuracy": 0.18200935274362565,
      "num_tokens": 30335389.0,
      "step": 13595
    },
    {
      "entropy": 5.565616607666016,
      "epoch": 1.5274330319537262,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004776367984377116,
      "loss": 5.3476,
      "mean_token_accuracy": 0.19087966978549958,
      "num_tokens": 30346627.0,
      "step": 13600
    },
    {
      "entropy": 5.504670524597168,
      "epoch": 1.5279946088616836,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047761934528139086,
      "loss": 5.4383,
      "mean_token_accuracy": 0.1850270837545395,
      "num_tokens": 30358215.0,
      "step": 13605
    },
    {
      "entropy": 5.568783950805664,
      "epoch": 1.528556185769641,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004776018856735916,
      "loss": 5.3368,
      "mean_token_accuracy": 0.1884455904364586,
      "num_tokens": 30369793.0,
      "step": 13610
    },
    {
      "entropy": 5.527515125274658,
      "epoch": 1.5291177626775987,
      "grad_norm": 1.375,
      "learning_rate": 0.0004775844196148698,
      "loss": 5.2814,
      "mean_token_accuracy": 0.1938552275300026,
      "num_tokens": 30380200.0,
      "step": 13615
    },
    {
      "entropy": 5.462940740585327,
      "epoch": 1.5296793395855564,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047756694710578146,
      "loss": 5.4801,
      "mean_token_accuracy": 0.17716196030378342,
      "num_tokens": 30392282.0,
      "step": 13620
    },
    {
      "entropy": 5.697913360595703,
      "epoch": 1.5302409164935138,
      "grad_norm": 5.25,
      "learning_rate": 0.000477549468146883,
      "loss": 5.6244,
      "mean_token_accuracy": 0.17258382290601731,
      "num_tokens": 30405365.0,
      "step": 13625
    },
    {
      "entropy": 5.589715194702149,
      "epoch": 1.5308024934014712,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004775319827387309,
      "loss": 5.4209,
      "mean_token_accuracy": 0.1872163236141205,
      "num_tokens": 30416513.0,
      "step": 13630
    },
    {
      "entropy": 5.475887632369995,
      "epoch": 1.531364070309429,
      "grad_norm": 1.5,
      "learning_rate": 0.0004775144908818818,
      "loss": 5.2946,
      "mean_token_accuracy": 0.19170508086681365,
      "num_tokens": 30427245.0,
      "step": 13635
    },
    {
      "entropy": 5.621194076538086,
      "epoch": 1.5319256472173866,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047749699257689296,
      "loss": 5.4364,
      "mean_token_accuracy": 0.18304508030414582,
      "num_tokens": 30438437.0,
      "step": 13640
    },
    {
      "entropy": 5.528850936889649,
      "epoch": 1.532487224125344,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004774794878243212,
      "loss": 5.3897,
      "mean_token_accuracy": 0.18640411049127578,
      "num_tokens": 30449726.0,
      "step": 13645
    },
    {
      "entropy": 5.555571460723877,
      "epoch": 1.5330488010333014,
      "grad_norm": 1.296875,
      "learning_rate": 0.000477461976624724,
      "loss": 5.4298,
      "mean_token_accuracy": 0.18499019593000413,
      "num_tokens": 30460024.0,
      "step": 13650
    },
    {
      "entropy": 5.652355289459228,
      "epoch": 1.533610377941259,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047744445897865905,
      "loss": 5.5162,
      "mean_token_accuracy": 0.17776869386434554,
      "num_tokens": 30471981.0,
      "step": 13655
    },
    {
      "entropy": 5.560834503173828,
      "epoch": 1.5341719548492168,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000477426934886684,
      "loss": 5.3272,
      "mean_token_accuracy": 0.19395865648984909,
      "num_tokens": 30483143.0,
      "step": 13660
    },
    {
      "entropy": 5.465796852111817,
      "epoch": 1.5347335317571742,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047740940434935677,
      "loss": 5.3583,
      "mean_token_accuracy": 0.18184324204921723,
      "num_tokens": 30495049.0,
      "step": 13665
    },
    {
      "entropy": 5.548905324935913,
      "epoch": 1.5352951086651316,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004773918673672356,
      "loss": 5.4642,
      "mean_token_accuracy": 0.18481016606092454,
      "num_tokens": 30506736.0,
      "step": 13670
    },
    {
      "entropy": 5.741988182067871,
      "epoch": 1.5358566855730893,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047737432394087887,
      "loss": 5.5595,
      "mean_token_accuracy": 0.18076975047588348,
      "num_tokens": 30520130.0,
      "step": 13675
    },
    {
      "entropy": 5.625166940689087,
      "epoch": 1.5364182624810467,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047735677407084525,
      "loss": 5.5307,
      "mean_token_accuracy": 0.17583049535751344,
      "num_tokens": 30531480.0,
      "step": 13680
    },
    {
      "entropy": 5.514678812026977,
      "epoch": 1.5369798393890042,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004773392177576934,
      "loss": 5.3671,
      "mean_token_accuracy": 0.18170706778764725,
      "num_tokens": 30541760.0,
      "step": 13685
    },
    {
      "entropy": 5.593460512161255,
      "epoch": 1.5375414162969618,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004773216550019824,
      "loss": 5.382,
      "mean_token_accuracy": 0.18787881731987,
      "num_tokens": 30551162.0,
      "step": 13690
    },
    {
      "entropy": 5.531304979324341,
      "epoch": 1.5381029932049195,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047730408580427136,
      "loss": 5.3488,
      "mean_token_accuracy": 0.18768645823001862,
      "num_tokens": 30562601.0,
      "step": 13695
    },
    {
      "entropy": 5.5491913795471195,
      "epoch": 1.538664570112877,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047728651016511994,
      "loss": 5.4301,
      "mean_token_accuracy": 0.18237154334783554,
      "num_tokens": 30573923.0,
      "step": 13700
    },
    {
      "entropy": 5.629022026062012,
      "epoch": 1.5392261470208344,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047726892808508744,
      "loss": 5.4916,
      "mean_token_accuracy": 0.17828229665756226,
      "num_tokens": 30586400.0,
      "step": 13705
    },
    {
      "entropy": 5.633077812194824,
      "epoch": 1.539787723928792,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004772513395647339,
      "loss": 5.4615,
      "mean_token_accuracy": 0.17952902764081954,
      "num_tokens": 30598317.0,
      "step": 13710
    },
    {
      "entropy": 5.614277791976929,
      "epoch": 1.5403493008367497,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004772337446046192,
      "loss": 5.4242,
      "mean_token_accuracy": 0.18551861345767975,
      "num_tokens": 30609413.0,
      "step": 13715
    },
    {
      "entropy": 5.6229565143585205,
      "epoch": 1.5409108777447071,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047721614320530366,
      "loss": 5.3359,
      "mean_token_accuracy": 0.18386495858430862,
      "num_tokens": 30619252.0,
      "step": 13720
    },
    {
      "entropy": 5.507887458801269,
      "epoch": 1.5414724546526646,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047719853536734764,
      "loss": 5.3582,
      "mean_token_accuracy": 0.19007731080055237,
      "num_tokens": 30630477.0,
      "step": 13725
    },
    {
      "entropy": 5.527295351028442,
      "epoch": 1.5420340315606222,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004771809210913119,
      "loss": 5.4153,
      "mean_token_accuracy": 0.1849319592118263,
      "num_tokens": 30641977.0,
      "step": 13730
    },
    {
      "entropy": 5.61675181388855,
      "epoch": 1.54259560846858,
      "grad_norm": 1.9921875,
      "learning_rate": 0.00047716330037775724,
      "loss": 5.355,
      "mean_token_accuracy": 0.18888724744319915,
      "num_tokens": 30653547.0,
      "step": 13735
    },
    {
      "entropy": 5.5287590503692625,
      "epoch": 1.5431571853765373,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004771456732272446,
      "loss": 5.3886,
      "mean_token_accuracy": 0.18013912588357925,
      "num_tokens": 30664270.0,
      "step": 13740
    },
    {
      "entropy": 5.659532165527343,
      "epoch": 1.5437187622844948,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047712803964033545,
      "loss": 5.5812,
      "mean_token_accuracy": 0.17014667391777039,
      "num_tokens": 30677172.0,
      "step": 13745
    },
    {
      "entropy": 5.616595029830933,
      "epoch": 1.5442803391924524,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047711039961759107,
      "loss": 5.448,
      "mean_token_accuracy": 0.18428623527288437,
      "num_tokens": 30688657.0,
      "step": 13750
    },
    {
      "entropy": 5.585419130325318,
      "epoch": 1.54484191610041,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004770927531595731,
      "loss": 5.4416,
      "mean_token_accuracy": 0.18118552267551422,
      "num_tokens": 30699103.0,
      "step": 13755
    },
    {
      "entropy": 5.558484268188477,
      "epoch": 1.5454034930083675,
      "grad_norm": 1.125,
      "learning_rate": 0.00047707510026684353,
      "loss": 5.3876,
      "mean_token_accuracy": 0.18183284252882004,
      "num_tokens": 30711352.0,
      "step": 13760
    },
    {
      "entropy": 5.675160837173462,
      "epoch": 1.545965069916325,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047705744093996447,
      "loss": 5.4765,
      "mean_token_accuracy": 0.18024851083755494,
      "num_tokens": 30722224.0,
      "step": 13765
    },
    {
      "entropy": 5.607077360153198,
      "epoch": 1.5465266468242826,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000477039775179498,
      "loss": 5.4422,
      "mean_token_accuracy": 0.17536765933036805,
      "num_tokens": 30733113.0,
      "step": 13770
    },
    {
      "entropy": 5.573582410812378,
      "epoch": 1.54708822373224,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047702210298600684,
      "loss": 5.4511,
      "mean_token_accuracy": 0.1805928409099579,
      "num_tokens": 30744903.0,
      "step": 13775
    },
    {
      "entropy": 5.633671283721924,
      "epoch": 1.5476498006401975,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047700442436005356,
      "loss": 5.4317,
      "mean_token_accuracy": 0.18018018305301667,
      "num_tokens": 30756034.0,
      "step": 13780
    },
    {
      "entropy": 5.542671060562133,
      "epoch": 1.5482113775481552,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000476986739302201,
      "loss": 5.3462,
      "mean_token_accuracy": 0.18649798780679702,
      "num_tokens": 30766139.0,
      "step": 13785
    },
    {
      "entropy": 5.567276573181152,
      "epoch": 1.5487729544561128,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047696904781301233,
      "loss": 5.4416,
      "mean_token_accuracy": 0.18002934902906417,
      "num_tokens": 30777505.0,
      "step": 13790
    },
    {
      "entropy": 5.5490563869476315,
      "epoch": 1.5493345313640703,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004769513498930508,
      "loss": 5.3241,
      "mean_token_accuracy": 0.18992604166269303,
      "num_tokens": 30788520.0,
      "step": 13795
    },
    {
      "entropy": 5.523218870162964,
      "epoch": 1.5498961082720277,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047693364554288,
      "loss": 5.3192,
      "mean_token_accuracy": 0.1855691209435463,
      "num_tokens": 30799051.0,
      "step": 13800
    },
    {
      "entropy": 5.521050930023193,
      "epoch": 1.5504576851799854,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047691593476306373,
      "loss": 5.336,
      "mean_token_accuracy": 0.19566950649023057,
      "num_tokens": 30809375.0,
      "step": 13805
    },
    {
      "entropy": 5.539036464691162,
      "epoch": 1.551019262087943,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004768982175541656,
      "loss": 5.3418,
      "mean_token_accuracy": 0.18851928114891053,
      "num_tokens": 30821330.0,
      "step": 13810
    },
    {
      "entropy": 5.633591508865356,
      "epoch": 1.5515808389959005,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047688049391675006,
      "loss": 5.4827,
      "mean_token_accuracy": 0.17657239884138107,
      "num_tokens": 30832366.0,
      "step": 13815
    },
    {
      "entropy": 5.594413089752197,
      "epoch": 1.552142415903858,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004768627638513813,
      "loss": 5.4264,
      "mean_token_accuracy": 0.17895372062921525,
      "num_tokens": 30843550.0,
      "step": 13820
    },
    {
      "entropy": 5.482576704025268,
      "epoch": 1.5527039928118156,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004768450273586237,
      "loss": 5.258,
      "mean_token_accuracy": 0.1935638576745987,
      "num_tokens": 30854727.0,
      "step": 13825
    },
    {
      "entropy": 5.589857721328736,
      "epoch": 1.5532655697197733,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047682728443904223,
      "loss": 5.3835,
      "mean_token_accuracy": 0.18434152752161026,
      "num_tokens": 30865594.0,
      "step": 13830
    },
    {
      "entropy": 5.580872392654419,
      "epoch": 1.5538271466277307,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047680953509320176,
      "loss": 5.3871,
      "mean_token_accuracy": 0.18392160683870315,
      "num_tokens": 30877386.0,
      "step": 13835
    },
    {
      "entropy": 5.657144498825073,
      "epoch": 1.5543887235356881,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004767917793216674,
      "loss": 5.4923,
      "mean_token_accuracy": 0.1839445412158966,
      "num_tokens": 30889052.0,
      "step": 13840
    },
    {
      "entropy": 5.562966823577881,
      "epoch": 1.5549503004436458,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047677401712500455,
      "loss": 5.3473,
      "mean_token_accuracy": 0.18103903234004975,
      "num_tokens": 30900524.0,
      "step": 13845
    },
    {
      "entropy": 5.576873540878296,
      "epoch": 1.5555118773516035,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004767562485037787,
      "loss": 5.4734,
      "mean_token_accuracy": 0.17559579908847808,
      "num_tokens": 30911770.0,
      "step": 13850
    },
    {
      "entropy": 5.4913770198822025,
      "epoch": 1.556073454259561,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047673847345855566,
      "loss": 5.2639,
      "mean_token_accuracy": 0.19188567548990249,
      "num_tokens": 30921662.0,
      "step": 13855
    },
    {
      "entropy": 5.6478245735168455,
      "epoch": 1.5566350311675183,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004767206919899014,
      "loss": 5.5509,
      "mean_token_accuracy": 0.18133361339569093,
      "num_tokens": 30934138.0,
      "step": 13860
    },
    {
      "entropy": 5.631950092315674,
      "epoch": 1.557196608075476,
      "grad_norm": 1.375,
      "learning_rate": 0.000476702904098382,
      "loss": 5.3804,
      "mean_token_accuracy": 0.1829290822148323,
      "num_tokens": 30945561.0,
      "step": 13865
    },
    {
      "entropy": 5.502339935302734,
      "epoch": 1.5577581849834334,
      "grad_norm": 1.25,
      "learning_rate": 0.000476685109784564,
      "loss": 5.2657,
      "mean_token_accuracy": 0.19422535449266434,
      "num_tokens": 30957210.0,
      "step": 13870
    },
    {
      "entropy": 5.546827745437622,
      "epoch": 1.5583197618913909,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047666730904901377,
      "loss": 5.3774,
      "mean_token_accuracy": 0.18659260720014573,
      "num_tokens": 30968782.0,
      "step": 13875
    },
    {
      "entropy": 5.553039169311523,
      "epoch": 1.5588813387993485,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047664950189229833,
      "loss": 5.42,
      "mean_token_accuracy": 0.18063327074050903,
      "num_tokens": 30981239.0,
      "step": 13880
    },
    {
      "entropy": 5.620796632766724,
      "epoch": 1.5594429157073062,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004766316883149845,
      "loss": 5.4496,
      "mean_token_accuracy": 0.18674004524946214,
      "num_tokens": 30991773.0,
      "step": 13885
    },
    {
      "entropy": 5.666515827178955,
      "epoch": 1.5600044926152636,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047661386831763955,
      "loss": 5.4382,
      "mean_token_accuracy": 0.1887378215789795,
      "num_tokens": 31002646.0,
      "step": 13890
    },
    {
      "entropy": 5.505196619033813,
      "epoch": 1.560566069523221,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047659604190083076,
      "loss": 5.4008,
      "mean_token_accuracy": 0.1821158915758133,
      "num_tokens": 31014405.0,
      "step": 13895
    },
    {
      "entropy": 5.5668925762176515,
      "epoch": 1.5611276464311787,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047657820906512585,
      "loss": 5.3414,
      "mean_token_accuracy": 0.18443161994218826,
      "num_tokens": 31025529.0,
      "step": 13900
    },
    {
      "entropy": 5.684357118606568,
      "epoch": 1.5616892233391364,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047656036981109256,
      "loss": 5.487,
      "mean_token_accuracy": 0.17803749591112136,
      "num_tokens": 31037558.0,
      "step": 13905
    },
    {
      "entropy": 5.430062532424927,
      "epoch": 1.5622508002470938,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047654252413929896,
      "loss": 5.2432,
      "mean_token_accuracy": 0.19305500388145447,
      "num_tokens": 31048918.0,
      "step": 13910
    },
    {
      "entropy": 5.454860258102417,
      "epoch": 1.5628123771550513,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047652467205031317,
      "loss": 5.3506,
      "mean_token_accuracy": 0.1870207667350769,
      "num_tokens": 31060149.0,
      "step": 13915
    },
    {
      "entropy": 5.548892307281494,
      "epoch": 1.563373954063009,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047650681354470363,
      "loss": 5.3699,
      "mean_token_accuracy": 0.18815428167581558,
      "num_tokens": 31071197.0,
      "step": 13920
    },
    {
      "entropy": 5.613787508010864,
      "epoch": 1.5639355309709666,
      "grad_norm": 1.375,
      "learning_rate": 0.00047648894862303907,
      "loss": 5.4866,
      "mean_token_accuracy": 0.18950826525688172,
      "num_tokens": 31083385.0,
      "step": 13925
    },
    {
      "entropy": 5.54557957649231,
      "epoch": 1.564497107878924,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047647107728588826,
      "loss": 5.39,
      "mean_token_accuracy": 0.18100087195634842,
      "num_tokens": 31094807.0,
      "step": 13930
    },
    {
      "entropy": 5.502994585037231,
      "epoch": 1.5650586847868815,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047645319953382024,
      "loss": 5.3025,
      "mean_token_accuracy": 0.19468299448490142,
      "num_tokens": 31105622.0,
      "step": 13935
    },
    {
      "entropy": 5.588700962066651,
      "epoch": 1.5656202616948391,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047643531536740405,
      "loss": 5.3365,
      "mean_token_accuracy": 0.18804485201835633,
      "num_tokens": 31115684.0,
      "step": 13940
    },
    {
      "entropy": 5.490893077850342,
      "epoch": 1.5661818386027968,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004764174247872094,
      "loss": 5.3033,
      "mean_token_accuracy": 0.19590633511543273,
      "num_tokens": 31126449.0,
      "step": 13945
    },
    {
      "entropy": 5.538293886184692,
      "epoch": 1.5667434155107542,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047639952779380586,
      "loss": 5.3283,
      "mean_token_accuracy": 0.19549835920333863,
      "num_tokens": 31136894.0,
      "step": 13950
    },
    {
      "entropy": 5.574567747116089,
      "epoch": 1.5673049924187117,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004763816243877632,
      "loss": 5.4235,
      "mean_token_accuracy": 0.18958259820938111,
      "num_tokens": 31147131.0,
      "step": 13955
    },
    {
      "entropy": 5.464069890975952,
      "epoch": 1.5678665693266693,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004763637145696515,
      "loss": 5.2629,
      "mean_token_accuracy": 0.19027214348316193,
      "num_tokens": 31157903.0,
      "step": 13960
    },
    {
      "entropy": 5.558705472946167,
      "epoch": 1.5684281462346268,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000476345798340041,
      "loss": 5.4313,
      "mean_token_accuracy": 0.1850793793797493,
      "num_tokens": 31169087.0,
      "step": 13965
    },
    {
      "entropy": 5.64756031036377,
      "epoch": 1.5689897231425842,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004763278756995022,
      "loss": 5.5032,
      "mean_token_accuracy": 0.17865439653396606,
      "num_tokens": 31180253.0,
      "step": 13970
    },
    {
      "entropy": 5.640758609771728,
      "epoch": 1.5695513000505419,
      "grad_norm": 1.640625,
      "learning_rate": 0.00047630994664860576,
      "loss": 5.4382,
      "mean_token_accuracy": 0.18289659172296524,
      "num_tokens": 31191857.0,
      "step": 13975
    },
    {
      "entropy": 5.55273962020874,
      "epoch": 1.5701128769584995,
      "grad_norm": 1.25,
      "learning_rate": 0.0004762920111879226,
      "loss": 5.4536,
      "mean_token_accuracy": 0.1757165178656578,
      "num_tokens": 31201868.0,
      "step": 13980
    },
    {
      "entropy": 5.563938236236572,
      "epoch": 1.570674453866457,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004762740693180237,
      "loss": 5.4015,
      "mean_token_accuracy": 0.18728701025247574,
      "num_tokens": 31212716.0,
      "step": 13985
    },
    {
      "entropy": 5.596206521987915,
      "epoch": 1.5712360307744144,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047625612103948035,
      "loss": 5.474,
      "mean_token_accuracy": 0.17948551028966903,
      "num_tokens": 31224823.0,
      "step": 13990
    },
    {
      "entropy": 5.570262765884399,
      "epoch": 1.571797607682372,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004762381663528641,
      "loss": 5.3656,
      "mean_token_accuracy": 0.1848146378993988,
      "num_tokens": 31236798.0,
      "step": 13995
    },
    {
      "entropy": 5.609161281585694,
      "epoch": 1.5723591845903298,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004762202052587466,
      "loss": 5.3998,
      "mean_token_accuracy": 0.1816373810172081,
      "num_tokens": 31248687.0,
      "step": 14000
    },
    {
      "entropy": 5.572255611419678,
      "epoch": 1.5729207614982872,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004762022377576996,
      "loss": 5.4599,
      "mean_token_accuracy": 0.18094970881938935,
      "num_tokens": 31258385.0,
      "step": 14005
    },
    {
      "entropy": 5.676636552810669,
      "epoch": 1.5734823384062446,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004761842638502954,
      "loss": 5.4744,
      "mean_token_accuracy": 0.1854086771607399,
      "num_tokens": 31269308.0,
      "step": 14010
    },
    {
      "entropy": 5.608185768127441,
      "epoch": 1.5740439153142023,
      "grad_norm": 1.25,
      "learning_rate": 0.0004761662835371062,
      "loss": 5.3744,
      "mean_token_accuracy": 0.18948292285203933,
      "num_tokens": 31279900.0,
      "step": 14015
    },
    {
      "entropy": 5.543855428695679,
      "epoch": 1.57460549222216,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004761482968187045,
      "loss": 5.3849,
      "mean_token_accuracy": 0.1922425866127014,
      "num_tokens": 31290456.0,
      "step": 14020
    },
    {
      "entropy": 5.491939544677734,
      "epoch": 1.5751670691301174,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004761303036956631,
      "loss": 5.3621,
      "mean_token_accuracy": 0.18260688036680223,
      "num_tokens": 31301398.0,
      "step": 14025
    },
    {
      "entropy": 5.658075141906738,
      "epoch": 1.5757286460380748,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004761123041685547,
      "loss": 5.4558,
      "mean_token_accuracy": 0.1853075996041298,
      "num_tokens": 31311149.0,
      "step": 14030
    },
    {
      "entropy": 5.464800119400024,
      "epoch": 1.5762902229460325,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004760942982379526,
      "loss": 5.2795,
      "mean_token_accuracy": 0.1985880345106125,
      "num_tokens": 31321099.0,
      "step": 14035
    },
    {
      "entropy": 5.521101665496826,
      "epoch": 1.5768517998539902,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047607628590443004,
      "loss": 5.413,
      "mean_token_accuracy": 0.1805889293551445,
      "num_tokens": 31331872.0,
      "step": 14040
    },
    {
      "entropy": 5.685874986648559,
      "epoch": 1.5774133767619476,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047605826716856056,
      "loss": 5.4529,
      "mean_token_accuracy": 0.1793512836098671,
      "num_tokens": 31342024.0,
      "step": 14045
    },
    {
      "entropy": 5.614808082580566,
      "epoch": 1.577974953669905,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004760402420309178,
      "loss": 5.2957,
      "mean_token_accuracy": 0.1913559079170227,
      "num_tokens": 31351857.0,
      "step": 14050
    },
    {
      "entropy": 5.503947019577026,
      "epoch": 1.5785365305778627,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004760222104920759,
      "loss": 5.3928,
      "mean_token_accuracy": 0.18860056549310683,
      "num_tokens": 31362609.0,
      "step": 14055
    },
    {
      "entropy": 5.478338956832886,
      "epoch": 1.5790981074858201,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004760041725526088,
      "loss": 5.3232,
      "mean_token_accuracy": 0.18444042205810546,
      "num_tokens": 31373129.0,
      "step": 14060
    },
    {
      "entropy": 5.550362205505371,
      "epoch": 1.5796596843937776,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047598612821309077,
      "loss": 5.2925,
      "mean_token_accuracy": 0.1911611244082451,
      "num_tokens": 31383109.0,
      "step": 14065
    },
    {
      "entropy": 5.484701251983642,
      "epoch": 1.5802212613017352,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004759680774740966,
      "loss": 5.2678,
      "mean_token_accuracy": 0.18637271523475646,
      "num_tokens": 31393603.0,
      "step": 14070
    },
    {
      "entropy": 5.60389404296875,
      "epoch": 1.580782838209693,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004759500203362008,
      "loss": 5.3491,
      "mean_token_accuracy": 0.18680666238069535,
      "num_tokens": 31404011.0,
      "step": 14075
    },
    {
      "entropy": 5.502894592285156,
      "epoch": 1.5813444151176503,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004759319567999785,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18432048857212066,
      "num_tokens": 31414239.0,
      "step": 14080
    },
    {
      "entropy": 5.53234281539917,
      "epoch": 1.5819059920256078,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004759138868660048,
      "loss": 5.4073,
      "mean_token_accuracy": 0.18384866714477538,
      "num_tokens": 31425055.0,
      "step": 14085
    },
    {
      "entropy": 5.602566194534302,
      "epoch": 1.5824675689335654,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047589581053485494,
      "loss": 5.3752,
      "mean_token_accuracy": 0.18901110738515853,
      "num_tokens": 31435537.0,
      "step": 14090
    },
    {
      "entropy": 5.62176251411438,
      "epoch": 1.583029145841523,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047587772780710455,
      "loss": 5.4438,
      "mean_token_accuracy": 0.18455122858285905,
      "num_tokens": 31446056.0,
      "step": 14095
    },
    {
      "entropy": 5.594608354568481,
      "epoch": 1.5835907227494805,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047585963868332945,
      "loss": 5.4112,
      "mean_token_accuracy": 0.18492064177989959,
      "num_tokens": 31457452.0,
      "step": 14100
    },
    {
      "entropy": 5.514397048950196,
      "epoch": 1.584152299657438,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004758415431641055,
      "loss": 5.359,
      "mean_token_accuracy": 0.1876969188451767,
      "num_tokens": 31469038.0,
      "step": 14105
    },
    {
      "entropy": 5.5730689525604244,
      "epoch": 1.5847138765653956,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004758234412500089,
      "loss": 5.4638,
      "mean_token_accuracy": 0.1851399928331375,
      "num_tokens": 31481069.0,
      "step": 14110
    },
    {
      "entropy": 5.519595813751221,
      "epoch": 1.5852754534733533,
      "grad_norm": 1.5,
      "learning_rate": 0.00047580533294161616,
      "loss": 5.4108,
      "mean_token_accuracy": 0.18213094621896744,
      "num_tokens": 31491393.0,
      "step": 14115
    },
    {
      "entropy": 5.528887176513672,
      "epoch": 1.5858370303813107,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004757872182395036,
      "loss": 5.4077,
      "mean_token_accuracy": 0.18127547353506088,
      "num_tokens": 31502627.0,
      "step": 14120
    },
    {
      "entropy": 5.699723100662231,
      "epoch": 1.5863986072892682,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047576909714424823,
      "loss": 5.4438,
      "mean_token_accuracy": 0.17827050983905793,
      "num_tokens": 31513995.0,
      "step": 14125
    },
    {
      "entropy": 5.525648593902588,
      "epoch": 1.5869601841972258,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047575096965642687,
      "loss": 5.3752,
      "mean_token_accuracy": 0.18316685557365417,
      "num_tokens": 31524625.0,
      "step": 14130
    },
    {
      "entropy": 5.481736183166504,
      "epoch": 1.5875217611051835,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047573283577661684,
      "loss": 5.3392,
      "mean_token_accuracy": 0.18936698287725448,
      "num_tokens": 31535940.0,
      "step": 14135
    },
    {
      "entropy": 5.566658544540405,
      "epoch": 1.588083338013141,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004757146955053955,
      "loss": 5.3906,
      "mean_token_accuracy": 0.1862411007285118,
      "num_tokens": 31546557.0,
      "step": 14140
    },
    {
      "entropy": 5.578667545318604,
      "epoch": 1.5886449149210984,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047569654884334036,
      "loss": 5.4245,
      "mean_token_accuracy": 0.17833538800477983,
      "num_tokens": 31559335.0,
      "step": 14145
    },
    {
      "entropy": 5.6002030849456785,
      "epoch": 1.589206491829056,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004756783957910292,
      "loss": 5.3658,
      "mean_token_accuracy": 0.18090360909700393,
      "num_tokens": 31570698.0,
      "step": 14150
    },
    {
      "entropy": 5.581079721450806,
      "epoch": 1.5897680687370135,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047566023634904015,
      "loss": 5.4318,
      "mean_token_accuracy": 0.18585970848798752,
      "num_tokens": 31582524.0,
      "step": 14155
    },
    {
      "entropy": 5.575134134292602,
      "epoch": 1.590329645644971,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004756420705179513,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18733209520578384,
      "num_tokens": 31593399.0,
      "step": 14160
    },
    {
      "entropy": 5.506226110458374,
      "epoch": 1.5908912225529286,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004756238982983412,
      "loss": 5.3904,
      "mean_token_accuracy": 0.17625253051519393,
      "num_tokens": 31604169.0,
      "step": 14165
    },
    {
      "entropy": 5.500355672836304,
      "epoch": 1.5914527994608862,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004756057196907883,
      "loss": 5.3312,
      "mean_token_accuracy": 0.19254590421915055,
      "num_tokens": 31614671.0,
      "step": 14170
    },
    {
      "entropy": 5.490681123733521,
      "epoch": 1.5920143763688437,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047558753469587154,
      "loss": 5.3185,
      "mean_token_accuracy": 0.19051374346017838,
      "num_tokens": 31625793.0,
      "step": 14175
    },
    {
      "entropy": 5.542859077453613,
      "epoch": 1.5925759532768011,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047556934331416986,
      "loss": 5.3358,
      "mean_token_accuracy": 0.19105246514081956,
      "num_tokens": 31637632.0,
      "step": 14180
    },
    {
      "entropy": 5.545540714263916,
      "epoch": 1.5931375301847588,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004755511455462624,
      "loss": 5.3291,
      "mean_token_accuracy": 0.19029556959867477,
      "num_tokens": 31648314.0,
      "step": 14185
    },
    {
      "entropy": 5.577873563766479,
      "epoch": 1.5936991070927164,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004755329413927287,
      "loss": 5.3908,
      "mean_token_accuracy": 0.18803092539310456,
      "num_tokens": 31659152.0,
      "step": 14190
    },
    {
      "entropy": 5.5810370445251465,
      "epoch": 1.5942606840006739,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004755147308541485,
      "loss": 5.3344,
      "mean_token_accuracy": 0.19057308733463288,
      "num_tokens": 31669837.0,
      "step": 14195
    },
    {
      "entropy": 5.511269760131836,
      "epoch": 1.5948222609086313,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047549651393110145,
      "loss": 5.3664,
      "mean_token_accuracy": 0.18276496827602387,
      "num_tokens": 31680663.0,
      "step": 14200
    },
    {
      "entropy": 5.553107500076294,
      "epoch": 1.595383837816589,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047547829062416757,
      "loss": 5.3588,
      "mean_token_accuracy": 0.18153503388166428,
      "num_tokens": 31691484.0,
      "step": 14205
    },
    {
      "entropy": 5.55496916770935,
      "epoch": 1.5959454147245467,
      "grad_norm": 1.25,
      "learning_rate": 0.00047546006093392726,
      "loss": 5.31,
      "mean_token_accuracy": 0.18103923350572587,
      "num_tokens": 31701642.0,
      "step": 14210
    },
    {
      "entropy": 5.492617607116699,
      "epoch": 1.596506991632504,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004754418248609608,
      "loss": 5.3719,
      "mean_token_accuracy": 0.1854472577571869,
      "num_tokens": 31714203.0,
      "step": 14215
    },
    {
      "entropy": 5.548641633987427,
      "epoch": 1.5970685685404615,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004754235824058489,
      "loss": 5.3546,
      "mean_token_accuracy": 0.18786930292844772,
      "num_tokens": 31725890.0,
      "step": 14220
    },
    {
      "entropy": 5.523428344726563,
      "epoch": 1.5976301454484192,
      "grad_norm": 1.625,
      "learning_rate": 0.0004754053335691724,
      "loss": 5.3097,
      "mean_token_accuracy": 0.18873031437397003,
      "num_tokens": 31738058.0,
      "step": 14225
    },
    {
      "entropy": 5.5686602115631105,
      "epoch": 1.5981917223563769,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047538707835151237,
      "loss": 5.3769,
      "mean_token_accuracy": 0.19276683628559113,
      "num_tokens": 31748983.0,
      "step": 14230
    },
    {
      "entropy": 5.562123250961304,
      "epoch": 1.5987532992643343,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047536881675345,
      "loss": 5.3753,
      "mean_token_accuracy": 0.18837305456399916,
      "num_tokens": 31760083.0,
      "step": 14235
    },
    {
      "entropy": 5.50814642906189,
      "epoch": 1.5993148761722917,
      "grad_norm": 1.5,
      "learning_rate": 0.0004753505487755669,
      "loss": 5.3182,
      "mean_token_accuracy": 0.19211919456720353,
      "num_tokens": 31771627.0,
      "step": 14240
    },
    {
      "entropy": 5.563792991638183,
      "epoch": 1.5998764530802494,
      "grad_norm": 1.375,
      "learning_rate": 0.00047533227441844453,
      "loss": 5.4112,
      "mean_token_accuracy": 0.19014624506235123,
      "num_tokens": 31783476.0,
      "step": 14245
    },
    {
      "entropy": 5.61142635345459,
      "epoch": 1.6004380299882068,
      "grad_norm": 1.765625,
      "learning_rate": 0.00047531399368266485,
      "loss": 5.4226,
      "mean_token_accuracy": 0.18333071917295457,
      "num_tokens": 31794072.0,
      "step": 14250
    },
    {
      "entropy": 5.547273302078247,
      "epoch": 1.6009996068961643,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004752957065688099,
      "loss": 5.3179,
      "mean_token_accuracy": 0.19077218621969222,
      "num_tokens": 31803613.0,
      "step": 14255
    },
    {
      "entropy": 5.63131275177002,
      "epoch": 1.601561183804122,
      "grad_norm": 1.515625,
      "learning_rate": 0.000475277413077462,
      "loss": 5.4515,
      "mean_token_accuracy": 0.18376434296369554,
      "num_tokens": 31815018.0,
      "step": 14260
    },
    {
      "entropy": 5.41245551109314,
      "epoch": 1.6021227607120796,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047525911320920363,
      "loss": 5.288,
      "mean_token_accuracy": 0.19167415499687196,
      "num_tokens": 31825603.0,
      "step": 14265
    },
    {
      "entropy": 5.483716344833374,
      "epoch": 1.602684337620037,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004752408069646174,
      "loss": 5.3519,
      "mean_token_accuracy": 0.18703959584236146,
      "num_tokens": 31835988.0,
      "step": 14270
    },
    {
      "entropy": 5.582781457901001,
      "epoch": 1.6032459145279945,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047522249434428623,
      "loss": 5.3838,
      "mean_token_accuracy": 0.1827508196234703,
      "num_tokens": 31846378.0,
      "step": 14275
    },
    {
      "entropy": 5.6581432819366455,
      "epoch": 1.6038074914359521,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004752041753487931,
      "loss": 5.4543,
      "mean_token_accuracy": 0.17981308996677398,
      "num_tokens": 31858836.0,
      "step": 14280
    },
    {
      "entropy": 5.505344343185425,
      "epoch": 1.6043690683439098,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004751858499787215,
      "loss": 5.3224,
      "mean_token_accuracy": 0.18830745071172714,
      "num_tokens": 31869315.0,
      "step": 14285
    },
    {
      "entropy": 5.605330228805542,
      "epoch": 1.6049306452518672,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004751675182346547,
      "loss": 5.5195,
      "mean_token_accuracy": 0.17846543937921525,
      "num_tokens": 31880527.0,
      "step": 14290
    },
    {
      "entropy": 5.636462545394897,
      "epoch": 1.6054922221598247,
      "grad_norm": 1.765625,
      "learning_rate": 0.00047514918011717656,
      "loss": 5.3828,
      "mean_token_accuracy": 0.18812797218561172,
      "num_tokens": 31891890.0,
      "step": 14295
    },
    {
      "entropy": 5.603710079193116,
      "epoch": 1.6060537990677823,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047513083562687085,
      "loss": 5.3857,
      "mean_token_accuracy": 0.18741926103830336,
      "num_tokens": 31902735.0,
      "step": 14300
    },
    {
      "entropy": 5.541609287261963,
      "epoch": 1.60661537597574,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004751124847643217,
      "loss": 5.437,
      "mean_token_accuracy": 0.1817302241921425,
      "num_tokens": 31914504.0,
      "step": 14305
    },
    {
      "entropy": 5.598568391799927,
      "epoch": 1.6071769528836974,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004750941275301135,
      "loss": 5.3956,
      "mean_token_accuracy": 0.18769104033708572,
      "num_tokens": 31925287.0,
      "step": 14310
    },
    {
      "entropy": 5.546432113647461,
      "epoch": 1.6077385297916549,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004750757639248306,
      "loss": 5.3295,
      "mean_token_accuracy": 0.19074969589710236,
      "num_tokens": 31935845.0,
      "step": 14315
    },
    {
      "entropy": 5.5763781547546385,
      "epoch": 1.6083001066996125,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047505739394905776,
      "loss": 5.415,
      "mean_token_accuracy": 0.18895251154899598,
      "num_tokens": 31946771.0,
      "step": 14320
    },
    {
      "entropy": 5.539247798919678,
      "epoch": 1.6088616836075702,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047503901760338003,
      "loss": 5.3768,
      "mean_token_accuracy": 0.18517448455095292,
      "num_tokens": 31957733.0,
      "step": 14325
    },
    {
      "entropy": 5.55397400856018,
      "epoch": 1.6094232605155276,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047502063488838225,
      "loss": 5.2744,
      "mean_token_accuracy": 0.19239411056041716,
      "num_tokens": 31967809.0,
      "step": 14330
    },
    {
      "entropy": 5.42648458480835,
      "epoch": 1.609984837423485,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047500224580464994,
      "loss": 5.2446,
      "mean_token_accuracy": 0.1935966953635216,
      "num_tokens": 31978673.0,
      "step": 14335
    },
    {
      "entropy": 5.491064739227295,
      "epoch": 1.6105464143314427,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047498385035276856,
      "loss": 5.3765,
      "mean_token_accuracy": 0.18715526312589645,
      "num_tokens": 31990535.0,
      "step": 14340
    },
    {
      "entropy": 5.580092477798462,
      "epoch": 1.6111079912394002,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004749654485333238,
      "loss": 5.4287,
      "mean_token_accuracy": 0.1823533996939659,
      "num_tokens": 32002164.0,
      "step": 14345
    },
    {
      "entropy": 5.6803051948547365,
      "epoch": 1.6116695681473576,
      "grad_norm": 2.109375,
      "learning_rate": 0.00047494704034690163,
      "loss": 5.4777,
      "mean_token_accuracy": 0.17761383950710297,
      "num_tokens": 32013651.0,
      "step": 14350
    },
    {
      "entropy": 5.604819202423096,
      "epoch": 1.6122311450553153,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047492862579408816,
      "loss": 5.3738,
      "mean_token_accuracy": 0.18794151097536088,
      "num_tokens": 32025128.0,
      "step": 14355
    },
    {
      "entropy": 5.550415658950806,
      "epoch": 1.612792721963273,
      "grad_norm": 1.828125,
      "learning_rate": 0.00047491020487546966,
      "loss": 5.407,
      "mean_token_accuracy": 0.17649219334125518,
      "num_tokens": 32036359.0,
      "step": 14360
    },
    {
      "entropy": 5.558035135269165,
      "epoch": 1.6133542988712304,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004748917775916327,
      "loss": 5.3206,
      "mean_token_accuracy": 0.18939928710460663,
      "num_tokens": 32047803.0,
      "step": 14365
    },
    {
      "entropy": 5.541605043411255,
      "epoch": 1.6139158757791878,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047487334394316405,
      "loss": 5.4889,
      "mean_token_accuracy": 0.183787302672863,
      "num_tokens": 32060156.0,
      "step": 14370
    },
    {
      "entropy": 5.596155071258545,
      "epoch": 1.6144774526871455,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004748549039306506,
      "loss": 5.3777,
      "mean_token_accuracy": 0.18946214765310287,
      "num_tokens": 32071398.0,
      "step": 14375
    },
    {
      "entropy": 5.495157909393311,
      "epoch": 1.6150390295951031,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004748364575546795,
      "loss": 5.2052,
      "mean_token_accuracy": 0.1963065445423126,
      "num_tokens": 32082472.0,
      "step": 14380
    },
    {
      "entropy": 5.486556148529052,
      "epoch": 1.6156006065030606,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047481800481583806,
      "loss": 5.3856,
      "mean_token_accuracy": 0.18367727398872374,
      "num_tokens": 32093103.0,
      "step": 14385
    },
    {
      "entropy": 5.569084644317627,
      "epoch": 1.616162183411018,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004747995457147139,
      "loss": 5.3978,
      "mean_token_accuracy": 0.1878577098250389,
      "num_tokens": 32104526.0,
      "step": 14390
    },
    {
      "entropy": 5.643805456161499,
      "epoch": 1.6167237603189757,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047478108025189467,
      "loss": 5.4471,
      "mean_token_accuracy": 0.1791172981262207,
      "num_tokens": 32116018.0,
      "step": 14395
    },
    {
      "entropy": 5.530964612960815,
      "epoch": 1.6172853372269334,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047476260842796837,
      "loss": 5.2966,
      "mean_token_accuracy": 0.19313062429428102,
      "num_tokens": 32127385.0,
      "step": 14400
    },
    {
      "entropy": 5.568222522735596,
      "epoch": 1.6178469141348908,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004747441302435231,
      "loss": 5.4237,
      "mean_token_accuracy": 0.18586568534374237,
      "num_tokens": 32137376.0,
      "step": 14405
    },
    {
      "entropy": 5.570278549194336,
      "epoch": 1.6184084910428482,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047472564569914726,
      "loss": 5.4094,
      "mean_token_accuracy": 0.18463828414678574,
      "num_tokens": 32148263.0,
      "step": 14410
    },
    {
      "entropy": 5.5898205757141115,
      "epoch": 1.618970067950806,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047470715479542945,
      "loss": 5.3361,
      "mean_token_accuracy": 0.18114308267831802,
      "num_tokens": 32161323.0,
      "step": 14415
    },
    {
      "entropy": 5.596181488037109,
      "epoch": 1.6195316448587636,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004746886575329584,
      "loss": 5.398,
      "mean_token_accuracy": 0.17863700091838836,
      "num_tokens": 32172294.0,
      "step": 14420
    },
    {
      "entropy": 5.478462219238281,
      "epoch": 1.620093221766721,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047467015391232305,
      "loss": 5.3537,
      "mean_token_accuracy": 0.17806696742773057,
      "num_tokens": 32182619.0,
      "step": 14425
    },
    {
      "entropy": 5.515846109390258,
      "epoch": 1.6206547986746784,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047465164393411245,
      "loss": 5.3377,
      "mean_token_accuracy": 0.18611229509115218,
      "num_tokens": 32193078.0,
      "step": 14430
    },
    {
      "entropy": 5.565969944000244,
      "epoch": 1.621216375582636,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00047463312759891615,
      "loss": 5.3396,
      "mean_token_accuracy": 0.18058524578809737,
      "num_tokens": 32204134.0,
      "step": 14435
    },
    {
      "entropy": 5.634787082672119,
      "epoch": 1.6217779524905935,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004746146049073236,
      "loss": 5.4189,
      "mean_token_accuracy": 0.1862504377961159,
      "num_tokens": 32215136.0,
      "step": 14440
    },
    {
      "entropy": 5.527179718017578,
      "epoch": 1.622339529398551,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047459607585992466,
      "loss": 5.4068,
      "mean_token_accuracy": 0.18511534482240677,
      "num_tokens": 32225727.0,
      "step": 14445
    },
    {
      "entropy": 5.558506059646606,
      "epoch": 1.6229011063065086,
      "grad_norm": 1.625,
      "learning_rate": 0.00047457754045730923,
      "loss": 5.3752,
      "mean_token_accuracy": 0.1840720921754837,
      "num_tokens": 32236376.0,
      "step": 14450
    },
    {
      "entropy": 5.51150336265564,
      "epoch": 1.6234626832144663,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047455899870006745,
      "loss": 5.3122,
      "mean_token_accuracy": 0.1918402850627899,
      "num_tokens": 32247202.0,
      "step": 14455
    },
    {
      "entropy": 5.516626405715942,
      "epoch": 1.6240242601224237,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004745404505887899,
      "loss": 5.4447,
      "mean_token_accuracy": 0.18239740282297134,
      "num_tokens": 32258625.0,
      "step": 14460
    },
    {
      "entropy": 5.598090410232544,
      "epoch": 1.6245858370303812,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004745218961240669,
      "loss": 5.3397,
      "mean_token_accuracy": 0.1882559359073639,
      "num_tokens": 32269288.0,
      "step": 14465
    },
    {
      "entropy": 5.580713844299316,
      "epoch": 1.6251474139383388,
      "grad_norm": 1.5,
      "learning_rate": 0.0004745033353064893,
      "loss": 5.3915,
      "mean_token_accuracy": 0.18425026386976243,
      "num_tokens": 32279888.0,
      "step": 14470
    },
    {
      "entropy": 5.590251159667969,
      "epoch": 1.6257089908462965,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004744847681366482,
      "loss": 5.3984,
      "mean_token_accuracy": 0.19061071127653123,
      "num_tokens": 32291153.0,
      "step": 14475
    },
    {
      "entropy": 5.473784732818603,
      "epoch": 1.626270567754254,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047446619461513466,
      "loss": 5.2835,
      "mean_token_accuracy": 0.18975742161273956,
      "num_tokens": 32302291.0,
      "step": 14480
    },
    {
      "entropy": 5.510013628005981,
      "epoch": 1.6268321446622114,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004744476147425402,
      "loss": 5.2811,
      "mean_token_accuracy": 0.19587058573961258,
      "num_tokens": 32312015.0,
      "step": 14485
    },
    {
      "entropy": 5.558365535736084,
      "epoch": 1.627393721570169,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047442902851945634,
      "loss": 5.4003,
      "mean_token_accuracy": 0.1832302913069725,
      "num_tokens": 32323967.0,
      "step": 14490
    },
    {
      "entropy": 5.5684449672698975,
      "epoch": 1.6279552984781267,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047441043594647486,
      "loss": 5.4173,
      "mean_token_accuracy": 0.17932213246822357,
      "num_tokens": 32336181.0,
      "step": 14495
    },
    {
      "entropy": 5.585696220397949,
      "epoch": 1.6285168753860841,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047439183702418765,
      "loss": 5.3534,
      "mean_token_accuracy": 0.1898641675710678,
      "num_tokens": 32346843.0,
      "step": 14500
    },
    {
      "entropy": 5.596896839141846,
      "epoch": 1.6290784522940416,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047437323175318716,
      "loss": 5.4467,
      "mean_token_accuracy": 0.18356406092643737,
      "num_tokens": 32358511.0,
      "step": 14505
    },
    {
      "entropy": 5.567827415466309,
      "epoch": 1.6296400292019992,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047435462013406554,
      "loss": 5.3398,
      "mean_token_accuracy": 0.18778470605611802,
      "num_tokens": 32370319.0,
      "step": 14510
    },
    {
      "entropy": 5.555706453323364,
      "epoch": 1.630201606109957,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004743360021674155,
      "loss": 5.4285,
      "mean_token_accuracy": 0.18525870591402055,
      "num_tokens": 32381455.0,
      "step": 14515
    },
    {
      "entropy": 5.579962825775146,
      "epoch": 1.6307631830179143,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047431737785382996,
      "loss": 5.4032,
      "mean_token_accuracy": 0.18517621904611586,
      "num_tokens": 32392553.0,
      "step": 14520
    },
    {
      "entropy": 5.733662414550781,
      "epoch": 1.6313247599258718,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047429874719390165,
      "loss": 5.4754,
      "mean_token_accuracy": 0.18362810909748079,
      "num_tokens": 32404736.0,
      "step": 14525
    },
    {
      "entropy": 5.5754382610321045,
      "epoch": 1.6318863368338294,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047428011018822404,
      "loss": 5.3613,
      "mean_token_accuracy": 0.18830402046442032,
      "num_tokens": 32415089.0,
      "step": 14530
    },
    {
      "entropy": 5.501427221298218,
      "epoch": 1.6324479137417869,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047426146683739043,
      "loss": 5.3667,
      "mean_token_accuracy": 0.18111147731542587,
      "num_tokens": 32426814.0,
      "step": 14535
    },
    {
      "entropy": 5.560883188247681,
      "epoch": 1.6330094906497443,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00047424281714199436,
      "loss": 5.4213,
      "mean_token_accuracy": 0.17635942846536637,
      "num_tokens": 32437173.0,
      "step": 14540
    },
    {
      "entropy": 5.584721612930298,
      "epoch": 1.633571067557702,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047422416110262975,
      "loss": 5.393,
      "mean_token_accuracy": 0.18085627853870392,
      "num_tokens": 32448597.0,
      "step": 14545
    },
    {
      "entropy": 5.672222852706909,
      "epoch": 1.6341326444656596,
      "grad_norm": 1.71875,
      "learning_rate": 0.00047420549871989053,
      "loss": 5.5749,
      "mean_token_accuracy": 0.17191092818975448,
      "num_tokens": 32460906.0,
      "step": 14550
    },
    {
      "entropy": 5.663802146911621,
      "epoch": 1.634694221373617,
      "grad_norm": 1.40625,
      "learning_rate": 0.000474186829994371,
      "loss": 5.3813,
      "mean_token_accuracy": 0.18450928777456282,
      "num_tokens": 32473756.0,
      "step": 14555
    },
    {
      "entropy": 5.5878105640411375,
      "epoch": 1.6352557982815745,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047416815492666564,
      "loss": 5.4435,
      "mean_token_accuracy": 0.1797104686498642,
      "num_tokens": 32483912.0,
      "step": 14560
    },
    {
      "entropy": 5.566701984405517,
      "epoch": 1.6358173751895322,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004741494735173689,
      "loss": 5.3784,
      "mean_token_accuracy": 0.17997499853372573,
      "num_tokens": 32493899.0,
      "step": 14565
    },
    {
      "entropy": 5.574539089202881,
      "epoch": 1.6363789520974898,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004741307857670757,
      "loss": 5.401,
      "mean_token_accuracy": 0.18635640144348145,
      "num_tokens": 32505215.0,
      "step": 14570
    },
    {
      "entropy": 5.566939735412598,
      "epoch": 1.6369405290054473,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047411209167638106,
      "loss": 5.4666,
      "mean_token_accuracy": 0.1814038038253784,
      "num_tokens": 32517288.0,
      "step": 14575
    },
    {
      "entropy": 5.639643812179566,
      "epoch": 1.6375021059134047,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004740933912458801,
      "loss": 5.4262,
      "mean_token_accuracy": 0.18364346027374268,
      "num_tokens": 32529705.0,
      "step": 14580
    },
    {
      "entropy": 5.718049049377441,
      "epoch": 1.6380636828213624,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047407468447616845,
      "loss": 5.4315,
      "mean_token_accuracy": 0.18514339476823807,
      "num_tokens": 32540868.0,
      "step": 14585
    },
    {
      "entropy": 5.495394134521485,
      "epoch": 1.63862525972932,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047405597136784166,
      "loss": 5.3822,
      "mean_token_accuracy": 0.18800296634435654,
      "num_tokens": 32552129.0,
      "step": 14590
    },
    {
      "entropy": 5.505461692810059,
      "epoch": 1.6391868366372775,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004740372519214955,
      "loss": 5.303,
      "mean_token_accuracy": 0.19333642572164536,
      "num_tokens": 32563059.0,
      "step": 14595
    },
    {
      "entropy": 5.553788805007935,
      "epoch": 1.639748413545235,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047401852613772604,
      "loss": 5.4939,
      "mean_token_accuracy": 0.17190300822257995,
      "num_tokens": 32576628.0,
      "step": 14600
    },
    {
      "entropy": 5.697915983200073,
      "epoch": 1.6403099904531926,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004739997940171295,
      "loss": 5.442,
      "mean_token_accuracy": 0.18500046133995057,
      "num_tokens": 32589138.0,
      "step": 14605
    },
    {
      "entropy": 5.6585431575775145,
      "epoch": 1.6408715673611503,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047398105556030234,
      "loss": 5.4269,
      "mean_token_accuracy": 0.18122628927230836,
      "num_tokens": 32601865.0,
      "step": 14610
    },
    {
      "entropy": 5.551347351074218,
      "epoch": 1.6414331442691077,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047396231076784123,
      "loss": 5.389,
      "mean_token_accuracy": 0.18326832950115204,
      "num_tokens": 32612439.0,
      "step": 14615
    },
    {
      "entropy": 5.5520836353302006,
      "epoch": 1.6419947211770651,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004739435596403429,
      "loss": 5.3862,
      "mean_token_accuracy": 0.18279039412736892,
      "num_tokens": 32623002.0,
      "step": 14620
    },
    {
      "entropy": 5.542112588882446,
      "epoch": 1.6425562980850228,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004739248021784046,
      "loss": 5.3594,
      "mean_token_accuracy": 0.1901555374264717,
      "num_tokens": 32632529.0,
      "step": 14625
    },
    {
      "entropy": 5.594372749328613,
      "epoch": 1.6431178749929805,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004739060383826234,
      "loss": 5.3717,
      "mean_token_accuracy": 0.18834180235862732,
      "num_tokens": 32643584.0,
      "step": 14630
    },
    {
      "entropy": 5.636145973205567,
      "epoch": 1.6436794519009377,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047388726825359673,
      "loss": 5.4677,
      "mean_token_accuracy": 0.17899191230535508,
      "num_tokens": 32655899.0,
      "step": 14635
    },
    {
      "entropy": 5.59476408958435,
      "epoch": 1.6442410288088953,
      "grad_norm": 1.5,
      "learning_rate": 0.00047386849179192235,
      "loss": 5.4077,
      "mean_token_accuracy": 0.18391844779253005,
      "num_tokens": 32666747.0,
      "step": 14640
    },
    {
      "entropy": 5.5709222793579105,
      "epoch": 1.644802605716853,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047384970899819813,
      "loss": 5.3547,
      "mean_token_accuracy": 0.18639909625053405,
      "num_tokens": 32678025.0,
      "step": 14645
    },
    {
      "entropy": 5.5783929347991945,
      "epoch": 1.6453641826248104,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004738309198730219,
      "loss": 5.3089,
      "mean_token_accuracy": 0.1884949326515198,
      "num_tokens": 32688894.0,
      "step": 14650
    },
    {
      "entropy": 5.545256853103638,
      "epoch": 1.6459257595327679,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047381212441699217,
      "loss": 5.4134,
      "mean_token_accuracy": 0.18252882212400437,
      "num_tokens": 32699530.0,
      "step": 14655
    },
    {
      "entropy": 5.611769962310791,
      "epoch": 1.6464873364407255,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004737933226307072,
      "loss": 5.4599,
      "mean_token_accuracy": 0.18351722806692122,
      "num_tokens": 32710969.0,
      "step": 14660
    },
    {
      "entropy": 5.597371339797974,
      "epoch": 1.6470489133486832,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047377451451476586,
      "loss": 5.3096,
      "mean_token_accuracy": 0.19139499962329865,
      "num_tokens": 32723715.0,
      "step": 14665
    },
    {
      "entropy": 5.516621255874634,
      "epoch": 1.6476104902566406,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004737557000697668,
      "loss": 5.3892,
      "mean_token_accuracy": 0.18492224663496018,
      "num_tokens": 32735406.0,
      "step": 14670
    },
    {
      "entropy": 5.529916048049927,
      "epoch": 1.648172067164598,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004737368792963092,
      "loss": 5.3963,
      "mean_token_accuracy": 0.180710369348526,
      "num_tokens": 32746123.0,
      "step": 14675
    },
    {
      "entropy": 5.5279628276824955,
      "epoch": 1.6487336440725557,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047371805219499223,
      "loss": 5.3562,
      "mean_token_accuracy": 0.18163390457630157,
      "num_tokens": 32757172.0,
      "step": 14680
    },
    {
      "entropy": 5.475903463363648,
      "epoch": 1.6492952209805134,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047369921876641547,
      "loss": 5.2677,
      "mean_token_accuracy": 0.18968770056962966,
      "num_tokens": 32767810.0,
      "step": 14685
    },
    {
      "entropy": 5.519390487670899,
      "epoch": 1.6498567978884708,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047368037901117854,
      "loss": 5.3575,
      "mean_token_accuracy": 0.1805579200387001,
      "num_tokens": 32779723.0,
      "step": 14690
    },
    {
      "entropy": 5.715040779113769,
      "epoch": 1.6504183747964283,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004736615329298812,
      "loss": 5.6407,
      "mean_token_accuracy": 0.16966511011123658,
      "num_tokens": 32793378.0,
      "step": 14695
    },
    {
      "entropy": 5.625489521026611,
      "epoch": 1.650979951704386,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047364268052312367,
      "loss": 5.3899,
      "mean_token_accuracy": 0.18817970603704454,
      "num_tokens": 32805643.0,
      "step": 14700
    },
    {
      "entropy": 5.6128757953643795,
      "epoch": 1.6515415286123436,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004736238217915061,
      "loss": 5.3579,
      "mean_token_accuracy": 0.18191513121128083,
      "num_tokens": 32818275.0,
      "step": 14705
    },
    {
      "entropy": 5.448544979095459,
      "epoch": 1.652103105520301,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047360495673562906,
      "loss": 5.2436,
      "mean_token_accuracy": 0.19300651252269746,
      "num_tokens": 32829548.0,
      "step": 14710
    },
    {
      "entropy": 5.503378009796142,
      "epoch": 1.6526646824282585,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004735860853560931,
      "loss": 5.438,
      "mean_token_accuracy": 0.18519551455974578,
      "num_tokens": 32841472.0,
      "step": 14715
    },
    {
      "entropy": 5.618911838531494,
      "epoch": 1.6532262593362161,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047356720765349926,
      "loss": 5.455,
      "mean_token_accuracy": 0.18254597783088683,
      "num_tokens": 32852589.0,
      "step": 14720
    },
    {
      "entropy": 5.575884246826172,
      "epoch": 1.6537878362441738,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004735483236284485,
      "loss": 5.3696,
      "mean_token_accuracy": 0.18488479107618333,
      "num_tokens": 32863100.0,
      "step": 14725
    },
    {
      "entropy": 5.642951726913452,
      "epoch": 1.654349413152131,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047352943328154203,
      "loss": 5.3704,
      "mean_token_accuracy": 0.18474053591489792,
      "num_tokens": 32874912.0,
      "step": 14730
    },
    {
      "entropy": 5.496847915649414,
      "epoch": 1.6549109900600887,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047351053661338145,
      "loss": 5.3078,
      "mean_token_accuracy": 0.18862911462783813,
      "num_tokens": 32886249.0,
      "step": 14735
    },
    {
      "entropy": 5.495778131484985,
      "epoch": 1.6554725669680463,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004734916336245684,
      "loss": 5.3942,
      "mean_token_accuracy": 0.17536802887916564,
      "num_tokens": 32898010.0,
      "step": 14740
    },
    {
      "entropy": 5.533234310150147,
      "epoch": 1.6560341438760038,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004734727243157047,
      "loss": 5.2644,
      "mean_token_accuracy": 0.18930499851703644,
      "num_tokens": 32909103.0,
      "step": 14745
    },
    {
      "entropy": 5.517409420013427,
      "epoch": 1.6565957207839612,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047345380868739253,
      "loss": 5.3499,
      "mean_token_accuracy": 0.18766181766986847,
      "num_tokens": 32919612.0,
      "step": 14750
    },
    {
      "entropy": 5.461494255065918,
      "epoch": 1.6571572976919189,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004734348867402341,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19135154634714127,
      "num_tokens": 32930072.0,
      "step": 14755
    },
    {
      "entropy": 5.66618070602417,
      "epoch": 1.6577188745998765,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004734159584748319,
      "loss": 5.5138,
      "mean_token_accuracy": 0.1844464957714081,
      "num_tokens": 32941984.0,
      "step": 14760
    },
    {
      "entropy": 5.592111968994141,
      "epoch": 1.658280451507834,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047339702389178874,
      "loss": 5.3599,
      "mean_token_accuracy": 0.19003164023160934,
      "num_tokens": 32952767.0,
      "step": 14765
    },
    {
      "entropy": 5.589803028106689,
      "epoch": 1.6588420284157914,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047337808299170725,
      "loss": 5.3993,
      "mean_token_accuracy": 0.18156627118587493,
      "num_tokens": 32964553.0,
      "step": 14770
    },
    {
      "entropy": 5.563896179199219,
      "epoch": 1.659403605323749,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00047335913577519075,
      "loss": 5.3401,
      "mean_token_accuracy": 0.1827165275812149,
      "num_tokens": 32975526.0,
      "step": 14775
    },
    {
      "entropy": 5.505002069473266,
      "epoch": 1.6599651822317067,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047334018224284244,
      "loss": 5.3922,
      "mean_token_accuracy": 0.18905986547470094,
      "num_tokens": 32987025.0,
      "step": 14780
    },
    {
      "entropy": 5.696928310394287,
      "epoch": 1.6605267591396642,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004733212223952657,
      "loss": 5.393,
      "mean_token_accuracy": 0.18647790253162383,
      "num_tokens": 32997447.0,
      "step": 14785
    },
    {
      "entropy": 5.665005445480347,
      "epoch": 1.6610883360476216,
      "grad_norm": 1.9921875,
      "learning_rate": 0.00047330225623306443,
      "loss": 5.4116,
      "mean_token_accuracy": 0.18931203484535217,
      "num_tokens": 33007432.0,
      "step": 14790
    },
    {
      "entropy": 5.550873756408691,
      "epoch": 1.6616499129555793,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047328328375684237,
      "loss": 5.4022,
      "mean_token_accuracy": 0.18748948872089385,
      "num_tokens": 33018547.0,
      "step": 14795
    },
    {
      "entropy": 5.556125783920288,
      "epoch": 1.662211489863537,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004732643049672036,
      "loss": 5.3341,
      "mean_token_accuracy": 0.1916821777820587,
      "num_tokens": 33029092.0,
      "step": 14800
    },
    {
      "entropy": 5.5294407367706295,
      "epoch": 1.6627730667714944,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004732453198647525,
      "loss": 5.3738,
      "mean_token_accuracy": 0.18623017370700837,
      "num_tokens": 33040498.0,
      "step": 14805
    },
    {
      "entropy": 5.674070644378662,
      "epoch": 1.6633346436794518,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004732263284500936,
      "loss": 5.4432,
      "mean_token_accuracy": 0.17861680388450624,
      "num_tokens": 33052203.0,
      "step": 14810
    },
    {
      "entropy": 5.612527561187744,
      "epoch": 1.6638962205874095,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004732073307238314,
      "loss": 5.3706,
      "mean_token_accuracy": 0.18829278647899628,
      "num_tokens": 33063453.0,
      "step": 14815
    },
    {
      "entropy": 5.528048753738403,
      "epoch": 1.6644577974953672,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000473188326686571,
      "loss": 5.4549,
      "mean_token_accuracy": 0.18797820508480073,
      "num_tokens": 33075511.0,
      "step": 14820
    },
    {
      "entropy": 5.544820976257324,
      "epoch": 1.6650193744033244,
      "grad_norm": 1.9921875,
      "learning_rate": 0.00047316931633891736,
      "loss": 5.4192,
      "mean_token_accuracy": 0.18570923805236816,
      "num_tokens": 33086094.0,
      "step": 14825
    },
    {
      "entropy": 5.723085737228393,
      "epoch": 1.665580951311282,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004731502996814759,
      "loss": 5.4785,
      "mean_token_accuracy": 0.18691678792238237,
      "num_tokens": 33096430.0,
      "step": 14830
    },
    {
      "entropy": 5.544121360778808,
      "epoch": 1.6661425282192397,
      "grad_norm": 1.5,
      "learning_rate": 0.00047313127671485194,
      "loss": 5.3575,
      "mean_token_accuracy": 0.18353654593229293,
      "num_tokens": 33108040.0,
      "step": 14835
    },
    {
      "entropy": 5.521466827392578,
      "epoch": 1.6667041051271971,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004731122474396514,
      "loss": 5.3552,
      "mean_token_accuracy": 0.18125558495521546,
      "num_tokens": 33118282.0,
      "step": 14840
    },
    {
      "entropy": 5.588000249862671,
      "epoch": 1.6672656820351546,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047309321185648,
      "loss": 5.4326,
      "mean_token_accuracy": 0.18425653874874115,
      "num_tokens": 33130308.0,
      "step": 14845
    },
    {
      "entropy": 5.67760157585144,
      "epoch": 1.6678272589431122,
      "grad_norm": 1.375,
      "learning_rate": 0.00047307416996594386,
      "loss": 5.4662,
      "mean_token_accuracy": 0.18178492337465285,
      "num_tokens": 33140260.0,
      "step": 14850
    },
    {
      "entropy": 5.555676698684692,
      "epoch": 1.66838883585107,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004730551217686495,
      "loss": 5.3738,
      "mean_token_accuracy": 0.1839437499642372,
      "num_tokens": 33152064.0,
      "step": 14855
    },
    {
      "entropy": 5.609837865829467,
      "epoch": 1.6689504127590273,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004730360672652031,
      "loss": 5.4331,
      "mean_token_accuracy": 0.18601180613040924,
      "num_tokens": 33164057.0,
      "step": 14860
    },
    {
      "entropy": 5.552616930007934,
      "epoch": 1.6695119896669848,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047301700645621154,
      "loss": 5.3969,
      "mean_token_accuracy": 0.186902716755867,
      "num_tokens": 33175281.0,
      "step": 14865
    },
    {
      "entropy": 5.562746953964234,
      "epoch": 1.6700735665749424,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004729979393422817,
      "loss": 5.3373,
      "mean_token_accuracy": 0.18874657303094863,
      "num_tokens": 33186615.0,
      "step": 14870
    },
    {
      "entropy": 5.572027587890625,
      "epoch": 1.6706351434829,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004729788659240206,
      "loss": 5.4406,
      "mean_token_accuracy": 0.18436575084924697,
      "num_tokens": 33198693.0,
      "step": 14875
    },
    {
      "entropy": 5.623181486129761,
      "epoch": 1.6711967203908575,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004729597862020358,
      "loss": 5.4047,
      "mean_token_accuracy": 0.1872449114918709,
      "num_tokens": 33209743.0,
      "step": 14880
    },
    {
      "entropy": 5.558723306655883,
      "epoch": 1.671758297298815,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047294070017693453,
      "loss": 5.3082,
      "mean_token_accuracy": 0.19150487333536148,
      "num_tokens": 33220907.0,
      "step": 14885
    },
    {
      "entropy": 5.617772722244263,
      "epoch": 1.6723198742067726,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047292160784932454,
      "loss": 5.4746,
      "mean_token_accuracy": 0.1800364747643471,
      "num_tokens": 33232278.0,
      "step": 14890
    },
    {
      "entropy": 5.626573467254639,
      "epoch": 1.6728814511147303,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047290250921981386,
      "loss": 5.3738,
      "mean_token_accuracy": 0.1844427466392517,
      "num_tokens": 33243920.0,
      "step": 14895
    },
    {
      "entropy": 5.598547506332397,
      "epoch": 1.6734430280226877,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047288340428901053,
      "loss": 5.4237,
      "mean_token_accuracy": 0.18580759763717652,
      "num_tokens": 33254467.0,
      "step": 14900
    },
    {
      "entropy": 5.502971839904785,
      "epoch": 1.6740046049306452,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047286429305752283,
      "loss": 5.3679,
      "mean_token_accuracy": 0.18248131275177001,
      "num_tokens": 33265536.0,
      "step": 14905
    },
    {
      "entropy": 5.5815403938293455,
      "epoch": 1.6745661818386028,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047284517552595933,
      "loss": 5.412,
      "mean_token_accuracy": 0.18008261770009995,
      "num_tokens": 33277233.0,
      "step": 14910
    },
    {
      "entropy": 5.517041254043579,
      "epoch": 1.6751277587465605,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004728260516949287,
      "loss": 5.2767,
      "mean_token_accuracy": 0.19094849675893782,
      "num_tokens": 33287873.0,
      "step": 14915
    },
    {
      "entropy": 5.530544853210449,
      "epoch": 1.6756893356545177,
      "grad_norm": 1.5,
      "learning_rate": 0.0004728069215650399,
      "loss": 5.3918,
      "mean_token_accuracy": 0.18779995441436767,
      "num_tokens": 33298888.0,
      "step": 14920
    },
    {
      "entropy": 5.499293565750122,
      "epoch": 1.6762509125624754,
      "grad_norm": 1.875,
      "learning_rate": 0.0004727877851369019,
      "loss": 5.2611,
      "mean_token_accuracy": 0.19046627432107927,
      "num_tokens": 33309607.0,
      "step": 14925
    },
    {
      "entropy": 5.550456094741821,
      "epoch": 1.676812489470433,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047276864241112417,
      "loss": 5.3027,
      "mean_token_accuracy": 0.19379159212112426,
      "num_tokens": 33320582.0,
      "step": 14930
    },
    {
      "entropy": 5.508336687088013,
      "epoch": 1.6773740663783905,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004727494933883161,
      "loss": 5.3578,
      "mean_token_accuracy": 0.19197508096694946,
      "num_tokens": 33332145.0,
      "step": 14935
    },
    {
      "entropy": 5.694448089599609,
      "epoch": 1.677935643286348,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004727303380690876,
      "loss": 5.6083,
      "mean_token_accuracy": 0.17880181670188905,
      "num_tokens": 33343963.0,
      "step": 14940
    },
    {
      "entropy": 5.568664169311523,
      "epoch": 1.6784972201943056,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004727111764540484,
      "loss": 5.4014,
      "mean_token_accuracy": 0.18977900594472885,
      "num_tokens": 33354931.0,
      "step": 14945
    },
    {
      "entropy": 5.5538811683654785,
      "epoch": 1.6790587971022632,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047269200854380854,
      "loss": 5.3958,
      "mean_token_accuracy": 0.1820443630218506,
      "num_tokens": 33365686.0,
      "step": 14950
    },
    {
      "entropy": 5.643840694427491,
      "epoch": 1.6796203740102207,
      "grad_norm": 1.25,
      "learning_rate": 0.00047267283433897854,
      "loss": 5.3946,
      "mean_token_accuracy": 0.18865703791379929,
      "num_tokens": 33377099.0,
      "step": 14955
    },
    {
      "entropy": 5.581523275375366,
      "epoch": 1.6801819509181781,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047265365384016887,
      "loss": 5.4437,
      "mean_token_accuracy": 0.1765981525182724,
      "num_tokens": 33389477.0,
      "step": 14960
    },
    {
      "entropy": 5.618612241744995,
      "epoch": 1.6807435278261358,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047263446704799017,
      "loss": 5.4663,
      "mean_token_accuracy": 0.18694839626550674,
      "num_tokens": 33401970.0,
      "step": 14965
    },
    {
      "entropy": 5.5506926536560055,
      "epoch": 1.6813051047340934,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047261527396305343,
      "loss": 5.3398,
      "mean_token_accuracy": 0.182860167324543,
      "num_tokens": 33412986.0,
      "step": 14970
    },
    {
      "entropy": 5.588017368316651,
      "epoch": 1.6818666816420509,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047259607458596966,
      "loss": 5.4763,
      "mean_token_accuracy": 0.18464904576539992,
      "num_tokens": 33424717.0,
      "step": 14975
    },
    {
      "entropy": 5.476791286468506,
      "epoch": 1.6824282585500083,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047257686891735023,
      "loss": 5.287,
      "mean_token_accuracy": 0.18902431577444076,
      "num_tokens": 33436713.0,
      "step": 14980
    },
    {
      "entropy": 5.656904935836792,
      "epoch": 1.682989835457966,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047255765695780665,
      "loss": 5.4831,
      "mean_token_accuracy": 0.17948793321847917,
      "num_tokens": 33447532.0,
      "step": 14985
    },
    {
      "entropy": 5.4818048000335695,
      "epoch": 1.6835514123659236,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004725384387079507,
      "loss": 5.3035,
      "mean_token_accuracy": 0.18263816684484482,
      "num_tokens": 33458530.0,
      "step": 14990
    },
    {
      "entropy": 5.551608562469482,
      "epoch": 1.684112989273881,
      "grad_norm": 1.5,
      "learning_rate": 0.00047251921416839425,
      "loss": 5.4123,
      "mean_token_accuracy": 0.17943487763404847,
      "num_tokens": 33469617.0,
      "step": 14995
    },
    {
      "entropy": 5.626010417938232,
      "epoch": 1.6846745661818385,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004724999833397494,
      "loss": 5.323,
      "mean_token_accuracy": 0.19285964518785476,
      "num_tokens": 33479622.0,
      "step": 15000
    },
    {
      "epoch": 1.6846745661818385,
      "eval_entropy": 5.383775010870264,
      "eval_loss": 5.4565653800964355,
      "eval_mean_token_accuracy": 0.1902181858315522,
      "eval_num_tokens": 33479622.0,
      "eval_runtime": 24.0402,
      "eval_samples_per_second": 1290.047,
      "eval_steps_per_second": 161.271,
      "step": 15000
    },
    {
      "entropy": 5.4471039295196535,
      "epoch": 1.6852361430897962,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004724807462226285,
      "loss": 5.2883,
      "mean_token_accuracy": 0.19635369181632994,
      "num_tokens": 33490907.0,
      "step": 15005
    },
    {
      "entropy": 5.551421070098877,
      "epoch": 1.6857977199977539,
      "grad_norm": 1.46875,
      "learning_rate": 0.000472461502817644,
      "loss": 5.4051,
      "mean_token_accuracy": 0.18559968769550322,
      "num_tokens": 33501884.0,
      "step": 15010
    },
    {
      "entropy": 5.557754468917847,
      "epoch": 1.6863592969057113,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004724422531254087,
      "loss": 5.2736,
      "mean_token_accuracy": 0.18997342735528946,
      "num_tokens": 33512106.0,
      "step": 15015
    },
    {
      "entropy": 5.582626485824585,
      "epoch": 1.6869208738136687,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004724229971465355,
      "loss": 5.3867,
      "mean_token_accuracy": 0.19005310386419297,
      "num_tokens": 33522889.0,
      "step": 15020
    },
    {
      "entropy": 5.546910095214844,
      "epoch": 1.6874824507216264,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004724037348816375,
      "loss": 5.3544,
      "mean_token_accuracy": 0.18871113210916518,
      "num_tokens": 33534601.0,
      "step": 15025
    },
    {
      "entropy": 5.513442134857177,
      "epoch": 1.6880440276295838,
      "grad_norm": 1.484375,
      "learning_rate": 0.000472384466331328,
      "loss": 5.4522,
      "mean_token_accuracy": 0.1847732901573181,
      "num_tokens": 33545300.0,
      "step": 15030
    },
    {
      "entropy": 5.687134075164795,
      "epoch": 1.6886056045375413,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004723651914962205,
      "loss": 5.4112,
      "mean_token_accuracy": 0.18240966796875,
      "num_tokens": 33555581.0,
      "step": 15035
    },
    {
      "entropy": 5.613533639907837,
      "epoch": 1.689167181445499,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047234591037692884,
      "loss": 5.4,
      "mean_token_accuracy": 0.18385293036699296,
      "num_tokens": 33566902.0,
      "step": 15040
    },
    {
      "entropy": 5.57280707359314,
      "epoch": 1.6897287583534566,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047232662297406683,
      "loss": 5.4161,
      "mean_token_accuracy": 0.18380822986364365,
      "num_tokens": 33578316.0,
      "step": 15045
    },
    {
      "entropy": 5.603354120254517,
      "epoch": 1.690290335261414,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004723073292882485,
      "loss": 5.4311,
      "mean_token_accuracy": 0.17819164395332338,
      "num_tokens": 33590242.0,
      "step": 15050
    },
    {
      "entropy": 5.548894309997559,
      "epoch": 1.6908519121693715,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047228802932008836,
      "loss": 5.3229,
      "mean_token_accuracy": 0.1852489396929741,
      "num_tokens": 33601277.0,
      "step": 15055
    },
    {
      "entropy": 5.528900814056397,
      "epoch": 1.6914134890773291,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004722687230702009,
      "loss": 5.3619,
      "mean_token_accuracy": 0.1814175859093666,
      "num_tokens": 33612423.0,
      "step": 15060
    },
    {
      "entropy": 5.672621393203736,
      "epoch": 1.6919750659852868,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004722494105392007,
      "loss": 5.4417,
      "mean_token_accuracy": 0.18055731058120728,
      "num_tokens": 33624250.0,
      "step": 15065
    },
    {
      "entropy": 5.6068737506866455,
      "epoch": 1.6925366428932442,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004722300917277029,
      "loss": 5.3778,
      "mean_token_accuracy": 0.18464804738759993,
      "num_tokens": 33636379.0,
      "step": 15070
    },
    {
      "entropy": 5.537733650207519,
      "epoch": 1.6930982198012017,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004722107666363224,
      "loss": 5.3834,
      "mean_token_accuracy": 0.18061763942241668,
      "num_tokens": 33648296.0,
      "step": 15075
    },
    {
      "entropy": 5.479118394851684,
      "epoch": 1.6936597967091593,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004721914352656746,
      "loss": 5.4005,
      "mean_token_accuracy": 0.18329438269138337,
      "num_tokens": 33659438.0,
      "step": 15080
    },
    {
      "entropy": 5.636709928512573,
      "epoch": 1.694221373617117,
      "grad_norm": 2.71875,
      "learning_rate": 0.000472172097616375,
      "loss": 5.4973,
      "mean_token_accuracy": 0.18170636296272277,
      "num_tokens": 33670893.0,
      "step": 15085
    },
    {
      "entropy": 5.621981287002564,
      "epoch": 1.6947829505250744,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004721527536890394,
      "loss": 5.3633,
      "mean_token_accuracy": 0.1880887269973755,
      "num_tokens": 33681487.0,
      "step": 15090
    },
    {
      "entropy": 5.557020998001098,
      "epoch": 1.6953445274330319,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004721334034842836,
      "loss": 5.3835,
      "mean_token_accuracy": 0.19438282400369644,
      "num_tokens": 33692925.0,
      "step": 15095
    },
    {
      "entropy": 5.584948348999023,
      "epoch": 1.6959061043409895,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004721140470027238,
      "loss": 5.3541,
      "mean_token_accuracy": 0.18742522597312927,
      "num_tokens": 33703559.0,
      "step": 15100
    },
    {
      "entropy": 5.504276752471924,
      "epoch": 1.6964676812489472,
      "grad_norm": 1.8125,
      "learning_rate": 0.00047209468424497627,
      "loss": 5.3614,
      "mean_token_accuracy": 0.18574756234884263,
      "num_tokens": 33715041.0,
      "step": 15105
    },
    {
      "entropy": 5.49846305847168,
      "epoch": 1.6970292581569046,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004720753152116576,
      "loss": 5.3071,
      "mean_token_accuracy": 0.18679920136928557,
      "num_tokens": 33726334.0,
      "step": 15110
    },
    {
      "entropy": 5.611520433425904,
      "epoch": 1.697590835064862,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047205593990338443,
      "loss": 5.3407,
      "mean_token_accuracy": 0.18786376416683198,
      "num_tokens": 33738663.0,
      "step": 15115
    },
    {
      "entropy": 5.489375638961792,
      "epoch": 1.6981524119728197,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004720365583207737,
      "loss": 5.3027,
      "mean_token_accuracy": 0.19285307973623275,
      "num_tokens": 33749202.0,
      "step": 15120
    },
    {
      "entropy": 5.4849128246307375,
      "epoch": 1.6987139888807772,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004720171704644425,
      "loss": 5.2708,
      "mean_token_accuracy": 0.19944576025009156,
      "num_tokens": 33759809.0,
      "step": 15125
    },
    {
      "entropy": 5.494880485534668,
      "epoch": 1.6992755657887346,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004719977763350083,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18570796847343446,
      "num_tokens": 33771374.0,
      "step": 15130
    },
    {
      "entropy": 5.537381267547607,
      "epoch": 1.6998371426966923,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047197837593308833,
      "loss": 5.3648,
      "mean_token_accuracy": 0.18862795382738112,
      "num_tokens": 33781645.0,
      "step": 15135
    },
    {
      "entropy": 5.565759325027466,
      "epoch": 1.70039871960465,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004719589692593006,
      "loss": 5.3458,
      "mean_token_accuracy": 0.18986233025789262,
      "num_tokens": 33793032.0,
      "step": 15140
    },
    {
      "entropy": 5.6880542755126955,
      "epoch": 1.7009602965126074,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004719395563142629,
      "loss": 5.5586,
      "mean_token_accuracy": 0.17638744413852692,
      "num_tokens": 33805116.0,
      "step": 15145
    },
    {
      "entropy": 5.646629333496094,
      "epoch": 1.7015218734205648,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004719201370985933,
      "loss": 5.3964,
      "mean_token_accuracy": 0.18295258432626724,
      "num_tokens": 33817135.0,
      "step": 15150
    },
    {
      "entropy": 5.574564695358276,
      "epoch": 1.7020834503285225,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047190071161291007,
      "loss": 5.3841,
      "mean_token_accuracy": 0.18291592746973037,
      "num_tokens": 33830187.0,
      "step": 15155
    },
    {
      "entropy": 5.458574628829956,
      "epoch": 1.7026450272364801,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047188127985783195,
      "loss": 5.3872,
      "mean_token_accuracy": 0.17977226823568343,
      "num_tokens": 33841297.0,
      "step": 15160
    },
    {
      "entropy": 5.59752893447876,
      "epoch": 1.7032066041444376,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047186184183397754,
      "loss": 5.3761,
      "mean_token_accuracy": 0.18661357313394547,
      "num_tokens": 33852888.0,
      "step": 15165
    },
    {
      "entropy": 5.568714380264282,
      "epoch": 1.703768181052395,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047184239754196564,
      "loss": 5.4437,
      "mean_token_accuracy": 0.18394217193126677,
      "num_tokens": 33865165.0,
      "step": 15170
    },
    {
      "entropy": 5.533801794052124,
      "epoch": 1.7043297579603527,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004718229469824155,
      "loss": 5.2487,
      "mean_token_accuracy": 0.19285314679145812,
      "num_tokens": 33875865.0,
      "step": 15175
    },
    {
      "entropy": 5.592979621887207,
      "epoch": 1.7048913348683103,
      "grad_norm": 2.265625,
      "learning_rate": 0.00047180349015594634,
      "loss": 5.482,
      "mean_token_accuracy": 0.17598691284656526,
      "num_tokens": 33887519.0,
      "step": 15180
    },
    {
      "entropy": 5.504102420806885,
      "epoch": 1.7054529117762678,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004717840270631778,
      "loss": 5.2914,
      "mean_token_accuracy": 0.19494041949510574,
      "num_tokens": 33898762.0,
      "step": 15185
    },
    {
      "entropy": 5.438482666015625,
      "epoch": 1.7060144886842252,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047176455770472946,
      "loss": 5.2926,
      "mean_token_accuracy": 0.1906571790575981,
      "num_tokens": 33911109.0,
      "step": 15190
    },
    {
      "entropy": 5.634068346023559,
      "epoch": 1.706576065592183,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047174508208122135,
      "loss": 5.5684,
      "mean_token_accuracy": 0.17408591955900193,
      "num_tokens": 33922741.0,
      "step": 15195
    },
    {
      "entropy": 5.581002140045166,
      "epoch": 1.7071376425001406,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004717256001932735,
      "loss": 5.3811,
      "mean_token_accuracy": 0.18276015520095826,
      "num_tokens": 33933234.0,
      "step": 15200
    },
    {
      "entropy": 5.526957702636719,
      "epoch": 1.707699219408098,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047170611204150624,
      "loss": 5.3139,
      "mean_token_accuracy": 0.18669446408748627,
      "num_tokens": 33943667.0,
      "step": 15205
    },
    {
      "entropy": 5.519064235687256,
      "epoch": 1.7082607963160554,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004716866176265401,
      "loss": 5.2847,
      "mean_token_accuracy": 0.19637755006551744,
      "num_tokens": 33954499.0,
      "step": 15210
    },
    {
      "entropy": 5.502477359771729,
      "epoch": 1.708822373224013,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047166711694899573,
      "loss": 5.3491,
      "mean_token_accuracy": 0.19001880586147307,
      "num_tokens": 33964700.0,
      "step": 15215
    },
    {
      "entropy": 5.514243841171265,
      "epoch": 1.7093839501319705,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047164761000949404,
      "loss": 5.4042,
      "mean_token_accuracy": 0.18515589386224746,
      "num_tokens": 33977759.0,
      "step": 15220
    },
    {
      "entropy": 5.6283937931060795,
      "epoch": 1.709945527039928,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004716280968086563,
      "loss": 5.4639,
      "mean_token_accuracy": 0.1797819197177887,
      "num_tokens": 33990019.0,
      "step": 15225
    },
    {
      "entropy": 5.613171672821045,
      "epoch": 1.7105071039478856,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004716085773471036,
      "loss": 5.3503,
      "mean_token_accuracy": 0.18897001296281815,
      "num_tokens": 34000663.0,
      "step": 15230
    },
    {
      "entropy": 5.459926891326904,
      "epoch": 1.7110686808558433,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047158905162545756,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19417371451854706,
      "num_tokens": 34012768.0,
      "step": 15235
    },
    {
      "entropy": 5.470266914367675,
      "epoch": 1.7116302577638007,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047156951964433984,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19037491083145142,
      "num_tokens": 34025801.0,
      "step": 15240
    },
    {
      "entropy": 5.491405773162842,
      "epoch": 1.7121918346717582,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00047154998140437246,
      "loss": 5.2082,
      "mean_token_accuracy": 0.20296344310045242,
      "num_tokens": 34036553.0,
      "step": 15245
    },
    {
      "entropy": 5.527611255645752,
      "epoch": 1.7127534115797158,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047153043690617737,
      "loss": 5.3886,
      "mean_token_accuracy": 0.18829963505268096,
      "num_tokens": 34048352.0,
      "step": 15250
    },
    {
      "entropy": 5.576293706893921,
      "epoch": 1.7133149884876735,
      "grad_norm": 1.34375,
      "learning_rate": 0.000471510886150377,
      "loss": 5.3432,
      "mean_token_accuracy": 0.19270518571138381,
      "num_tokens": 34059617.0,
      "step": 15255
    },
    {
      "entropy": 5.562987804412842,
      "epoch": 1.713876565395631,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004714913291375938,
      "loss": 5.4001,
      "mean_token_accuracy": 0.18346952199935912,
      "num_tokens": 34071176.0,
      "step": 15260
    },
    {
      "entropy": 5.4848121166229244,
      "epoch": 1.7144381423035884,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047147176586845036,
      "loss": 5.3518,
      "mean_token_accuracy": 0.18714902997016908,
      "num_tokens": 34081883.0,
      "step": 15265
    },
    {
      "entropy": 5.551849412918091,
      "epoch": 1.714999719211546,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004714521963435698,
      "loss": 5.3679,
      "mean_token_accuracy": 0.18345563858747482,
      "num_tokens": 34092938.0,
      "step": 15270
    },
    {
      "entropy": 5.604917812347412,
      "epoch": 1.7155612961195037,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047143262056357506,
      "loss": 5.3316,
      "mean_token_accuracy": 0.19288102239370347,
      "num_tokens": 34104758.0,
      "step": 15275
    },
    {
      "entropy": 5.555082178115844,
      "epoch": 1.7161228730274611,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047141303852908965,
      "loss": 5.2539,
      "mean_token_accuracy": 0.19242660254240035,
      "num_tokens": 34114752.0,
      "step": 15280
    },
    {
      "entropy": 5.581843376159668,
      "epoch": 1.7166844499354186,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004713934502407367,
      "loss": 5.4126,
      "mean_token_accuracy": 0.18007928133010864,
      "num_tokens": 34125547.0,
      "step": 15285
    },
    {
      "entropy": 5.675565767288208,
      "epoch": 1.7172460268433762,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004713738556991402,
      "loss": 5.4546,
      "mean_token_accuracy": 0.1848657414317131,
      "num_tokens": 34136085.0,
      "step": 15290
    },
    {
      "entropy": 5.5911674976348875,
      "epoch": 1.717807603751334,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000471354254904924,
      "loss": 5.3871,
      "mean_token_accuracy": 0.19026076495647432,
      "num_tokens": 34147536.0,
      "step": 15295
    },
    {
      "entropy": 5.579338026046753,
      "epoch": 1.7183691806592913,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047133464785871216,
      "loss": 5.5024,
      "mean_token_accuracy": 0.1849495455622673,
      "num_tokens": 34158238.0,
      "step": 15300
    },
    {
      "entropy": 5.500206089019775,
      "epoch": 1.7189307575672488,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000471315034561129,
      "loss": 5.2063,
      "mean_token_accuracy": 0.1989926964044571,
      "num_tokens": 34169793.0,
      "step": 15305
    },
    {
      "entropy": 5.59913649559021,
      "epoch": 1.7194923344752064,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047129541501279897,
      "loss": 5.4041,
      "mean_token_accuracy": 0.19394327849149703,
      "num_tokens": 34179753.0,
      "step": 15310
    },
    {
      "entropy": 5.600598907470703,
      "epoch": 1.7200539113831639,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004712757892143468,
      "loss": 5.4824,
      "mean_token_accuracy": 0.18532252162694932,
      "num_tokens": 34189947.0,
      "step": 15315
    },
    {
      "entropy": 5.636711502075196,
      "epoch": 1.7206154882911213,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047125615716639744,
      "loss": 5.4322,
      "mean_token_accuracy": 0.1880556896328926,
      "num_tokens": 34201197.0,
      "step": 15320
    },
    {
      "entropy": 5.590401458740234,
      "epoch": 1.721177065199079,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004712365188695758,
      "loss": 5.4262,
      "mean_token_accuracy": 0.19257451444864274,
      "num_tokens": 34212762.0,
      "step": 15325
    },
    {
      "entropy": 5.569433212280273,
      "epoch": 1.7217386421070366,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004712168743245074,
      "loss": 5.3809,
      "mean_token_accuracy": 0.18759627640247345,
      "num_tokens": 34223928.0,
      "step": 15330
    },
    {
      "entropy": 5.529983615875244,
      "epoch": 1.722300219014994,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047119722353181754,
      "loss": 5.283,
      "mean_token_accuracy": 0.20313519090414048,
      "num_tokens": 34233888.0,
      "step": 15335
    },
    {
      "entropy": 5.627472019195556,
      "epoch": 1.7228617959229515,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047117756649213204,
      "loss": 5.3958,
      "mean_token_accuracy": 0.18413452059030533,
      "num_tokens": 34245894.0,
      "step": 15340
    },
    {
      "entropy": 5.563360166549683,
      "epoch": 1.7234233728309092,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004711579032060767,
      "loss": 5.2997,
      "mean_token_accuracy": 0.19482649117708206,
      "num_tokens": 34256288.0,
      "step": 15345
    },
    {
      "entropy": 5.566926193237305,
      "epoch": 1.7239849497388668,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004711382336742777,
      "loss": 5.351,
      "mean_token_accuracy": 0.1786728397011757,
      "num_tokens": 34266515.0,
      "step": 15350
    },
    {
      "entropy": 5.59462161064148,
      "epoch": 1.7245465266468243,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004711185578973612,
      "loss": 5.4231,
      "mean_token_accuracy": 0.1808302104473114,
      "num_tokens": 34278681.0,
      "step": 15355
    },
    {
      "entropy": 5.609726810455323,
      "epoch": 1.7251081035547817,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004710988758759538,
      "loss": 5.3395,
      "mean_token_accuracy": 0.19018030911684036,
      "num_tokens": 34289161.0,
      "step": 15360
    },
    {
      "entropy": 5.545761442184448,
      "epoch": 1.7256696804627394,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004710791876106821,
      "loss": 5.319,
      "mean_token_accuracy": 0.18373594731092452,
      "num_tokens": 34300323.0,
      "step": 15365
    },
    {
      "entropy": 5.518750762939453,
      "epoch": 1.726231257370697,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047105949310217304,
      "loss": 5.2944,
      "mean_token_accuracy": 0.18588107079267502,
      "num_tokens": 34311558.0,
      "step": 15370
    },
    {
      "entropy": 5.4772566795349125,
      "epoch": 1.7267928342786545,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004710397923510537,
      "loss": 5.4157,
      "mean_token_accuracy": 0.18117000311613082,
      "num_tokens": 34323573.0,
      "step": 15375
    },
    {
      "entropy": 5.546475267410278,
      "epoch": 1.727354411186612,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047102008535795134,
      "loss": 5.3554,
      "mean_token_accuracy": 0.18172081112861632,
      "num_tokens": 34334698.0,
      "step": 15380
    },
    {
      "entropy": 5.602615737915039,
      "epoch": 1.7279159880945696,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004710003721234934,
      "loss": 5.3293,
      "mean_token_accuracy": 0.1962783992290497,
      "num_tokens": 34345948.0,
      "step": 15385
    },
    {
      "entropy": 5.559959602355957,
      "epoch": 1.7284775650025272,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047098065264830754,
      "loss": 5.3905,
      "mean_token_accuracy": 0.18383126109838485,
      "num_tokens": 34359067.0,
      "step": 15390
    },
    {
      "entropy": 5.403591728210449,
      "epoch": 1.7290391419104847,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047096092693302174,
      "loss": 5.2503,
      "mean_token_accuracy": 0.19293245822191238,
      "num_tokens": 34370895.0,
      "step": 15395
    },
    {
      "entropy": 5.506108713150025,
      "epoch": 1.7296007188184421,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047094119497826406,
      "loss": 5.3551,
      "mean_token_accuracy": 0.18479404747486114,
      "num_tokens": 34382039.0,
      "step": 15400
    },
    {
      "entropy": 5.632439422607422,
      "epoch": 1.7301622957263998,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047092145678466273,
      "loss": 5.4327,
      "mean_token_accuracy": 0.18405346870422362,
      "num_tokens": 34392939.0,
      "step": 15405
    },
    {
      "entropy": 5.674262189865113,
      "epoch": 1.7307238726343572,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004709017123528462,
      "loss": 5.4423,
      "mean_token_accuracy": 0.18372531682252885,
      "num_tokens": 34404402.0,
      "step": 15410
    },
    {
      "entropy": 5.486842155456543,
      "epoch": 1.7312854495423147,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047088196168344315,
      "loss": 5.3434,
      "mean_token_accuracy": 0.19193651378154755,
      "num_tokens": 34415563.0,
      "step": 15415
    },
    {
      "entropy": 5.603653287887573,
      "epoch": 1.7318470264502723,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00047086220477708256,
      "loss": 5.4514,
      "mean_token_accuracy": 0.17992481589317322,
      "num_tokens": 34426993.0,
      "step": 15420
    },
    {
      "entropy": 5.6662970066070555,
      "epoch": 1.73240860335823,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047084244163439335,
      "loss": 5.3841,
      "mean_token_accuracy": 0.19330619722604753,
      "num_tokens": 34436697.0,
      "step": 15425
    },
    {
      "entropy": 5.587507629394532,
      "epoch": 1.7329701802661874,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004708226722560049,
      "loss": 5.3638,
      "mean_token_accuracy": 0.1874844491481781,
      "num_tokens": 34448076.0,
      "step": 15430
    },
    {
      "entropy": 5.508535099029541,
      "epoch": 1.7335317571741449,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004708028966425465,
      "loss": 5.4078,
      "mean_token_accuracy": 0.18446434140205384,
      "num_tokens": 34460369.0,
      "step": 15435
    },
    {
      "entropy": 5.63477840423584,
      "epoch": 1.7340933340821025,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047078311479464797,
      "loss": 5.4748,
      "mean_token_accuracy": 0.1780938521027565,
      "num_tokens": 34472693.0,
      "step": 15440
    },
    {
      "entropy": 5.7154499053955075,
      "epoch": 1.7346549109900602,
      "grad_norm": 2.484375,
      "learning_rate": 0.00047076332671293917,
      "loss": 5.4107,
      "mean_token_accuracy": 0.19212453812360764,
      "num_tokens": 34485258.0,
      "step": 15445
    },
    {
      "entropy": 5.629126977920532,
      "epoch": 1.7352164878980176,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047074353239805005,
      "loss": 5.387,
      "mean_token_accuracy": 0.18570349365472794,
      "num_tokens": 34496450.0,
      "step": 15450
    },
    {
      "entropy": 5.530909729003906,
      "epoch": 1.735778064805975,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004707237318506111,
      "loss": 5.4207,
      "mean_token_accuracy": 0.18627371042966842,
      "num_tokens": 34508062.0,
      "step": 15455
    },
    {
      "entropy": 5.5551708221435545,
      "epoch": 1.7363396417139327,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047070392507125245,
      "loss": 5.3716,
      "mean_token_accuracy": 0.19078274071216583,
      "num_tokens": 34519817.0,
      "step": 15460
    },
    {
      "entropy": 5.569928073883057,
      "epoch": 1.7369012186218904,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000470684112060605,
      "loss": 5.3789,
      "mean_token_accuracy": 0.1850120022892952,
      "num_tokens": 34532628.0,
      "step": 15465
    },
    {
      "entropy": 5.620344400405884,
      "epoch": 1.7374627955298478,
      "grad_norm": 2.40625,
      "learning_rate": 0.00047066429281929944,
      "loss": 5.4333,
      "mean_token_accuracy": 0.1882002368569374,
      "num_tokens": 34543558.0,
      "step": 15470
    },
    {
      "entropy": 5.661240339279175,
      "epoch": 1.7380243724378053,
      "grad_norm": 1.328125,
      "learning_rate": 0.000470644467347967,
      "loss": 5.4705,
      "mean_token_accuracy": 0.18023901879787446,
      "num_tokens": 34553720.0,
      "step": 15475
    },
    {
      "entropy": 5.57817497253418,
      "epoch": 1.738585949345763,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047062463564723884,
      "loss": 5.3512,
      "mean_token_accuracy": 0.1927105724811554,
      "num_tokens": 34564641.0,
      "step": 15480
    },
    {
      "entropy": 5.629330730438232,
      "epoch": 1.7391475262537206,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004706047977177464,
      "loss": 5.4506,
      "mean_token_accuracy": 0.17920246571302414,
      "num_tokens": 34576470.0,
      "step": 15485
    },
    {
      "entropy": 5.660240888595581,
      "epoch": 1.739709103161678,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047058495356012135,
      "loss": 5.4717,
      "mean_token_accuracy": 0.18309042900800704,
      "num_tokens": 34589158.0,
      "step": 15490
    },
    {
      "entropy": 5.585483837127685,
      "epoch": 1.7402706800696355,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047056510317499545,
      "loss": 5.2674,
      "mean_token_accuracy": 0.19525233805179595,
      "num_tokens": 34599192.0,
      "step": 15495
    },
    {
      "entropy": 5.536190557479858,
      "epoch": 1.7408322569775931,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047054524656300087,
      "loss": 5.3848,
      "mean_token_accuracy": 0.18656711727380754,
      "num_tokens": 34611922.0,
      "step": 15500
    },
    {
      "entropy": 5.612058496475219,
      "epoch": 1.7413938338855506,
      "grad_norm": 1.375,
      "learning_rate": 0.00047052538372476985,
      "loss": 5.4424,
      "mean_token_accuracy": 0.18274508863687516,
      "num_tokens": 34623618.0,
      "step": 15505
    },
    {
      "entropy": 5.63532280921936,
      "epoch": 1.741955410793508,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047050551466093466,
      "loss": 5.448,
      "mean_token_accuracy": 0.18087286949157716,
      "num_tokens": 34635057.0,
      "step": 15510
    },
    {
      "entropy": 5.588735723495484,
      "epoch": 1.7425169877014657,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004704856393721281,
      "loss": 5.3885,
      "mean_token_accuracy": 0.18542587012052536,
      "num_tokens": 34647527.0,
      "step": 15515
    },
    {
      "entropy": 5.525556564331055,
      "epoch": 1.7430785646094233,
      "grad_norm": 2.0625,
      "learning_rate": 0.00047046575785898297,
      "loss": 5.3501,
      "mean_token_accuracy": 0.18491790890693666,
      "num_tokens": 34658615.0,
      "step": 15520
    },
    {
      "entropy": 5.6457047939300535,
      "epoch": 1.7436401415173808,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004704458701221323,
      "loss": 5.3801,
      "mean_token_accuracy": 0.18849466741085052,
      "num_tokens": 34668485.0,
      "step": 15525
    },
    {
      "entropy": 5.581352806091308,
      "epoch": 1.7442017184253382,
      "grad_norm": 1.75,
      "learning_rate": 0.0004704259761622094,
      "loss": 5.3613,
      "mean_token_accuracy": 0.1857620045542717,
      "num_tokens": 34681181.0,
      "step": 15530
    },
    {
      "entropy": 5.473740434646606,
      "epoch": 1.7447632953332959,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047040607597984745,
      "loss": 5.3262,
      "mean_token_accuracy": 0.1866876021027565,
      "num_tokens": 34692954.0,
      "step": 15535
    },
    {
      "entropy": 5.446321678161621,
      "epoch": 1.7453248722412535,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004703861695756804,
      "loss": 5.2443,
      "mean_token_accuracy": 0.19230304658412933,
      "num_tokens": 34705120.0,
      "step": 15540
    },
    {
      "entropy": 5.566729164123535,
      "epoch": 1.745886449149211,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004703662569503419,
      "loss": 5.3736,
      "mean_token_accuracy": 0.1882579505443573,
      "num_tokens": 34716589.0,
      "step": 15545
    },
    {
      "entropy": 5.573837327957153,
      "epoch": 1.7464480260571684,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047034633810446604,
      "loss": 5.3529,
      "mean_token_accuracy": 0.18860515356063842,
      "num_tokens": 34727551.0,
      "step": 15550
    },
    {
      "entropy": 5.502696371078491,
      "epoch": 1.747009602965126,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004703264130386869,
      "loss": 5.3878,
      "mean_token_accuracy": 0.1905357763171196,
      "num_tokens": 34738887.0,
      "step": 15555
    },
    {
      "entropy": 5.505575323104859,
      "epoch": 1.7475711798730837,
      "grad_norm": 2.125,
      "learning_rate": 0.00047030648175363913,
      "loss": 5.324,
      "mean_token_accuracy": 0.1926727384328842,
      "num_tokens": 34750264.0,
      "step": 15560
    },
    {
      "entropy": 5.564579248428345,
      "epoch": 1.7481327567810412,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004702865442499572,
      "loss": 5.401,
      "mean_token_accuracy": 0.18318630903959274,
      "num_tokens": 34761335.0,
      "step": 15565
    },
    {
      "entropy": 5.5248524188995365,
      "epoch": 1.7486943336889986,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004702666005282759,
      "loss": 5.3712,
      "mean_token_accuracy": 0.18553073108196258,
      "num_tokens": 34772237.0,
      "step": 15570
    },
    {
      "entropy": 5.5324851989746096,
      "epoch": 1.7492559105969563,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004702466505892304,
      "loss": 5.2694,
      "mean_token_accuracy": 0.19213333427906037,
      "num_tokens": 34784648.0,
      "step": 15575
    },
    {
      "entropy": 5.604431247711181,
      "epoch": 1.749817487504914,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047022669443345575,
      "loss": 5.3849,
      "mean_token_accuracy": 0.1935294434428215,
      "num_tokens": 34794505.0,
      "step": 15580
    },
    {
      "entropy": 5.572837781906128,
      "epoch": 1.7503790644128714,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047020673206158743,
      "loss": 5.3806,
      "mean_token_accuracy": 0.1911494642496109,
      "num_tokens": 34804887.0,
      "step": 15585
    },
    {
      "entropy": 5.637475967407227,
      "epoch": 1.7509406413208288,
      "grad_norm": 1.984375,
      "learning_rate": 0.000470186763474261,
      "loss": 5.3844,
      "mean_token_accuracy": 0.18640556037425995,
      "num_tokens": 34816814.0,
      "step": 15590
    },
    {
      "entropy": 5.561314725875855,
      "epoch": 1.7515022182287865,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004701667886721125,
      "loss": 5.3015,
      "mean_token_accuracy": 0.19368523210287095,
      "num_tokens": 34827251.0,
      "step": 15595
    },
    {
      "entropy": 5.576226997375488,
      "epoch": 1.752063795136744,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004701468076557776,
      "loss": 5.4243,
      "mean_token_accuracy": 0.18055089116096495,
      "num_tokens": 34839010.0,
      "step": 15600
    },
    {
      "entropy": 5.589766645431519,
      "epoch": 1.7526253720447014,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047012682042589277,
      "loss": 5.4225,
      "mean_token_accuracy": 0.18299496173858643,
      "num_tokens": 34850435.0,
      "step": 15605
    },
    {
      "entropy": 5.545932340621948,
      "epoch": 1.753186948952659,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047010682698309423,
      "loss": 5.2576,
      "mean_token_accuracy": 0.19708157777786256,
      "num_tokens": 34861222.0,
      "step": 15610
    },
    {
      "entropy": 5.637009000778198,
      "epoch": 1.7537485258606167,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004700868273280186,
      "loss": 5.4984,
      "mean_token_accuracy": 0.1692656770348549,
      "num_tokens": 34872650.0,
      "step": 15615
    },
    {
      "entropy": 5.627308702468872,
      "epoch": 1.7543101027685741,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047006682146130273,
      "loss": 5.3025,
      "mean_token_accuracy": 0.19773675948381425,
      "num_tokens": 34883201.0,
      "step": 15620
    },
    {
      "entropy": 5.656827211380005,
      "epoch": 1.7548716796765316,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004700468093835837,
      "loss": 5.5374,
      "mean_token_accuracy": 0.17450018972158432,
      "num_tokens": 34895936.0,
      "step": 15625
    },
    {
      "entropy": 5.571184539794922,
      "epoch": 1.7554332565844892,
      "grad_norm": 1.375,
      "learning_rate": 0.00047002679109549864,
      "loss": 5.3689,
      "mean_token_accuracy": 0.19319046586751937,
      "num_tokens": 34906491.0,
      "step": 15630
    },
    {
      "entropy": 5.555408477783203,
      "epoch": 1.755994833492447,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047000676659768477,
      "loss": 5.3749,
      "mean_token_accuracy": 0.18406349867582322,
      "num_tokens": 34918300.0,
      "step": 15635
    },
    {
      "entropy": 5.471654033660888,
      "epoch": 1.7565564104004043,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004699867358907799,
      "loss": 5.3172,
      "mean_token_accuracy": 0.1846435323357582,
      "num_tokens": 34928573.0,
      "step": 15640
    },
    {
      "entropy": 5.605314159393311,
      "epoch": 1.7571179873083618,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004699666989754217,
      "loss": 5.3651,
      "mean_token_accuracy": 0.18511402755975723,
      "num_tokens": 34940378.0,
      "step": 15645
    },
    {
      "entropy": 5.564402055740357,
      "epoch": 1.7576795642163194,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004699466558522482,
      "loss": 5.4219,
      "mean_token_accuracy": 0.18588894307613374,
      "num_tokens": 34951627.0,
      "step": 15650
    },
    {
      "entropy": 5.560405826568603,
      "epoch": 1.758241141124277,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046992660652189755,
      "loss": 5.3548,
      "mean_token_accuracy": 0.1835366114974022,
      "num_tokens": 34961428.0,
      "step": 15655
    },
    {
      "entropy": 5.6877467155456545,
      "epoch": 1.7588027180322345,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046990655098500824,
      "loss": 5.4902,
      "mean_token_accuracy": 0.18185962587594987,
      "num_tokens": 34973736.0,
      "step": 15660
    },
    {
      "entropy": 5.570611047744751,
      "epoch": 1.759364294940192,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046988648924221866,
      "loss": 5.3559,
      "mean_token_accuracy": 0.19374988228082657,
      "num_tokens": 34984510.0,
      "step": 15665
    },
    {
      "entropy": 5.577480983734131,
      "epoch": 1.7599258718481496,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004698664212941677,
      "loss": 5.3475,
      "mean_token_accuracy": 0.1899129033088684,
      "num_tokens": 34995218.0,
      "step": 15670
    },
    {
      "entropy": 5.426718807220459,
      "epoch": 1.7604874487561073,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046984634714149437,
      "loss": 5.1614,
      "mean_token_accuracy": 0.19524444937705993,
      "num_tokens": 35006681.0,
      "step": 15675
    },
    {
      "entropy": 5.510039281845093,
      "epoch": 1.7610490256640647,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004698262667848377,
      "loss": 5.4053,
      "mean_token_accuracy": 0.18200305998325347,
      "num_tokens": 35017341.0,
      "step": 15680
    },
    {
      "entropy": 5.5395721912384035,
      "epoch": 1.7616106025720222,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046980618022483714,
      "loss": 5.2968,
      "mean_token_accuracy": 0.19315295964479445,
      "num_tokens": 35027718.0,
      "step": 15685
    },
    {
      "entropy": 5.602697229385376,
      "epoch": 1.7621721794799798,
      "grad_norm": 1.734375,
      "learning_rate": 0.00046978608746213224,
      "loss": 5.3811,
      "mean_token_accuracy": 0.19060612320899964,
      "num_tokens": 35038100.0,
      "step": 15690
    },
    {
      "entropy": 5.52765793800354,
      "epoch": 1.7627337563879373,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004697659884973628,
      "loss": 5.4062,
      "mean_token_accuracy": 0.18741343915462494,
      "num_tokens": 35049077.0,
      "step": 15695
    },
    {
      "entropy": 5.6450201034545895,
      "epoch": 1.7632953332958947,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004697458833311687,
      "loss": 5.5438,
      "mean_token_accuracy": 0.1795817032456398,
      "num_tokens": 35060685.0,
      "step": 15700
    },
    {
      "entropy": 5.549354219436646,
      "epoch": 1.7638569102038524,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004697257719641902,
      "loss": 5.3062,
      "mean_token_accuracy": 0.19198077917099,
      "num_tokens": 35072122.0,
      "step": 15705
    },
    {
      "entropy": 5.583201789855957,
      "epoch": 1.76441848711181,
      "grad_norm": 1.375,
      "learning_rate": 0.0004697056543970676,
      "loss": 5.3449,
      "mean_token_accuracy": 0.19073683619499207,
      "num_tokens": 35082427.0,
      "step": 15710
    },
    {
      "entropy": 5.568373775482177,
      "epoch": 1.7649800640197675,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046968553063044146,
      "loss": 5.3858,
      "mean_token_accuracy": 0.1903146043419838,
      "num_tokens": 35093061.0,
      "step": 15715
    },
    {
      "entropy": 5.621543550491333,
      "epoch": 1.765541640927725,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046966540066495247,
      "loss": 5.4004,
      "mean_token_accuracy": 0.18613336980342865,
      "num_tokens": 35103226.0,
      "step": 15720
    },
    {
      "entropy": 5.614335441589356,
      "epoch": 1.7661032178356826,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00046964526450124164,
      "loss": 5.3784,
      "mean_token_accuracy": 0.18761180490255355,
      "num_tokens": 35113977.0,
      "step": 15725
    },
    {
      "entropy": 5.5389328956604,
      "epoch": 1.7666647947436402,
      "grad_norm": 1.953125,
      "learning_rate": 0.00046962512213995,
      "loss": 5.361,
      "mean_token_accuracy": 0.18948362320661544,
      "num_tokens": 35125132.0,
      "step": 15730
    },
    {
      "entropy": 5.5051295280456545,
      "epoch": 1.7672263716515977,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046960497358171913,
      "loss": 5.3016,
      "mean_token_accuracy": 0.18765170276165008,
      "num_tokens": 35135266.0,
      "step": 15735
    },
    {
      "entropy": 5.596678066253662,
      "epoch": 1.7677879485595551,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046958481882719035,
      "loss": 5.3735,
      "mean_token_accuracy": 0.19410445392131806,
      "num_tokens": 35146268.0,
      "step": 15740
    },
    {
      "entropy": 5.561468648910522,
      "epoch": 1.7683495254675128,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046956465787700554,
      "loss": 5.297,
      "mean_token_accuracy": 0.1890968218445778,
      "num_tokens": 35157046.0,
      "step": 15745
    },
    {
      "entropy": 5.57606053352356,
      "epoch": 1.7689111023754704,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046954449073180644,
      "loss": 5.4801,
      "mean_token_accuracy": 0.18432846516370774,
      "num_tokens": 35167624.0,
      "step": 15750
    },
    {
      "entropy": 5.592984199523926,
      "epoch": 1.7694726792834279,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004695243173922354,
      "loss": 5.4458,
      "mean_token_accuracy": 0.18281312584877013,
      "num_tokens": 35178987.0,
      "step": 15755
    },
    {
      "entropy": 5.515202236175537,
      "epoch": 1.7700342561913853,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046950413785893465,
      "loss": 5.3005,
      "mean_token_accuracy": 0.1893632858991623,
      "num_tokens": 35189804.0,
      "step": 15760
    },
    {
      "entropy": 5.620057249069214,
      "epoch": 1.770595833099343,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046948395213254667,
      "loss": 5.363,
      "mean_token_accuracy": 0.18398898243904113,
      "num_tokens": 35200677.0,
      "step": 15765
    },
    {
      "entropy": 5.463645362854004,
      "epoch": 1.7711574100073006,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004694637602137143,
      "loss": 5.3334,
      "mean_token_accuracy": 0.190482397377491,
      "num_tokens": 35211570.0,
      "step": 15770
    },
    {
      "entropy": 5.477284479141235,
      "epoch": 1.771718986915258,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004694435621030803,
      "loss": 5.3739,
      "mean_token_accuracy": 0.1894892379641533,
      "num_tokens": 35222451.0,
      "step": 15775
    },
    {
      "entropy": 5.678949451446533,
      "epoch": 1.7722805638232155,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046942335780128785,
      "loss": 5.4806,
      "mean_token_accuracy": 0.18306827545166016,
      "num_tokens": 35233588.0,
      "step": 15780
    },
    {
      "entropy": 5.648662281036377,
      "epoch": 1.7728421407311732,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046940314730898035,
      "loss": 5.4473,
      "mean_token_accuracy": 0.17744902074337005,
      "num_tokens": 35245766.0,
      "step": 15785
    },
    {
      "entropy": 5.507218599319458,
      "epoch": 1.7734037176391306,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004693829306268012,
      "loss": 5.3599,
      "mean_token_accuracy": 0.19044422805309297,
      "num_tokens": 35255635.0,
      "step": 15790
    },
    {
      "entropy": 5.595747947692871,
      "epoch": 1.773965294547088,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004693627077553942,
      "loss": 5.3711,
      "mean_token_accuracy": 0.1829087480902672,
      "num_tokens": 35269342.0,
      "step": 15795
    },
    {
      "entropy": 5.600680637359619,
      "epoch": 1.7745268714550457,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004693424786954031,
      "loss": 5.3806,
      "mean_token_accuracy": 0.18079588264226915,
      "num_tokens": 35280735.0,
      "step": 15800
    },
    {
      "entropy": 5.570738458633423,
      "epoch": 1.7750884483630034,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046932224344747224,
      "loss": 5.3907,
      "mean_token_accuracy": 0.1817556396126747,
      "num_tokens": 35290803.0,
      "step": 15805
    },
    {
      "entropy": 5.615657567977905,
      "epoch": 1.7756500252709608,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046930200201224573,
      "loss": 5.3779,
      "mean_token_accuracy": 0.1841071903705597,
      "num_tokens": 35301811.0,
      "step": 15810
    },
    {
      "entropy": 5.597724437713623,
      "epoch": 1.7762116021789183,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004692817543903681,
      "loss": 5.4635,
      "mean_token_accuracy": 0.17897940576076507,
      "num_tokens": 35313374.0,
      "step": 15815
    },
    {
      "entropy": 5.610829925537109,
      "epoch": 1.776773179086876,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000469261500582484,
      "loss": 5.324,
      "mean_token_accuracy": 0.19385354369878768,
      "num_tokens": 35323272.0,
      "step": 15820
    },
    {
      "entropy": 5.6619956493377686,
      "epoch": 1.7773347559948336,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046924124058923843,
      "loss": 5.458,
      "mean_token_accuracy": 0.1730327844619751,
      "num_tokens": 35335720.0,
      "step": 15825
    },
    {
      "entropy": 5.5761558532714846,
      "epoch": 1.777896332902791,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004692209744112765,
      "loss": 5.4976,
      "mean_token_accuracy": 0.1847943678498268,
      "num_tokens": 35347621.0,
      "step": 15830
    },
    {
      "entropy": 5.718498373031617,
      "epoch": 1.7784579098107485,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004692007020492433,
      "loss": 5.5732,
      "mean_token_accuracy": 0.17302506268024445,
      "num_tokens": 35360611.0,
      "step": 15835
    },
    {
      "entropy": 5.602990007400512,
      "epoch": 1.7790194867187061,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004691804235037845,
      "loss": 5.4609,
      "mean_token_accuracy": 0.18267596065998076,
      "num_tokens": 35372242.0,
      "step": 15840
    },
    {
      "entropy": 5.611075305938721,
      "epoch": 1.7795810636266638,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004691601387755456,
      "loss": 5.3618,
      "mean_token_accuracy": 0.19045183807611465,
      "num_tokens": 35385718.0,
      "step": 15845
    },
    {
      "entropy": 5.603568363189697,
      "epoch": 1.7801426405346212,
      "grad_norm": 1.25,
      "learning_rate": 0.00046913984786517275,
      "loss": 5.4757,
      "mean_token_accuracy": 0.17845510244369506,
      "num_tokens": 35398298.0,
      "step": 15850
    },
    {
      "entropy": 5.565671443939209,
      "epoch": 1.7807042174425787,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046911955077331177,
      "loss": 5.2554,
      "mean_token_accuracy": 0.1875331446528435,
      "num_tokens": 35408765.0,
      "step": 15855
    },
    {
      "entropy": 5.49656195640564,
      "epoch": 1.7812657943505363,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046909924750060893,
      "loss": 5.2602,
      "mean_token_accuracy": 0.19083157181739807,
      "num_tokens": 35418547.0,
      "step": 15860
    },
    {
      "entropy": 5.5733802795410154,
      "epoch": 1.781827371258494,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004690789380477109,
      "loss": 5.381,
      "mean_token_accuracy": 0.1814642384648323,
      "num_tokens": 35428743.0,
      "step": 15865
    },
    {
      "entropy": 5.482580661773682,
      "epoch": 1.7823889481664514,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004690586224152641,
      "loss": 5.2973,
      "mean_token_accuracy": 0.1893284648656845,
      "num_tokens": 35439557.0,
      "step": 15870
    },
    {
      "entropy": 5.544387578964233,
      "epoch": 1.7829505250744089,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004690383006039156,
      "loss": 5.4617,
      "mean_token_accuracy": 0.18211542963981628,
      "num_tokens": 35452041.0,
      "step": 15875
    },
    {
      "entropy": 5.624589967727661,
      "epoch": 1.7835121019823665,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004690179726143123,
      "loss": 5.3825,
      "mean_token_accuracy": 0.19349778741598128,
      "num_tokens": 35463283.0,
      "step": 15880
    },
    {
      "entropy": 5.676436138153076,
      "epoch": 1.784073678890324,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004689976384471014,
      "loss": 5.5068,
      "mean_token_accuracy": 0.1909259706735611,
      "num_tokens": 35475802.0,
      "step": 15885
    },
    {
      "entropy": 5.500989723205566,
      "epoch": 1.7846352557982814,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004689772981029306,
      "loss": 5.3019,
      "mean_token_accuracy": 0.1928515240550041,
      "num_tokens": 35486322.0,
      "step": 15890
    },
    {
      "entropy": 5.553930282592773,
      "epoch": 1.785196832706239,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004689569515824473,
      "loss": 5.3452,
      "mean_token_accuracy": 0.19019032567739486,
      "num_tokens": 35497021.0,
      "step": 15895
    },
    {
      "entropy": 5.555098581314087,
      "epoch": 1.7857584096141967,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004689365988862995,
      "loss": 5.3244,
      "mean_token_accuracy": 0.1960470959544182,
      "num_tokens": 35507579.0,
      "step": 15900
    },
    {
      "entropy": 5.478539276123047,
      "epoch": 1.7863199865221542,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046891624001513515,
      "loss": 5.2773,
      "mean_token_accuracy": 0.18713027983903885,
      "num_tokens": 35518131.0,
      "step": 15905
    },
    {
      "entropy": 5.560598850250244,
      "epoch": 1.7868815634301116,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046889587496960257,
      "loss": 5.4211,
      "mean_token_accuracy": 0.18091222941875457,
      "num_tokens": 35529577.0,
      "step": 15910
    },
    {
      "entropy": 5.609899806976318,
      "epoch": 1.7874431403380693,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046887550375034997,
      "loss": 5.3448,
      "mean_token_accuracy": 0.1910657912492752,
      "num_tokens": 35541555.0,
      "step": 15915
    },
    {
      "entropy": 5.5346800804138185,
      "epoch": 1.788004717246027,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046885512635802627,
      "loss": 5.2772,
      "mean_token_accuracy": 0.1943323016166687,
      "num_tokens": 35552596.0,
      "step": 15920
    },
    {
      "entropy": 5.407305765151977,
      "epoch": 1.7885662941539844,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004688347427932801,
      "loss": 5.3629,
      "mean_token_accuracy": 0.18243516087532044,
      "num_tokens": 35564962.0,
      "step": 15925
    },
    {
      "entropy": 5.618680953979492,
      "epoch": 1.7891278710619418,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046881435305676055,
      "loss": 5.4494,
      "mean_token_accuracy": 0.17935291826725006,
      "num_tokens": 35576436.0,
      "step": 15930
    },
    {
      "entropy": 5.633065271377563,
      "epoch": 1.7896894479698995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046879395714911686,
      "loss": 5.3908,
      "mean_token_accuracy": 0.18873339146375656,
      "num_tokens": 35587467.0,
      "step": 15935
    },
    {
      "entropy": 5.633873844146729,
      "epoch": 1.7902510248778571,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046877355507099846,
      "loss": 5.4434,
      "mean_token_accuracy": 0.17776518166065217,
      "num_tokens": 35600302.0,
      "step": 15940
    },
    {
      "entropy": 5.563335609436035,
      "epoch": 1.7908126017858146,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004687531468230549,
      "loss": 5.3727,
      "mean_token_accuracy": 0.1925714984536171,
      "num_tokens": 35611757.0,
      "step": 15945
    },
    {
      "entropy": 5.531421899795532,
      "epoch": 1.791374178693772,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004687327324059359,
      "loss": 5.4276,
      "mean_token_accuracy": 0.18253857344388963,
      "num_tokens": 35621991.0,
      "step": 15950
    },
    {
      "entropy": 5.577724266052246,
      "epoch": 1.7919357556017297,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004687123118202916,
      "loss": 5.3962,
      "mean_token_accuracy": 0.18589919358491896,
      "num_tokens": 35633744.0,
      "step": 15955
    },
    {
      "entropy": 5.53998908996582,
      "epoch": 1.7924973325096873,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004686918850667722,
      "loss": 5.2911,
      "mean_token_accuracy": 0.1892552986741066,
      "num_tokens": 35645523.0,
      "step": 15960
    },
    {
      "entropy": 5.590689468383789,
      "epoch": 1.7930589094176448,
      "grad_norm": 1.3125,
      "learning_rate": 0.000468671452146028,
      "loss": 5.4468,
      "mean_token_accuracy": 0.1826014667749405,
      "num_tokens": 35657670.0,
      "step": 15965
    },
    {
      "entropy": 5.529779481887817,
      "epoch": 1.7936204863256022,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046865101305870965,
      "loss": 5.2516,
      "mean_token_accuracy": 0.1899636507034302,
      "num_tokens": 35668149.0,
      "step": 15970
    },
    {
      "entropy": 5.494525718688965,
      "epoch": 1.7941820632335599,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000468630567805468,
      "loss": 5.34,
      "mean_token_accuracy": 0.18510247766971588,
      "num_tokens": 35680225.0,
      "step": 15975
    },
    {
      "entropy": 5.430608177185059,
      "epoch": 1.7947436401415173,
      "grad_norm": 1.65625,
      "learning_rate": 0.000468610116386954,
      "loss": 5.249,
      "mean_token_accuracy": 0.195359405875206,
      "num_tokens": 35691042.0,
      "step": 15980
    },
    {
      "entropy": 5.530224704742432,
      "epoch": 1.7953052170494748,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004685896588038187,
      "loss": 5.3675,
      "mean_token_accuracy": 0.1856921836733818,
      "num_tokens": 35702895.0,
      "step": 15985
    },
    {
      "entropy": 5.588540363311767,
      "epoch": 1.7958667939574324,
      "grad_norm": 1.5,
      "learning_rate": 0.0004685691950567136,
      "loss": 5.3112,
      "mean_token_accuracy": 0.18628643453121185,
      "num_tokens": 35714765.0,
      "step": 15990
    },
    {
      "entropy": 5.632132482528687,
      "epoch": 1.79642837086539,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046854872514629035,
      "loss": 5.4812,
      "mean_token_accuracy": 0.1774725615978241,
      "num_tokens": 35725832.0,
      "step": 15995
    },
    {
      "entropy": 5.540455150604248,
      "epoch": 1.7969899477733475,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004685282490732005,
      "loss": 5.3074,
      "mean_token_accuracy": 0.19379405677318573,
      "num_tokens": 35736619.0,
      "step": 16000
    },
    {
      "entropy": 5.50766134262085,
      "epoch": 1.797551524681305,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046850776683809623,
      "loss": 5.262,
      "mean_token_accuracy": 0.18469182550907134,
      "num_tokens": 35748243.0,
      "step": 16005
    },
    {
      "entropy": 5.541708898544312,
      "epoch": 1.7981131015892626,
      "grad_norm": 1.25,
      "learning_rate": 0.0004684872784416295,
      "loss": 5.3748,
      "mean_token_accuracy": 0.18503991067409514,
      "num_tokens": 35759126.0,
      "step": 16010
    },
    {
      "entropy": 5.648258066177368,
      "epoch": 1.7986746784972203,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004684667838844529,
      "loss": 5.3932,
      "mean_token_accuracy": 0.1819484680891037,
      "num_tokens": 35770998.0,
      "step": 16015
    },
    {
      "entropy": 5.525299501419068,
      "epoch": 1.7992362554051777,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004684462831672188,
      "loss": 5.329,
      "mean_token_accuracy": 0.1874153807759285,
      "num_tokens": 35782835.0,
      "step": 16020
    },
    {
      "entropy": 5.549390172958374,
      "epoch": 1.7997978323131352,
      "grad_norm": 1.375,
      "learning_rate": 0.00046842577629058006,
      "loss": 5.3128,
      "mean_token_accuracy": 0.1882465064525604,
      "num_tokens": 35792828.0,
      "step": 16025
    },
    {
      "entropy": 5.642633104324341,
      "epoch": 1.8003594092210928,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004684052632551896,
      "loss": 5.5487,
      "mean_token_accuracy": 0.1810038223862648,
      "num_tokens": 35804449.0,
      "step": 16030
    },
    {
      "entropy": 5.574741077423096,
      "epoch": 1.8009209861290505,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046838474406170046,
      "loss": 5.302,
      "mean_token_accuracy": 0.18773899972438812,
      "num_tokens": 35815903.0,
      "step": 16035
    },
    {
      "entropy": 5.491513681411743,
      "epoch": 1.801482563037008,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046836421871076616,
      "loss": 5.3141,
      "mean_token_accuracy": 0.19241884350776672,
      "num_tokens": 35827872.0,
      "step": 16040
    },
    {
      "entropy": 5.521188020706177,
      "epoch": 1.8020441399449654,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004683436872030401,
      "loss": 5.3027,
      "mean_token_accuracy": 0.19239161610603334,
      "num_tokens": 35838864.0,
      "step": 16045
    },
    {
      "entropy": 5.490436601638794,
      "epoch": 1.802605716852923,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046832314953917605,
      "loss": 5.2212,
      "mean_token_accuracy": 0.18229456543922423,
      "num_tokens": 35850178.0,
      "step": 16050
    },
    {
      "entropy": 5.544579124450683,
      "epoch": 1.8031672937608807,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046830260571982795,
      "loss": 5.3763,
      "mean_token_accuracy": 0.1876939445734024,
      "num_tokens": 35861904.0,
      "step": 16055
    },
    {
      "entropy": 5.4166254043579105,
      "epoch": 1.8037288706688381,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004682820557456499,
      "loss": 5.2574,
      "mean_token_accuracy": 0.19551395773887634,
      "num_tokens": 35872977.0,
      "step": 16060
    },
    {
      "entropy": 5.621083164215088,
      "epoch": 1.8042904475767956,
      "grad_norm": 1.125,
      "learning_rate": 0.00046826149961729625,
      "loss": 5.4566,
      "mean_token_accuracy": 0.18301086127758026,
      "num_tokens": 35886094.0,
      "step": 16065
    },
    {
      "entropy": 5.537100791931152,
      "epoch": 1.8048520244847532,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046824093733542144,
      "loss": 5.279,
      "mean_token_accuracy": 0.19515191167593002,
      "num_tokens": 35897557.0,
      "step": 16070
    },
    {
      "entropy": 5.579884958267212,
      "epoch": 1.8054136013927107,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046822036890068033,
      "loss": 5.4973,
      "mean_token_accuracy": 0.17925168871879577,
      "num_tokens": 35910605.0,
      "step": 16075
    },
    {
      "entropy": 5.554397964477539,
      "epoch": 1.8059751783006681,
      "grad_norm": 1.25,
      "learning_rate": 0.00046819979431372773,
      "loss": 5.2688,
      "mean_token_accuracy": 0.1950139656662941,
      "num_tokens": 35921062.0,
      "step": 16080
    },
    {
      "entropy": 5.548288822174072,
      "epoch": 1.8065367552086258,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046817921357521865,
      "loss": 5.3555,
      "mean_token_accuracy": 0.1850862443447113,
      "num_tokens": 35932385.0,
      "step": 16085
    },
    {
      "entropy": 5.47045931816101,
      "epoch": 1.8070983321165834,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046815862668580854,
      "loss": 5.3434,
      "mean_token_accuracy": 0.19163644313812256,
      "num_tokens": 35943628.0,
      "step": 16090
    },
    {
      "entropy": 5.60513014793396,
      "epoch": 1.8076599090245409,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004681380336461528,
      "loss": 5.4026,
      "mean_token_accuracy": 0.1782854288816452,
      "num_tokens": 35955074.0,
      "step": 16095
    },
    {
      "entropy": 5.593359375,
      "epoch": 1.8082214859324983,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046811743445690724,
      "loss": 5.3292,
      "mean_token_accuracy": 0.1858907625079155,
      "num_tokens": 35966200.0,
      "step": 16100
    },
    {
      "entropy": 5.467005634307862,
      "epoch": 1.808783062840456,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004680968291187276,
      "loss": 5.2271,
      "mean_token_accuracy": 0.19870145320892335,
      "num_tokens": 35975638.0,
      "step": 16105
    },
    {
      "entropy": 5.513053369522095,
      "epoch": 1.8093446397484136,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004680762176322701,
      "loss": 5.3487,
      "mean_token_accuracy": 0.1926051288843155,
      "num_tokens": 35986486.0,
      "step": 16110
    },
    {
      "entropy": 5.490559530258179,
      "epoch": 1.809906216656371,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046805559999819093,
      "loss": 5.2737,
      "mean_token_accuracy": 0.19124430269002915,
      "num_tokens": 35996612.0,
      "step": 16115
    },
    {
      "entropy": 5.415689134597779,
      "epoch": 1.8104677935643285,
      "grad_norm": 1.375,
      "learning_rate": 0.0004680349762171465,
      "loss": 5.3365,
      "mean_token_accuracy": 0.18895064294338226,
      "num_tokens": 36006936.0,
      "step": 16120
    },
    {
      "entropy": 5.6507463455200195,
      "epoch": 1.8110293704722862,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004680143462897936,
      "loss": 5.461,
      "mean_token_accuracy": 0.18231729716062545,
      "num_tokens": 36019562.0,
      "step": 16125
    },
    {
      "entropy": 5.540613079071045,
      "epoch": 1.8115909473802438,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004679937102167891,
      "loss": 5.2565,
      "mean_token_accuracy": 0.19357980042696,
      "num_tokens": 36029585.0,
      "step": 16130
    },
    {
      "entropy": 5.472490930557251,
      "epoch": 1.8121525242882013,
      "grad_norm": 1.25,
      "learning_rate": 0.00046797306799879,
      "loss": 5.3391,
      "mean_token_accuracy": 0.19227437376976014,
      "num_tokens": 36041495.0,
      "step": 16135
    },
    {
      "entropy": 5.490143156051635,
      "epoch": 1.8127141011961587,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046795241963645356,
      "loss": 5.3793,
      "mean_token_accuracy": 0.18855390399694444,
      "num_tokens": 36052087.0,
      "step": 16140
    },
    {
      "entropy": 5.549077892303467,
      "epoch": 1.8132756781041164,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004679317651304373,
      "loss": 5.3439,
      "mean_token_accuracy": 0.1903028041124344,
      "num_tokens": 36065011.0,
      "step": 16145
    },
    {
      "entropy": 5.59675784111023,
      "epoch": 1.813837255012074,
      "grad_norm": 1.625,
      "learning_rate": 0.0004679111044813987,
      "loss": 5.4232,
      "mean_token_accuracy": 0.18762087523937226,
      "num_tokens": 36075912.0,
      "step": 16150
    },
    {
      "entropy": 5.610179901123047,
      "epoch": 1.8143988319200315,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046789043768999573,
      "loss": 5.3904,
      "mean_token_accuracy": 0.18435035198926925,
      "num_tokens": 36086757.0,
      "step": 16155
    },
    {
      "entropy": 5.515399742126465,
      "epoch": 1.814960408827989,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004678697647568865,
      "loss": 5.2643,
      "mean_token_accuracy": 0.19338752776384355,
      "num_tokens": 36097319.0,
      "step": 16160
    },
    {
      "entropy": 5.522013235092163,
      "epoch": 1.8155219857359466,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00046784908568272903,
      "loss": 5.4083,
      "mean_token_accuracy": 0.18334012925624849,
      "num_tokens": 36108416.0,
      "step": 16165
    },
    {
      "entropy": 5.463136053085327,
      "epoch": 1.8160835626439042,
      "grad_norm": 1.296875,
      "learning_rate": 0.000467828400468182,
      "loss": 5.3351,
      "mean_token_accuracy": 0.18444680869579316,
      "num_tokens": 36120833.0,
      "step": 16170
    },
    {
      "entropy": 5.610969066619873,
      "epoch": 1.8166451395518615,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046780770911390387,
      "loss": 5.4616,
      "mean_token_accuracy": 0.18258167505264283,
      "num_tokens": 36132348.0,
      "step": 16175
    },
    {
      "entropy": 5.541339874267578,
      "epoch": 1.8172067164598191,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004677870116205534,
      "loss": 5.2378,
      "mean_token_accuracy": 0.19122178256511688,
      "num_tokens": 36142829.0,
      "step": 16180
    },
    {
      "entropy": 5.566759204864502,
      "epoch": 1.8177682933677768,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046776630798878985,
      "loss": 5.5088,
      "mean_token_accuracy": 0.18053208142518998,
      "num_tokens": 36155931.0,
      "step": 16185
    },
    {
      "entropy": 5.606925439834595,
      "epoch": 1.8183298702757342,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004677455982192723,
      "loss": 5.4102,
      "mean_token_accuracy": 0.17741843014955522,
      "num_tokens": 36167591.0,
      "step": 16190
    },
    {
      "entropy": 5.513416719436646,
      "epoch": 1.8188914471836917,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046772488231266004,
      "loss": 5.2858,
      "mean_token_accuracy": 0.1894127681851387,
      "num_tokens": 36178867.0,
      "step": 16195
    },
    {
      "entropy": 5.39373745918274,
      "epoch": 1.8194530240916493,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046770416026961284,
      "loss": 5.2578,
      "mean_token_accuracy": 0.2004848062992096,
      "num_tokens": 36190691.0,
      "step": 16200
    },
    {
      "entropy": 5.625265455245971,
      "epoch": 1.820014600999607,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004676834320907904,
      "loss": 5.4731,
      "mean_token_accuracy": 0.18121488094329835,
      "num_tokens": 36202222.0,
      "step": 16205
    },
    {
      "entropy": 5.610524082183838,
      "epoch": 1.8205761779075644,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004676626977768528,
      "loss": 5.2545,
      "mean_token_accuracy": 0.19390959143638611,
      "num_tokens": 36214303.0,
      "step": 16210
    },
    {
      "entropy": 5.452145576477051,
      "epoch": 1.8211377548155219,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046764195732846013,
      "loss": 5.2808,
      "mean_token_accuracy": 0.18949127495288848,
      "num_tokens": 36224440.0,
      "step": 16215
    },
    {
      "entropy": 5.421571397781372,
      "epoch": 1.8216993317234795,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046762121074627285,
      "loss": 5.2891,
      "mean_token_accuracy": 0.1903606489300728,
      "num_tokens": 36235445.0,
      "step": 16220
    },
    {
      "entropy": 5.569859313964844,
      "epoch": 1.8222609086314372,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046760045803095153,
      "loss": 5.3156,
      "mean_token_accuracy": 0.19223127663135528,
      "num_tokens": 36247024.0,
      "step": 16225
    },
    {
      "entropy": 5.606548643112182,
      "epoch": 1.8228224855393946,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004675796991831569,
      "loss": 5.3367,
      "mean_token_accuracy": 0.18688896000385286,
      "num_tokens": 36257406.0,
      "step": 16230
    },
    {
      "entropy": 5.524360275268554,
      "epoch": 1.823384062447352,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046755893420355,
      "loss": 5.3498,
      "mean_token_accuracy": 0.18490929454565047,
      "num_tokens": 36270266.0,
      "step": 16235
    },
    {
      "entropy": 5.566007661819458,
      "epoch": 1.8239456393553097,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004675381630927919,
      "loss": 5.4291,
      "mean_token_accuracy": 0.1858387365937233,
      "num_tokens": 36281470.0,
      "step": 16240
    },
    {
      "entropy": 5.534808826446533,
      "epoch": 1.8245072162632674,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046751738585154405,
      "loss": 5.3248,
      "mean_token_accuracy": 0.18495461493730544,
      "num_tokens": 36293315.0,
      "step": 16245
    },
    {
      "entropy": 5.557805871963501,
      "epoch": 1.8250687931712248,
      "grad_norm": 1.5,
      "learning_rate": 0.00046749660248046794,
      "loss": 5.3685,
      "mean_token_accuracy": 0.19024549275636674,
      "num_tokens": 36304688.0,
      "step": 16250
    },
    {
      "entropy": 5.6081033706665036,
      "epoch": 1.8256303700791823,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046747581298022537,
      "loss": 5.434,
      "mean_token_accuracy": 0.18934669643640517,
      "num_tokens": 36315773.0,
      "step": 16255
    },
    {
      "entropy": 5.626693439483643,
      "epoch": 1.82619194698714,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004674550173514782,
      "loss": 5.3524,
      "mean_token_accuracy": 0.1872216284275055,
      "num_tokens": 36327914.0,
      "step": 16260
    },
    {
      "entropy": 5.603176641464233,
      "epoch": 1.8267535238950976,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046743421559488875,
      "loss": 5.4365,
      "mean_token_accuracy": 0.17932053208351134,
      "num_tokens": 36339617.0,
      "step": 16265
    },
    {
      "entropy": 5.630232667922973,
      "epoch": 1.8273151008030548,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046741340771111917,
      "loss": 5.4864,
      "mean_token_accuracy": 0.17964668124914168,
      "num_tokens": 36350768.0,
      "step": 16270
    },
    {
      "entropy": 5.530461740493775,
      "epoch": 1.8278766777110125,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046739259370083205,
      "loss": 5.3582,
      "mean_token_accuracy": 0.18999661654233932,
      "num_tokens": 36361267.0,
      "step": 16275
    },
    {
      "entropy": 5.583748197555542,
      "epoch": 1.8284382546189701,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004673717735646902,
      "loss": 5.389,
      "mean_token_accuracy": 0.18332965970039367,
      "num_tokens": 36373259.0,
      "step": 16280
    },
    {
      "entropy": 5.590427303314209,
      "epoch": 1.8289998315269276,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004673509473033564,
      "loss": 5.3283,
      "mean_token_accuracy": 0.18604336231946944,
      "num_tokens": 36384295.0,
      "step": 16285
    },
    {
      "entropy": 5.571167469024658,
      "epoch": 1.829561408434885,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046733011491749383,
      "loss": 5.3173,
      "mean_token_accuracy": 0.19359821528196336,
      "num_tokens": 36394928.0,
      "step": 16290
    },
    {
      "entropy": 5.515721654891967,
      "epoch": 1.8301229853428427,
      "grad_norm": 1.25,
      "learning_rate": 0.0004673092764077658,
      "loss": 5.3815,
      "mean_token_accuracy": 0.19220682084560395,
      "num_tokens": 36407038.0,
      "step": 16295
    },
    {
      "entropy": 5.661236715316773,
      "epoch": 1.8306845622508003,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004672884317748359,
      "loss": 5.3604,
      "mean_token_accuracy": 0.18396896719932557,
      "num_tokens": 36418221.0,
      "step": 16300
    },
    {
      "entropy": 5.525816154479981,
      "epoch": 1.8312461391587578,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004672675810193677,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1906868487596512,
      "num_tokens": 36429036.0,
      "step": 16305
    },
    {
      "entropy": 5.512045383453369,
      "epoch": 1.8318077160667152,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046724672414202514,
      "loss": 5.3631,
      "mean_token_accuracy": 0.18964605480432511,
      "num_tokens": 36439851.0,
      "step": 16310
    },
    {
      "entropy": 5.591065883636475,
      "epoch": 1.8323692929746729,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004672258611434724,
      "loss": 5.3747,
      "mean_token_accuracy": 0.18101801127195358,
      "num_tokens": 36451612.0,
      "step": 16315
    },
    {
      "entropy": 5.631992769241333,
      "epoch": 1.8329308698826305,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046720499202437354,
      "loss": 5.4328,
      "mean_token_accuracy": 0.1815750315785408,
      "num_tokens": 36462883.0,
      "step": 16320
    },
    {
      "entropy": 5.5189462184906,
      "epoch": 1.833492446790588,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004671841167853933,
      "loss": 5.324,
      "mean_token_accuracy": 0.19218890368938446,
      "num_tokens": 36473609.0,
      "step": 16325
    },
    {
      "entropy": 5.573424434661865,
      "epoch": 1.8340540236985454,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004671632354271962,
      "loss": 5.3145,
      "mean_token_accuracy": 0.18836932480335236,
      "num_tokens": 36484058.0,
      "step": 16330
    },
    {
      "entropy": 5.640398120880127,
      "epoch": 1.834615600606503,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046714234795044724,
      "loss": 5.36,
      "mean_token_accuracy": 0.18595891892910005,
      "num_tokens": 36496445.0,
      "step": 16335
    },
    {
      "entropy": 5.503306198120117,
      "epoch": 1.8351771775144607,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004671214543558113,
      "loss": 5.2861,
      "mean_token_accuracy": 0.19019471257925033,
      "num_tokens": 36507553.0,
      "step": 16340
    },
    {
      "entropy": 5.503566837310791,
      "epoch": 1.8357387544224182,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004671005546439538,
      "loss": 5.3336,
      "mean_token_accuracy": 0.1920084074139595,
      "num_tokens": 36517887.0,
      "step": 16345
    },
    {
      "entropy": 5.527010250091553,
      "epoch": 1.8363003313303756,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046707964881554005,
      "loss": 5.2848,
      "mean_token_accuracy": 0.18495176434516908,
      "num_tokens": 36529185.0,
      "step": 16350
    },
    {
      "entropy": 5.528638744354248,
      "epoch": 1.8368619082383333,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046705873687123586,
      "loss": 5.3722,
      "mean_token_accuracy": 0.19100251346826552,
      "num_tokens": 36539473.0,
      "step": 16355
    },
    {
      "entropy": 5.5641679763793945,
      "epoch": 1.837423485146291,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046703781881170703,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18921186923980712,
      "num_tokens": 36549603.0,
      "step": 16360
    },
    {
      "entropy": 5.534250068664551,
      "epoch": 1.8379850620542482,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046701689463761947,
      "loss": 5.3379,
      "mean_token_accuracy": 0.19235465973615645,
      "num_tokens": 36560394.0,
      "step": 16365
    },
    {
      "entropy": 5.61123833656311,
      "epoch": 1.8385466389622058,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046699596434963953,
      "loss": 5.3686,
      "mean_token_accuracy": 0.18490458726882936,
      "num_tokens": 36570657.0,
      "step": 16370
    },
    {
      "entropy": 5.4530211925506595,
      "epoch": 1.8391082158701635,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004669750279484337,
      "loss": 5.2226,
      "mean_token_accuracy": 0.19429712146520614,
      "num_tokens": 36580810.0,
      "step": 16375
    },
    {
      "entropy": 5.473446083068848,
      "epoch": 1.839669792778121,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004669540854346685,
      "loss": 5.2702,
      "mean_token_accuracy": 0.18894062936306,
      "num_tokens": 36592981.0,
      "step": 16380
    },
    {
      "entropy": 5.488171720504761,
      "epoch": 1.8402313696860784,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046693313680901073,
      "loss": 5.2653,
      "mean_token_accuracy": 0.1925550326704979,
      "num_tokens": 36604130.0,
      "step": 16385
    },
    {
      "entropy": 5.470244550704956,
      "epoch": 1.840792946594036,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046691218207212744,
      "loss": 5.2851,
      "mean_token_accuracy": 0.1861992135643959,
      "num_tokens": 36614833.0,
      "step": 16390
    },
    {
      "entropy": 5.596527481079102,
      "epoch": 1.8413545235019937,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004668912212246859,
      "loss": 5.2951,
      "mean_token_accuracy": 0.19008202254772186,
      "num_tokens": 36625306.0,
      "step": 16395
    },
    {
      "entropy": 5.712037181854248,
      "epoch": 1.8419161004099511,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046687025426735335,
      "loss": 5.5507,
      "mean_token_accuracy": 0.17697899043560028,
      "num_tokens": 36637396.0,
      "step": 16400
    },
    {
      "entropy": 5.578679132461548,
      "epoch": 1.8424776773179086,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004668492812007976,
      "loss": 5.339,
      "mean_token_accuracy": 0.18426618576049805,
      "num_tokens": 36648175.0,
      "step": 16405
    },
    {
      "entropy": 5.5641011714935305,
      "epoch": 1.8430392542258662,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004668283020256863,
      "loss": 5.3825,
      "mean_token_accuracy": 0.18374981880187988,
      "num_tokens": 36659876.0,
      "step": 16410
    },
    {
      "entropy": 5.546352672576904,
      "epoch": 1.843600831133824,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046680731674268744,
      "loss": 5.388,
      "mean_token_accuracy": 0.1864971250295639,
      "num_tokens": 36671168.0,
      "step": 16415
    },
    {
      "entropy": 5.492926692962646,
      "epoch": 1.8441624080417813,
      "grad_norm": 1.25,
      "learning_rate": 0.0004667863253524692,
      "loss": 5.3088,
      "mean_token_accuracy": 0.19539319574832917,
      "num_tokens": 36680987.0,
      "step": 16420
    },
    {
      "entropy": 5.5537152767181395,
      "epoch": 1.8447239849497388,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046676532785569995,
      "loss": 5.376,
      "mean_token_accuracy": 0.185614974796772,
      "num_tokens": 36692396.0,
      "step": 16425
    },
    {
      "entropy": 5.494251823425293,
      "epoch": 1.8452855618576964,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004667443242530484,
      "loss": 5.2681,
      "mean_token_accuracy": 0.19782126992940902,
      "num_tokens": 36702696.0,
      "step": 16430
    },
    {
      "entropy": 5.4757472515106205,
      "epoch": 1.845847138765654,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046672331454518314,
      "loss": 5.2973,
      "mean_token_accuracy": 0.19498538821935654,
      "num_tokens": 36713582.0,
      "step": 16435
    },
    {
      "entropy": 5.494393634796142,
      "epoch": 1.8464087156736115,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004667022987327732,
      "loss": 5.2826,
      "mean_token_accuracy": 0.1949489161372185,
      "num_tokens": 36724324.0,
      "step": 16440
    },
    {
      "entropy": 5.568372821807861,
      "epoch": 1.846970292581569,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004666812768164877,
      "loss": 5.3701,
      "mean_token_accuracy": 0.19605543613433837,
      "num_tokens": 36736369.0,
      "step": 16445
    },
    {
      "entropy": 5.556389713287354,
      "epoch": 1.8475318694895266,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046666024879699606,
      "loss": 5.3992,
      "mean_token_accuracy": 0.19371566474437713,
      "num_tokens": 36748047.0,
      "step": 16450
    },
    {
      "entropy": 5.557503271102905,
      "epoch": 1.8480934463974843,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046663921467496766,
      "loss": 5.3201,
      "mean_token_accuracy": 0.19459699243307113,
      "num_tokens": 36758943.0,
      "step": 16455
    },
    {
      "entropy": 5.612880420684815,
      "epoch": 1.8486550233054415,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004666181744510724,
      "loss": 5.3868,
      "mean_token_accuracy": 0.1820188581943512,
      "num_tokens": 36769807.0,
      "step": 16460
    },
    {
      "entropy": 5.468385457992554,
      "epoch": 1.8492166002133992,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046659712812598017,
      "loss": 5.2466,
      "mean_token_accuracy": 0.19445807486772537,
      "num_tokens": 36780370.0,
      "step": 16465
    },
    {
      "entropy": 5.431660175323486,
      "epoch": 1.8497781771213568,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046657607570036103,
      "loss": 5.2679,
      "mean_token_accuracy": 0.19591842740774154,
      "num_tokens": 36789757.0,
      "step": 16470
    },
    {
      "entropy": 5.495207262039185,
      "epoch": 1.8503397540293143,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046655501717488533,
      "loss": 5.3225,
      "mean_token_accuracy": 0.1907737597823143,
      "num_tokens": 36800531.0,
      "step": 16475
    },
    {
      "entropy": 5.57843246459961,
      "epoch": 1.8509013309372717,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004665339525502236,
      "loss": 5.4184,
      "mean_token_accuracy": 0.18473077565431595,
      "num_tokens": 36811666.0,
      "step": 16480
    },
    {
      "entropy": 5.522574234008789,
      "epoch": 1.8514629078452294,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046651288182704653,
      "loss": 5.2653,
      "mean_token_accuracy": 0.18938798904418946,
      "num_tokens": 36822443.0,
      "step": 16485
    },
    {
      "entropy": 5.5522459030151365,
      "epoch": 1.852024484753187,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004664918050060251,
      "loss": 5.3293,
      "mean_token_accuracy": 0.1872788116335869,
      "num_tokens": 36835517.0,
      "step": 16490
    },
    {
      "entropy": 5.519955110549927,
      "epoch": 1.8525860616611445,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004664707220878302,
      "loss": 5.3371,
      "mean_token_accuracy": 0.19462635815143586,
      "num_tokens": 36845177.0,
      "step": 16495
    },
    {
      "entropy": 5.438331222534179,
      "epoch": 1.853147638569102,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004664496330731334,
      "loss": 5.2798,
      "mean_token_accuracy": 0.18872665315866471,
      "num_tokens": 36855821.0,
      "step": 16500
    },
    {
      "entropy": 5.556100273132325,
      "epoch": 1.8537092154770596,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004664285379626059,
      "loss": 5.389,
      "mean_token_accuracy": 0.1765715792775154,
      "num_tokens": 36866254.0,
      "step": 16505
    },
    {
      "entropy": 5.507565927505493,
      "epoch": 1.8542707923850172,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046640743675691954,
      "loss": 5.2211,
      "mean_token_accuracy": 0.19270853847265243,
      "num_tokens": 36876348.0,
      "step": 16510
    },
    {
      "entropy": 5.560690546035767,
      "epoch": 1.8548323692929747,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004663863294567462,
      "loss": 5.3788,
      "mean_token_accuracy": 0.1828359082341194,
      "num_tokens": 36888088.0,
      "step": 16515
    },
    {
      "entropy": 5.677534580230713,
      "epoch": 1.8553939462009321,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004663652160627579,
      "loss": 5.4796,
      "mean_token_accuracy": 0.18595814406871797,
      "num_tokens": 36900203.0,
      "step": 16520
    },
    {
      "entropy": 5.587461280822754,
      "epoch": 1.8559555231088898,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046634409657562683,
      "loss": 5.348,
      "mean_token_accuracy": 0.1860669359564781,
      "num_tokens": 36911321.0,
      "step": 16525
    },
    {
      "entropy": 5.57177586555481,
      "epoch": 1.8565171000168474,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004663229709960255,
      "loss": 5.4303,
      "mean_token_accuracy": 0.18599589318037033,
      "num_tokens": 36922562.0,
      "step": 16530
    },
    {
      "entropy": 5.52012825012207,
      "epoch": 1.8570786769248049,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046630183932462664,
      "loss": 5.2886,
      "mean_token_accuracy": 0.18841566741466523,
      "num_tokens": 36933923.0,
      "step": 16535
    },
    {
      "entropy": 5.484059810638428,
      "epoch": 1.8576402538327623,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004662807015621031,
      "loss": 5.3064,
      "mean_token_accuracy": 0.1898571565747261,
      "num_tokens": 36945203.0,
      "step": 16540
    },
    {
      "entropy": 5.523033905029297,
      "epoch": 1.85820183074072,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046625955770912764,
      "loss": 5.2992,
      "mean_token_accuracy": 0.18876135647296904,
      "num_tokens": 36956094.0,
      "step": 16545
    },
    {
      "entropy": 5.586569976806641,
      "epoch": 1.8587634076486776,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046623840776637384,
      "loss": 5.3545,
      "mean_token_accuracy": 0.19304639995098113,
      "num_tokens": 36967375.0,
      "step": 16550
    },
    {
      "entropy": 5.418436574935913,
      "epoch": 1.859324984556635,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004662172517345149,
      "loss": 5.247,
      "mean_token_accuracy": 0.19597518146038057,
      "num_tokens": 36977674.0,
      "step": 16555
    },
    {
      "entropy": 5.527724647521973,
      "epoch": 1.8598865614645925,
      "grad_norm": 1.5,
      "learning_rate": 0.0004661960896142245,
      "loss": 5.4098,
      "mean_token_accuracy": 0.19291428476572037,
      "num_tokens": 36990241.0,
      "step": 16560
    },
    {
      "entropy": 5.545258331298828,
      "epoch": 1.8604481383725502,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046617492140617654,
      "loss": 5.3528,
      "mean_token_accuracy": 0.1865066796541214,
      "num_tokens": 37002322.0,
      "step": 16565
    },
    {
      "entropy": 5.5665429592132565,
      "epoch": 1.8610097152805076,
      "grad_norm": 1.375,
      "learning_rate": 0.0004661537471110448,
      "loss": 5.3697,
      "mean_token_accuracy": 0.1901166707277298,
      "num_tokens": 37014524.0,
      "step": 16570
    },
    {
      "entropy": 5.5150861740112305,
      "epoch": 1.861571292188465,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004661325667295037,
      "loss": 5.2295,
      "mean_token_accuracy": 0.19875368624925613,
      "num_tokens": 37025057.0,
      "step": 16575
    },
    {
      "entropy": 5.495872449874878,
      "epoch": 1.8621328690964227,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004661113802622275,
      "loss": 5.382,
      "mean_token_accuracy": 0.1921732857823372,
      "num_tokens": 37035311.0,
      "step": 16580
    },
    {
      "entropy": 5.515810537338257,
      "epoch": 1.8626944460043804,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004660901877098908,
      "loss": 5.3676,
      "mean_token_accuracy": 0.19142916351556777,
      "num_tokens": 37046275.0,
      "step": 16585
    },
    {
      "entropy": 5.527347612380981,
      "epoch": 1.8632560229123378,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004660689890731685,
      "loss": 5.2279,
      "mean_token_accuracy": 0.19131554663181305,
      "num_tokens": 37057658.0,
      "step": 16590
    },
    {
      "entropy": 5.513562154769898,
      "epoch": 1.8638175998202953,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00046604778435273546,
      "loss": 5.2489,
      "mean_token_accuracy": 0.19852979779243468,
      "num_tokens": 37067946.0,
      "step": 16595
    },
    {
      "entropy": 5.453700017929077,
      "epoch": 1.864379176728253,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046602657354926686,
      "loss": 5.2634,
      "mean_token_accuracy": 0.19169158935546876,
      "num_tokens": 37078881.0,
      "step": 16600
    },
    {
      "entropy": 5.513870763778686,
      "epoch": 1.8649407536362106,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000466005356663438,
      "loss": 5.3893,
      "mean_token_accuracy": 0.18352641761302949,
      "num_tokens": 37090474.0,
      "step": 16605
    },
    {
      "entropy": 5.539599275588989,
      "epoch": 1.865502330544168,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046598413369592463,
      "loss": 5.4367,
      "mean_token_accuracy": 0.19055111408233644,
      "num_tokens": 37101527.0,
      "step": 16610
    },
    {
      "entropy": 5.585215616226196,
      "epoch": 1.8660639074521255,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004659629046474023,
      "loss": 5.355,
      "mean_token_accuracy": 0.18252760022878647,
      "num_tokens": 37112615.0,
      "step": 16615
    },
    {
      "entropy": 5.5295312881469725,
      "epoch": 1.8666254843600831,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000465941669518547,
      "loss": 5.2566,
      "mean_token_accuracy": 0.19171176999807357,
      "num_tokens": 37122922.0,
      "step": 16620
    },
    {
      "entropy": 5.498215818405152,
      "epoch": 1.8671870612680408,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004659204283100349,
      "loss": 5.2786,
      "mean_token_accuracy": 0.19974020570516587,
      "num_tokens": 37133215.0,
      "step": 16625
    },
    {
      "entropy": 5.424301290512085,
      "epoch": 1.8677486381759982,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046589918102254237,
      "loss": 5.1965,
      "mean_token_accuracy": 0.19507750272750854,
      "num_tokens": 37144292.0,
      "step": 16630
    },
    {
      "entropy": 5.477900838851928,
      "epoch": 1.8683102150839557,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004658779276567458,
      "loss": 5.3558,
      "mean_token_accuracy": 0.18585943430662155,
      "num_tokens": 37155623.0,
      "step": 16635
    },
    {
      "entropy": 5.630840015411377,
      "epoch": 1.8688717919919133,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046585666821332203,
      "loss": 5.3456,
      "mean_token_accuracy": 0.190969055891037,
      "num_tokens": 37165406.0,
      "step": 16640
    },
    {
      "entropy": 5.563094186782837,
      "epoch": 1.869433368899871,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046583540269294794,
      "loss": 5.3656,
      "mean_token_accuracy": 0.19431325793266296,
      "num_tokens": 37175522.0,
      "step": 16645
    },
    {
      "entropy": 5.492715883255005,
      "epoch": 1.8699949458078284,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004658141310963006,
      "loss": 5.3912,
      "mean_token_accuracy": 0.19230775833129882,
      "num_tokens": 37186745.0,
      "step": 16650
    },
    {
      "entropy": 5.479142761230468,
      "epoch": 1.8705565227157859,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004657928534240573,
      "loss": 5.2403,
      "mean_token_accuracy": 0.19582124352455138,
      "num_tokens": 37197357.0,
      "step": 16655
    },
    {
      "entropy": 5.554664564132691,
      "epoch": 1.8711180996237435,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004657715696768955,
      "loss": 5.3881,
      "mean_token_accuracy": 0.18799951076507568,
      "num_tokens": 37209839.0,
      "step": 16660
    },
    {
      "entropy": 5.596767520904541,
      "epoch": 1.871679676531701,
      "grad_norm": 1.15625,
      "learning_rate": 0.000465750279855493,
      "loss": 5.4143,
      "mean_token_accuracy": 0.1821955546736717,
      "num_tokens": 37221314.0,
      "step": 16665
    },
    {
      "entropy": 5.569704675674439,
      "epoch": 1.8722412534396584,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004657289839605276,
      "loss": 5.2706,
      "mean_token_accuracy": 0.1906728371977806,
      "num_tokens": 37233372.0,
      "step": 16670
    },
    {
      "entropy": 5.593374586105346,
      "epoch": 1.872802830347616,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004657076819926774,
      "loss": 5.3701,
      "mean_token_accuracy": 0.18732914477586746,
      "num_tokens": 37244758.0,
      "step": 16675
    },
    {
      "entropy": 5.568263387680053,
      "epoch": 1.8733644072555737,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004656863739526206,
      "loss": 5.3281,
      "mean_token_accuracy": 0.18720599859952927,
      "num_tokens": 37255848.0,
      "step": 16680
    },
    {
      "entropy": 5.488857936859131,
      "epoch": 1.8739259841635312,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046566505984103575,
      "loss": 5.2712,
      "mean_token_accuracy": 0.1951414555311203,
      "num_tokens": 37266010.0,
      "step": 16685
    },
    {
      "entropy": 5.4241259574890135,
      "epoch": 1.8744875610714886,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004656437396586014,
      "loss": 5.2931,
      "mean_token_accuracy": 0.18768802732229234,
      "num_tokens": 37278445.0,
      "step": 16690
    },
    {
      "entropy": 5.642972850799561,
      "epoch": 1.8750491379794463,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004656224134059964,
      "loss": 5.4392,
      "mean_token_accuracy": 0.18230196088552475,
      "num_tokens": 37291181.0,
      "step": 16695
    },
    {
      "entropy": 5.6833336353302,
      "epoch": 1.875610714887404,
      "grad_norm": 1.25,
      "learning_rate": 0.00046560108108389985,
      "loss": 5.52,
      "mean_token_accuracy": 0.18521885722875595,
      "num_tokens": 37302333.0,
      "step": 16700
    },
    {
      "entropy": 5.516350030899048,
      "epoch": 1.8761722917953614,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000465579742692991,
      "loss": 5.2709,
      "mean_token_accuracy": 0.1961539715528488,
      "num_tokens": 37312754.0,
      "step": 16705
    },
    {
      "entropy": 5.5581494808197025,
      "epoch": 1.8767338687033188,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004655583982339492,
      "loss": 5.3492,
      "mean_token_accuracy": 0.19225821644067764,
      "num_tokens": 37323335.0,
      "step": 16710
    },
    {
      "entropy": 5.565476655960083,
      "epoch": 1.8772954456112765,
      "grad_norm": 1.25,
      "learning_rate": 0.0004655370477074541,
      "loss": 5.3435,
      "mean_token_accuracy": 0.1916085883975029,
      "num_tokens": 37334332.0,
      "step": 16715
    },
    {
      "entropy": 5.5233629703521725,
      "epoch": 1.8778570225192341,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004655156911141855,
      "loss": 5.3511,
      "mean_token_accuracy": 0.18915074169635773,
      "num_tokens": 37344770.0,
      "step": 16720
    },
    {
      "entropy": 5.549921321868896,
      "epoch": 1.8784185994271916,
      "grad_norm": 1.375,
      "learning_rate": 0.0004654943284548234,
      "loss": 5.2933,
      "mean_token_accuracy": 0.1868213325738907,
      "num_tokens": 37356009.0,
      "step": 16725
    },
    {
      "entropy": 5.4536461353302,
      "epoch": 1.878980176335149,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046547295973004807,
      "loss": 5.2729,
      "mean_token_accuracy": 0.19684419333934783,
      "num_tokens": 37367823.0,
      "step": 16730
    },
    {
      "entropy": 5.45973949432373,
      "epoch": 1.8795417532431067,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046545158494053966,
      "loss": 5.3001,
      "mean_token_accuracy": 0.18995605856180192,
      "num_tokens": 37378156.0,
      "step": 16735
    },
    {
      "entropy": 5.555756092071533,
      "epoch": 1.8801033301510643,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046543020408697905,
      "loss": 5.339,
      "mean_token_accuracy": 0.1906617671251297,
      "num_tokens": 37388378.0,
      "step": 16740
    },
    {
      "entropy": 5.4716565132141115,
      "epoch": 1.8806649070590218,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046540881717004685,
      "loss": 5.2365,
      "mean_token_accuracy": 0.18968347012996672,
      "num_tokens": 37398457.0,
      "step": 16745
    },
    {
      "entropy": 5.553845930099487,
      "epoch": 1.8812264839669792,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004653874241904241,
      "loss": 5.4273,
      "mean_token_accuracy": 0.1854670152068138,
      "num_tokens": 37409341.0,
      "step": 16750
    },
    {
      "entropy": 5.5429408073425295,
      "epoch": 1.8817880608749369,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004653660251487919,
      "loss": 5.2391,
      "mean_token_accuracy": 0.19419020563364028,
      "num_tokens": 37420081.0,
      "step": 16755
    },
    {
      "entropy": 5.446212291717529,
      "epoch": 1.8823496377828943,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046534462004583154,
      "loss": 5.2324,
      "mean_token_accuracy": 0.20230178385972977,
      "num_tokens": 37430707.0,
      "step": 16760
    },
    {
      "entropy": 5.4975244998931885,
      "epoch": 1.8829112146908518,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004653232088822248,
      "loss": 5.3504,
      "mean_token_accuracy": 0.18768247067928315,
      "num_tokens": 37441250.0,
      "step": 16765
    },
    {
      "entropy": 5.561507081985473,
      "epoch": 1.8834727915988094,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004653017916586532,
      "loss": 5.4027,
      "mean_token_accuracy": 0.18384319096803664,
      "num_tokens": 37452909.0,
      "step": 16770
    },
    {
      "entropy": 5.5425133228302,
      "epoch": 1.884034368506767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004652803683757987,
      "loss": 5.3365,
      "mean_token_accuracy": 0.1840794563293457,
      "num_tokens": 37463833.0,
      "step": 16775
    },
    {
      "entropy": 5.441729974746704,
      "epoch": 1.8845959454147245,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004652589390343436,
      "loss": 5.2705,
      "mean_token_accuracy": 0.19357890188694,
      "num_tokens": 37473967.0,
      "step": 16780
    },
    {
      "entropy": 5.564308452606201,
      "epoch": 1.885157522322682,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046523750363497,
      "loss": 5.3053,
      "mean_token_accuracy": 0.19836509376764297,
      "num_tokens": 37484748.0,
      "step": 16785
    },
    {
      "entropy": 5.549997997283936,
      "epoch": 1.8857190992306396,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004652160621783605,
      "loss": 5.3301,
      "mean_token_accuracy": 0.1822197899222374,
      "num_tokens": 37495757.0,
      "step": 16790
    },
    {
      "entropy": 5.497464179992676,
      "epoch": 1.8862806761385973,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046519461466519785,
      "loss": 5.3374,
      "mean_token_accuracy": 0.1866649031639099,
      "num_tokens": 37506612.0,
      "step": 16795
    },
    {
      "entropy": 5.467371129989624,
      "epoch": 1.8868422530465547,
      "grad_norm": 1.5,
      "learning_rate": 0.0004651731610961649,
      "loss": 5.2538,
      "mean_token_accuracy": 0.19866592735052108,
      "num_tokens": 37518639.0,
      "step": 16800
    },
    {
      "entropy": 5.52176194190979,
      "epoch": 1.8874038299545122,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004651517014719446,
      "loss": 5.3212,
      "mean_token_accuracy": 0.18694213777780533,
      "num_tokens": 37530020.0,
      "step": 16805
    },
    {
      "entropy": 5.483829069137573,
      "epoch": 1.8879654068624698,
      "grad_norm": 1.375,
      "learning_rate": 0.0004651302357932206,
      "loss": 5.2563,
      "mean_token_accuracy": 0.1908920630812645,
      "num_tokens": 37540737.0,
      "step": 16810
    },
    {
      "entropy": 5.427550506591797,
      "epoch": 1.8885269837704275,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000465108764060676,
      "loss": 5.275,
      "mean_token_accuracy": 0.18921370804309845,
      "num_tokens": 37551091.0,
      "step": 16815
    },
    {
      "entropy": 5.542868661880493,
      "epoch": 1.889088560678385,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046508728627499464,
      "loss": 5.3951,
      "mean_token_accuracy": 0.18349259197711945,
      "num_tokens": 37561904.0,
      "step": 16820
    },
    {
      "entropy": 5.62216272354126,
      "epoch": 1.8896501375863424,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004650658024368604,
      "loss": 5.4204,
      "mean_token_accuracy": 0.1816960707306862,
      "num_tokens": 37573532.0,
      "step": 16825
    },
    {
      "entropy": 5.659744358062744,
      "epoch": 1.8902117144943,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004650443125469573,
      "loss": 5.4083,
      "mean_token_accuracy": 0.19345381557941438,
      "num_tokens": 37584886.0,
      "step": 16830
    },
    {
      "entropy": 5.502018308639526,
      "epoch": 1.8907732914022577,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004650228166059695,
      "loss": 5.3367,
      "mean_token_accuracy": 0.18997133523225784,
      "num_tokens": 37595190.0,
      "step": 16835
    },
    {
      "entropy": 5.446289587020874,
      "epoch": 1.8913348683102151,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004650013146145816,
      "loss": 5.2862,
      "mean_token_accuracy": 0.19220385253429412,
      "num_tokens": 37605776.0,
      "step": 16840
    },
    {
      "entropy": 5.606544780731201,
      "epoch": 1.8918964452181726,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004649798065734781,
      "loss": 5.3897,
      "mean_token_accuracy": 0.18499674052000045,
      "num_tokens": 37616882.0,
      "step": 16845
    },
    {
      "entropy": 5.664537143707276,
      "epoch": 1.8924580221261302,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046495829248334393,
      "loss": 5.4515,
      "mean_token_accuracy": 0.18400414884090424,
      "num_tokens": 37628528.0,
      "step": 16850
    },
    {
      "entropy": 5.551404237747192,
      "epoch": 1.8930195990340877,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004649367723448641,
      "loss": 5.3587,
      "mean_token_accuracy": 0.18622815757989883,
      "num_tokens": 37640502.0,
      "step": 16855
    },
    {
      "entropy": 5.4735455989837645,
      "epoch": 1.8935811759420451,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046491524615872366,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1845741242170334,
      "num_tokens": 37652845.0,
      "step": 16860
    },
    {
      "entropy": 5.5555400371551515,
      "epoch": 1.8941427528500028,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046489371392560816,
      "loss": 5.451,
      "mean_token_accuracy": 0.1814900740981102,
      "num_tokens": 37665418.0,
      "step": 16865
    },
    {
      "entropy": 5.5555469512939455,
      "epoch": 1.8947043297579604,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046487217564620317,
      "loss": 5.3789,
      "mean_token_accuracy": 0.1876147910952568,
      "num_tokens": 37676429.0,
      "step": 16870
    },
    {
      "entropy": 5.571235418319702,
      "epoch": 1.8952659066659179,
      "grad_norm": 1.375,
      "learning_rate": 0.0004648506313211945,
      "loss": 5.3486,
      "mean_token_accuracy": 0.19341038316488265,
      "num_tokens": 37686779.0,
      "step": 16875
    },
    {
      "entropy": 5.491058588027954,
      "epoch": 1.8958274835738753,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046482908095126803,
      "loss": 5.3897,
      "mean_token_accuracy": 0.18769828379154205,
      "num_tokens": 37697665.0,
      "step": 16880
    },
    {
      "entropy": 5.5995521068573,
      "epoch": 1.896389060481833,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046480752453711,
      "loss": 5.3992,
      "mean_token_accuracy": 0.18566225469112396,
      "num_tokens": 37708987.0,
      "step": 16885
    },
    {
      "entropy": 5.495259618759155,
      "epoch": 1.8969506373897906,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004647859620794068,
      "loss": 5.2325,
      "mean_token_accuracy": 0.1924973323941231,
      "num_tokens": 37720335.0,
      "step": 16890
    },
    {
      "entropy": 5.579281902313232,
      "epoch": 1.897512214297748,
      "grad_norm": 1.375,
      "learning_rate": 0.00046476439357884497,
      "loss": 5.4145,
      "mean_token_accuracy": 0.18679735511541368,
      "num_tokens": 37731586.0,
      "step": 16895
    },
    {
      "entropy": 5.562653350830078,
      "epoch": 1.8980737912057055,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046474281903611126,
      "loss": 5.3104,
      "mean_token_accuracy": 0.19232197701931,
      "num_tokens": 37742907.0,
      "step": 16900
    },
    {
      "entropy": 5.62323112487793,
      "epoch": 1.8986353681136632,
      "grad_norm": 1.25,
      "learning_rate": 0.00046472123845189253,
      "loss": 5.3992,
      "mean_token_accuracy": 0.18631621599197387,
      "num_tokens": 37753787.0,
      "step": 16905
    },
    {
      "entropy": 5.490570402145385,
      "epoch": 1.8991969450216208,
      "grad_norm": 1.21875,
      "learning_rate": 0.000464699651826876,
      "loss": 5.3081,
      "mean_token_accuracy": 0.1899639844894409,
      "num_tokens": 37766047.0,
      "step": 16910
    },
    {
      "entropy": 5.505640983581543,
      "epoch": 1.8997585219295783,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046467805916174907,
      "loss": 5.2912,
      "mean_token_accuracy": 0.1874990016222,
      "num_tokens": 37777651.0,
      "step": 16915
    },
    {
      "entropy": 5.486234426498413,
      "epoch": 1.9003200988375357,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046465646045719915,
      "loss": 5.3304,
      "mean_token_accuracy": 0.1921023830771446,
      "num_tokens": 37788527.0,
      "step": 16920
    },
    {
      "entropy": 5.444620609283447,
      "epoch": 1.9008816757454934,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004646348557139139,
      "loss": 5.2755,
      "mean_token_accuracy": 0.18400513976812363,
      "num_tokens": 37800727.0,
      "step": 16925
    },
    {
      "entropy": 5.529443979263306,
      "epoch": 1.901443252653451,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046461324493258133,
      "loss": 5.2671,
      "mean_token_accuracy": 0.1931947410106659,
      "num_tokens": 37812332.0,
      "step": 16930
    },
    {
      "entropy": 5.5595433712005615,
      "epoch": 1.9020048295614085,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046459162811388955,
      "loss": 5.3685,
      "mean_token_accuracy": 0.18835092037916185,
      "num_tokens": 37823665.0,
      "step": 16935
    },
    {
      "entropy": 5.451412010192871,
      "epoch": 1.902566406469366,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004645700052585267,
      "loss": 5.1889,
      "mean_token_accuracy": 0.20055983066558838,
      "num_tokens": 37833793.0,
      "step": 16940
    },
    {
      "entropy": 5.42030701637268,
      "epoch": 1.9031279833773236,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004645483763671815,
      "loss": 5.222,
      "mean_token_accuracy": 0.1994650512933731,
      "num_tokens": 37844104.0,
      "step": 16945
    },
    {
      "entropy": 5.556173849105835,
      "epoch": 1.903689560285281,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046452674144054236,
      "loss": 5.3015,
      "mean_token_accuracy": 0.19548871517181396,
      "num_tokens": 37853920.0,
      "step": 16950
    },
    {
      "entropy": 5.505234622955323,
      "epoch": 1.9042511371932385,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046450510047929826,
      "loss": 5.3638,
      "mean_token_accuracy": 0.18661977350711823,
      "num_tokens": 37865749.0,
      "step": 16955
    },
    {
      "entropy": 5.4916908740997314,
      "epoch": 1.9048127141011961,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046448345348413833,
      "loss": 5.3213,
      "mean_token_accuracy": 0.18939271867275237,
      "num_tokens": 37876259.0,
      "step": 16960
    },
    {
      "entropy": 5.591541385650634,
      "epoch": 1.9053742910091538,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004644618004557517,
      "loss": 5.3973,
      "mean_token_accuracy": 0.18937608003616332,
      "num_tokens": 37887186.0,
      "step": 16965
    },
    {
      "entropy": 5.450802040100098,
      "epoch": 1.9059358679171112,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004644401413948279,
      "loss": 5.2058,
      "mean_token_accuracy": 0.19802374243736268,
      "num_tokens": 37898210.0,
      "step": 16970
    },
    {
      "entropy": 5.5113931655883786,
      "epoch": 1.9064974448250687,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004644184763020565,
      "loss": 5.3294,
      "mean_token_accuracy": 0.1850786715745926,
      "num_tokens": 37909332.0,
      "step": 16975
    },
    {
      "entropy": 5.499251031875611,
      "epoch": 1.9070590217330263,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004643968051781273,
      "loss": 5.2756,
      "mean_token_accuracy": 0.1952410012483597,
      "num_tokens": 37919796.0,
      "step": 16980
    },
    {
      "entropy": 5.546873712539673,
      "epoch": 1.907620598640984,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004643751280237305,
      "loss": 5.4289,
      "mean_token_accuracy": 0.18502222895622253,
      "num_tokens": 37931135.0,
      "step": 16985
    },
    {
      "entropy": 5.632438373565674,
      "epoch": 1.9081821755489414,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004643534448395561,
      "loss": 5.4213,
      "mean_token_accuracy": 0.18655699491500854,
      "num_tokens": 37942302.0,
      "step": 16990
    },
    {
      "entropy": 5.503334045410156,
      "epoch": 1.9087437524568989,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004643317556262946,
      "loss": 5.3274,
      "mean_token_accuracy": 0.18913712799549104,
      "num_tokens": 37954188.0,
      "step": 16995
    },
    {
      "entropy": 5.523032188415527,
      "epoch": 1.9093053293648565,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046431006038463653,
      "loss": 5.3595,
      "mean_token_accuracy": 0.18022434711456298,
      "num_tokens": 37964778.0,
      "step": 17000
    },
    {
      "entropy": 5.638865041732788,
      "epoch": 1.9098669062728142,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004642883591152727,
      "loss": 5.3805,
      "mean_token_accuracy": 0.1816907823085785,
      "num_tokens": 37976384.0,
      "step": 17005
    },
    {
      "entropy": 5.5630089282989506,
      "epoch": 1.9104284831807716,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004642666518188942,
      "loss": 5.3749,
      "mean_token_accuracy": 0.18493264317512512,
      "num_tokens": 37986975.0,
      "step": 17010
    },
    {
      "entropy": 5.65526819229126,
      "epoch": 1.910990060088729,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046424493849619205,
      "loss": 5.486,
      "mean_token_accuracy": 0.18746909946203233,
      "num_tokens": 37999255.0,
      "step": 17015
    },
    {
      "entropy": 5.624090242385864,
      "epoch": 1.9115516369966867,
      "grad_norm": 2.125,
      "learning_rate": 0.00046422321914785757,
      "loss": 5.4296,
      "mean_token_accuracy": 0.18602934181690217,
      "num_tokens": 38010436.0,
      "step": 17020
    },
    {
      "entropy": 5.434043741226196,
      "epoch": 1.9121132139046444,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046420149377458255,
      "loss": 5.2446,
      "mean_token_accuracy": 0.19463708102703095,
      "num_tokens": 38021537.0,
      "step": 17025
    },
    {
      "entropy": 5.514421558380127,
      "epoch": 1.9126747908126018,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046417976237705846,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18595751821994783,
      "num_tokens": 38033396.0,
      "step": 17030
    },
    {
      "entropy": 5.533701801300049,
      "epoch": 1.9132363677205593,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046415802495597744,
      "loss": 5.2384,
      "mean_token_accuracy": 0.20455720126628876,
      "num_tokens": 38044917.0,
      "step": 17035
    },
    {
      "entropy": 5.525361251831055,
      "epoch": 1.913797944628517,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00046413628151203154,
      "loss": 5.4001,
      "mean_token_accuracy": 0.19001386910676957,
      "num_tokens": 38057002.0,
      "step": 17040
    },
    {
      "entropy": 5.448178672790528,
      "epoch": 1.9143595215364744,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046411453204591303,
      "loss": 5.1667,
      "mean_token_accuracy": 0.20019909888505935,
      "num_tokens": 38068023.0,
      "step": 17045
    },
    {
      "entropy": 5.618009853363037,
      "epoch": 1.9149210984444318,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046409277655831455,
      "loss": 5.4787,
      "mean_token_accuracy": 0.17797086536884307,
      "num_tokens": 38080137.0,
      "step": 17050
    },
    {
      "entropy": 5.601627683639526,
      "epoch": 1.9154826753523895,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004640710150499287,
      "loss": 5.2752,
      "mean_token_accuracy": 0.19400590062141418,
      "num_tokens": 38090078.0,
      "step": 17055
    },
    {
      "entropy": 5.5804304599761965,
      "epoch": 1.9160442522603471,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046404924752144834,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18672227412462233,
      "num_tokens": 38101410.0,
      "step": 17060
    },
    {
      "entropy": 5.512361478805542,
      "epoch": 1.9166058291683046,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046402747397356666,
      "loss": 5.2792,
      "mean_token_accuracy": 0.19851553738117217,
      "num_tokens": 38111610.0,
      "step": 17065
    },
    {
      "entropy": 5.558873319625855,
      "epoch": 1.917167406076262,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004640056944069768,
      "loss": 5.3149,
      "mean_token_accuracy": 0.19130940288305281,
      "num_tokens": 38122947.0,
      "step": 17070
    },
    {
      "entropy": 5.499215221405029,
      "epoch": 1.9177289829842197,
      "grad_norm": 1.25,
      "learning_rate": 0.0004639839088223725,
      "loss": 5.2677,
      "mean_token_accuracy": 0.20023803561925888,
      "num_tokens": 38133083.0,
      "step": 17075
    },
    {
      "entropy": 5.53978419303894,
      "epoch": 1.9182905598921773,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046396211722044716,
      "loss": 5.315,
      "mean_token_accuracy": 0.1979520857334137,
      "num_tokens": 38142941.0,
      "step": 17080
    },
    {
      "entropy": 5.56115345954895,
      "epoch": 1.9188521368001348,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046394031960189473,
      "loss": 5.3485,
      "mean_token_accuracy": 0.18756425976753235,
      "num_tokens": 38154269.0,
      "step": 17085
    },
    {
      "entropy": 5.49371976852417,
      "epoch": 1.9194137137080922,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004639185159674093,
      "loss": 5.266,
      "mean_token_accuracy": 0.19421560764312745,
      "num_tokens": 38164153.0,
      "step": 17090
    },
    {
      "entropy": 5.554946804046631,
      "epoch": 1.9199752906160499,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046389670631768494,
      "loss": 5.3929,
      "mean_token_accuracy": 0.1864183157682419,
      "num_tokens": 38174704.0,
      "step": 17095
    },
    {
      "entropy": 5.675718641281128,
      "epoch": 1.9205368675240075,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004638748906534162,
      "loss": 5.423,
      "mean_token_accuracy": 0.18479039371013642,
      "num_tokens": 38186292.0,
      "step": 17100
    },
    {
      "entropy": 5.5812060832977295,
      "epoch": 1.921098444431965,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004638530689752977,
      "loss": 5.3528,
      "mean_token_accuracy": 0.1936494678258896,
      "num_tokens": 38197419.0,
      "step": 17105
    },
    {
      "entropy": 5.524025583267212,
      "epoch": 1.9216600213399224,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046383124128402433,
      "loss": 5.359,
      "mean_token_accuracy": 0.18838028460741044,
      "num_tokens": 38208276.0,
      "step": 17110
    },
    {
      "entropy": 5.435537338256836,
      "epoch": 1.92222159824788,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046380940758029095,
      "loss": 5.2467,
      "mean_token_accuracy": 0.1975184515118599,
      "num_tokens": 38220276.0,
      "step": 17115
    },
    {
      "entropy": 5.510719680786133,
      "epoch": 1.9227831751558377,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046378756786479274,
      "loss": 5.3592,
      "mean_token_accuracy": 0.1928332716226578,
      "num_tokens": 38230760.0,
      "step": 17120
    },
    {
      "entropy": 5.502442646026611,
      "epoch": 1.9233447520637952,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004637657221382252,
      "loss": 5.2511,
      "mean_token_accuracy": 0.19449742287397384,
      "num_tokens": 38241769.0,
      "step": 17125
    },
    {
      "entropy": 5.562748861312866,
      "epoch": 1.9239063289717526,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046374387040128384,
      "loss": 5.3496,
      "mean_token_accuracy": 0.18859248012304305,
      "num_tokens": 38252800.0,
      "step": 17130
    },
    {
      "entropy": 5.461896514892578,
      "epoch": 1.9244679058797103,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046372201265466444,
      "loss": 5.2612,
      "mean_token_accuracy": 0.1905917689204216,
      "num_tokens": 38264319.0,
      "step": 17135
    },
    {
      "entropy": 5.46633825302124,
      "epoch": 1.9250294827876677,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000463700148899063,
      "loss": 5.3072,
      "mean_token_accuracy": 0.1901361256837845,
      "num_tokens": 38275118.0,
      "step": 17140
    },
    {
      "entropy": 5.656360721588134,
      "epoch": 1.9255910596956252,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004636782791351756,
      "loss": 5.4582,
      "mean_token_accuracy": 0.18556168526411057,
      "num_tokens": 38286264.0,
      "step": 17145
    },
    {
      "entropy": 5.5240942478179935,
      "epoch": 1.9261526366035828,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004636564033636987,
      "loss": 5.2513,
      "mean_token_accuracy": 0.19560887068510055,
      "num_tokens": 38297603.0,
      "step": 17150
    },
    {
      "entropy": 5.538367128372192,
      "epoch": 1.9267142135115405,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046363452158532857,
      "loss": 5.3964,
      "mean_token_accuracy": 0.1822678416967392,
      "num_tokens": 38308760.0,
      "step": 17155
    },
    {
      "entropy": 5.582892894744873,
      "epoch": 1.927275790419498,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004636126338007622,
      "loss": 5.3517,
      "mean_token_accuracy": 0.18947700262069703,
      "num_tokens": 38320807.0,
      "step": 17160
    },
    {
      "entropy": 5.46394453048706,
      "epoch": 1.9278373673274554,
      "grad_norm": 1.375,
      "learning_rate": 0.0004635907400106965,
      "loss": 5.3272,
      "mean_token_accuracy": 0.18540936410427095,
      "num_tokens": 38331087.0,
      "step": 17165
    },
    {
      "entropy": 5.478472423553467,
      "epoch": 1.928398944235413,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004635688402158285,
      "loss": 5.2903,
      "mean_token_accuracy": 0.19315948486328124,
      "num_tokens": 38341659.0,
      "step": 17170
    },
    {
      "entropy": 5.470145225524902,
      "epoch": 1.9289605211433707,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046354693441685537,
      "loss": 5.254,
      "mean_token_accuracy": 0.1987090140581131,
      "num_tokens": 38351371.0,
      "step": 17175
    },
    {
      "entropy": 5.468095541000366,
      "epoch": 1.9295220980513281,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046352502261447475,
      "loss": 5.262,
      "mean_token_accuracy": 0.1989993393421173,
      "num_tokens": 38362677.0,
      "step": 17180
    },
    {
      "entropy": 5.46759262084961,
      "epoch": 1.9300836749592856,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004635031048093843,
      "loss": 5.3095,
      "mean_token_accuracy": 0.19176963865756988,
      "num_tokens": 38372882.0,
      "step": 17185
    },
    {
      "entropy": 5.596657609939575,
      "epoch": 1.9306452518672432,
      "grad_norm": 1.25,
      "learning_rate": 0.0004634811810022819,
      "loss": 5.4029,
      "mean_token_accuracy": 0.18605346381664276,
      "num_tokens": 38383401.0,
      "step": 17190
    },
    {
      "entropy": 5.599388122558594,
      "epoch": 1.931206828775201,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046345925119386556,
      "loss": 5.432,
      "mean_token_accuracy": 0.1843906819820404,
      "num_tokens": 38394733.0,
      "step": 17195
    },
    {
      "entropy": 5.609647226333618,
      "epoch": 1.9317684056831583,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004634373153848336,
      "loss": 5.3993,
      "mean_token_accuracy": 0.19139508306980133,
      "num_tokens": 38406482.0,
      "step": 17200
    },
    {
      "entropy": 5.578036069869995,
      "epoch": 1.9323299825911158,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004634153735758845,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19059739410877227,
      "num_tokens": 38417366.0,
      "step": 17205
    },
    {
      "entropy": 5.554816341400146,
      "epoch": 1.9328915594990734,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004633934257677167,
      "loss": 5.37,
      "mean_token_accuracy": 0.19673199206590652,
      "num_tokens": 38428440.0,
      "step": 17210
    },
    {
      "entropy": 5.5706367015838625,
      "epoch": 1.933453136407031,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004633714719610292,
      "loss": 5.3233,
      "mean_token_accuracy": 0.1908646509051323,
      "num_tokens": 38438471.0,
      "step": 17215
    },
    {
      "entropy": 5.588924694061279,
      "epoch": 1.9340147133149885,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046334951215652097,
      "loss": 5.3018,
      "mean_token_accuracy": 0.19423958659172058,
      "num_tokens": 38448990.0,
      "step": 17220
    },
    {
      "entropy": 5.582559108734131,
      "epoch": 1.934576290222946,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046332754635489126,
      "loss": 5.3657,
      "mean_token_accuracy": 0.18794859498739241,
      "num_tokens": 38460990.0,
      "step": 17225
    },
    {
      "entropy": 5.569936895370484,
      "epoch": 1.9351378671309036,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004633055745568394,
      "loss": 5.4139,
      "mean_token_accuracy": 0.1809059590101242,
      "num_tokens": 38471672.0,
      "step": 17230
    },
    {
      "entropy": 5.4709044933319095,
      "epoch": 1.935699444038861,
      "grad_norm": 1.375,
      "learning_rate": 0.00046328359676306506,
      "loss": 5.2884,
      "mean_token_accuracy": 0.1941262200474739,
      "num_tokens": 38483248.0,
      "step": 17235
    },
    {
      "entropy": 5.566699647903443,
      "epoch": 1.9362610209468185,
      "grad_norm": 1.390625,
      "learning_rate": 0.000463261612974268,
      "loss": 5.3048,
      "mean_token_accuracy": 0.19185658544301987,
      "num_tokens": 38494795.0,
      "step": 17240
    },
    {
      "entropy": 5.485395050048828,
      "epoch": 1.9368225978547762,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004632396231911481,
      "loss": 5.3053,
      "mean_token_accuracy": 0.19147204756736755,
      "num_tokens": 38505276.0,
      "step": 17245
    },
    {
      "entropy": 5.625464916229248,
      "epoch": 1.9373841747627338,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004632176274144056,
      "loss": 5.3527,
      "mean_token_accuracy": 0.18187369257211686,
      "num_tokens": 38515536.0,
      "step": 17250
    },
    {
      "entropy": 5.529721355438232,
      "epoch": 1.9379457516706913,
      "grad_norm": 1.640625,
      "learning_rate": 0.00046319562564474083,
      "loss": 5.3186,
      "mean_token_accuracy": 0.19153500944375992,
      "num_tokens": 38525579.0,
      "step": 17255
    },
    {
      "entropy": 5.624566459655762,
      "epoch": 1.9385073285786487,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046317361788285436,
      "loss": 5.4925,
      "mean_token_accuracy": 0.185320383310318,
      "num_tokens": 38536963.0,
      "step": 17260
    },
    {
      "entropy": 5.534394025802612,
      "epoch": 1.9390689054866064,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004631516041294469,
      "loss": 5.3881,
      "mean_token_accuracy": 0.18485718071460724,
      "num_tokens": 38547497.0,
      "step": 17265
    },
    {
      "entropy": 5.5887068748474125,
      "epoch": 1.939630482394564,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004631295843852194,
      "loss": 5.2752,
      "mean_token_accuracy": 0.1898737668991089,
      "num_tokens": 38558840.0,
      "step": 17270
    },
    {
      "entropy": 5.5194587230682375,
      "epoch": 1.9401920593025215,
      "grad_norm": 1.34375,
      "learning_rate": 0.000463107558650873,
      "loss": 5.2474,
      "mean_token_accuracy": 0.19837641566991807,
      "num_tokens": 38568670.0,
      "step": 17275
    },
    {
      "entropy": 5.5173698425292965,
      "epoch": 1.940753636210479,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046308552692710895,
      "loss": 5.3187,
      "mean_token_accuracy": 0.19002542942762374,
      "num_tokens": 38578565.0,
      "step": 17280
    },
    {
      "entropy": 5.474986362457275,
      "epoch": 1.9413152131184366,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046306348921462876,
      "loss": 5.288,
      "mean_token_accuracy": 0.1900528132915497,
      "num_tokens": 38589126.0,
      "step": 17285
    },
    {
      "entropy": 5.606071805953979,
      "epoch": 1.9418767900263942,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004630414455141342,
      "loss": 5.4005,
      "mean_token_accuracy": 0.18429171591997145,
      "num_tokens": 38601008.0,
      "step": 17290
    },
    {
      "entropy": 5.630123853683472,
      "epoch": 1.9424383669343517,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000463019395826327,
      "loss": 5.3852,
      "mean_token_accuracy": 0.18718257397413254,
      "num_tokens": 38611756.0,
      "step": 17295
    },
    {
      "entropy": 5.587280988693237,
      "epoch": 1.9429999438423091,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046299734015190934,
      "loss": 5.2927,
      "mean_token_accuracy": 0.18605223894119263,
      "num_tokens": 38623156.0,
      "step": 17300
    },
    {
      "entropy": 5.6102910995483395,
      "epoch": 1.9435615207502668,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046297527849158344,
      "loss": 5.3622,
      "mean_token_accuracy": 0.19393292516469957,
      "num_tokens": 38633920.0,
      "step": 17305
    },
    {
      "entropy": 5.530622434616089,
      "epoch": 1.9441230976582244,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004629532108460519,
      "loss": 5.28,
      "mean_token_accuracy": 0.18657930195331573,
      "num_tokens": 38644744.0,
      "step": 17310
    },
    {
      "entropy": 5.489866828918457,
      "epoch": 1.9446846745661819,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046293113721601706,
      "loss": 5.2804,
      "mean_token_accuracy": 0.19437275379896163,
      "num_tokens": 38656524.0,
      "step": 17315
    },
    {
      "entropy": 5.491071319580078,
      "epoch": 1.9452462514741393,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046290905760218193,
      "loss": 5.3309,
      "mean_token_accuracy": 0.19527771025896073,
      "num_tokens": 38668386.0,
      "step": 17320
    },
    {
      "entropy": 5.539380073547363,
      "epoch": 1.945807828382097,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046288697200524964,
      "loss": 5.2443,
      "mean_token_accuracy": 0.18934347331523896,
      "num_tokens": 38678498.0,
      "step": 17325
    },
    {
      "entropy": 5.673765754699707,
      "epoch": 1.9463694052900544,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046286488042592325,
      "loss": 5.4794,
      "mean_token_accuracy": 0.17748477309942245,
      "num_tokens": 38690084.0,
      "step": 17330
    },
    {
      "entropy": 5.50318169593811,
      "epoch": 1.9469309821980119,
      "grad_norm": 1.375,
      "learning_rate": 0.0004628427828649062,
      "loss": 5.3193,
      "mean_token_accuracy": 0.18965124636888503,
      "num_tokens": 38701282.0,
      "step": 17335
    },
    {
      "entropy": 5.447755956649781,
      "epoch": 1.9474925591059695,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046282067932290204,
      "loss": 5.2044,
      "mean_token_accuracy": 0.19539620876312255,
      "num_tokens": 38713012.0,
      "step": 17340
    },
    {
      "entropy": 5.437713670730591,
      "epoch": 1.9480541360139272,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004627985698006146,
      "loss": 5.2604,
      "mean_token_accuracy": 0.19580089300870895,
      "num_tokens": 38723179.0,
      "step": 17345
    },
    {
      "entropy": 5.55149245262146,
      "epoch": 1.9486157129218846,
      "grad_norm": 1.46875,
      "learning_rate": 0.000462776454298748,
      "loss": 5.2899,
      "mean_token_accuracy": 0.1989804908633232,
      "num_tokens": 38733729.0,
      "step": 17350
    },
    {
      "entropy": 5.560542345046997,
      "epoch": 1.949177289829842,
      "grad_norm": 1.765625,
      "learning_rate": 0.00046275433281800613,
      "loss": 5.2477,
      "mean_token_accuracy": 0.19693769216537477,
      "num_tokens": 38743139.0,
      "step": 17355
    },
    {
      "entropy": 5.517525005340576,
      "epoch": 1.9497388667377997,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004627322053590935,
      "loss": 5.3276,
      "mean_token_accuracy": 0.18538037538528443,
      "num_tokens": 38754697.0,
      "step": 17360
    },
    {
      "entropy": 5.4607631206512455,
      "epoch": 1.9503004436457574,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004627100719227147,
      "loss": 5.3546,
      "mean_token_accuracy": 0.1893941804766655,
      "num_tokens": 38765999.0,
      "step": 17365
    },
    {
      "entropy": 5.526936388015747,
      "epoch": 1.9508620205537148,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046268793250957437,
      "loss": 5.3623,
      "mean_token_accuracy": 0.18579252511262895,
      "num_tokens": 38776912.0,
      "step": 17370
    },
    {
      "entropy": 5.6296142578125,
      "epoch": 1.9514235974616723,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046266578712037747,
      "loss": 5.3569,
      "mean_token_accuracy": 0.18661235719919206,
      "num_tokens": 38787296.0,
      "step": 17375
    },
    {
      "entropy": 5.546997785568237,
      "epoch": 1.95198517436963,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046264363575582916,
      "loss": 5.3471,
      "mean_token_accuracy": 0.18177139461040498,
      "num_tokens": 38798013.0,
      "step": 17380
    },
    {
      "entropy": 5.458870315551758,
      "epoch": 1.9525467512775876,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004626214784166347,
      "loss": 5.2857,
      "mean_token_accuracy": 0.1914055496454239,
      "num_tokens": 38808842.0,
      "step": 17385
    },
    {
      "entropy": 5.490812921524048,
      "epoch": 1.953108328185545,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004625993151034995,
      "loss": 5.23,
      "mean_token_accuracy": 0.19331282377243042,
      "num_tokens": 38819094.0,
      "step": 17390
    },
    {
      "entropy": 5.512252330780029,
      "epoch": 1.9536699050935025,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004625771458171294,
      "loss": 5.3282,
      "mean_token_accuracy": 0.1918247625231743,
      "num_tokens": 38830348.0,
      "step": 17395
    },
    {
      "entropy": 5.566256141662597,
      "epoch": 1.9542314820014601,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046255497055823025,
      "loss": 5.3399,
      "mean_token_accuracy": 0.1917381003499031,
      "num_tokens": 38841927.0,
      "step": 17400
    },
    {
      "entropy": 5.427985715866089,
      "epoch": 1.9547930589094178,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004625327893275082,
      "loss": 5.1462,
      "mean_token_accuracy": 0.19760164022445678,
      "num_tokens": 38852817.0,
      "step": 17405
    },
    {
      "entropy": 5.3728128433227536,
      "epoch": 1.9553546358173752,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046251060212566927,
      "loss": 5.2575,
      "mean_token_accuracy": 0.1909310519695282,
      "num_tokens": 38864181.0,
      "step": 17410
    },
    {
      "entropy": 5.542944860458374,
      "epoch": 1.9559162127253327,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004624884089534201,
      "loss": 5.3244,
      "mean_token_accuracy": 0.19081388413906097,
      "num_tokens": 38873827.0,
      "step": 17415
    },
    {
      "entropy": 5.567887163162231,
      "epoch": 1.9564777896332903,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004624662098114672,
      "loss": 5.3318,
      "mean_token_accuracy": 0.18709559440612794,
      "num_tokens": 38884540.0,
      "step": 17420
    },
    {
      "entropy": 5.431196308135986,
      "epoch": 1.9570393665412478,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046244400470051754,
      "loss": 5.1872,
      "mean_token_accuracy": 0.1930992603302002,
      "num_tokens": 38895600.0,
      "step": 17425
    },
    {
      "entropy": 5.472609519958496,
      "epoch": 1.9576009434492052,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000462421793621278,
      "loss": 5.2644,
      "mean_token_accuracy": 0.1968981981277466,
      "num_tokens": 38906166.0,
      "step": 17430
    },
    {
      "entropy": 5.4246618270874025,
      "epoch": 1.9581625203571629,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000462399576574456,
      "loss": 5.1718,
      "mean_token_accuracy": 0.19477488696575165,
      "num_tokens": 38916247.0,
      "step": 17435
    },
    {
      "entropy": 5.590512895584107,
      "epoch": 1.9587240972651205,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004623773535607587,
      "loss": 5.3942,
      "mean_token_accuracy": 0.18519411832094193,
      "num_tokens": 38928276.0,
      "step": 17440
    },
    {
      "entropy": 5.571941566467285,
      "epoch": 1.959285674173078,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004623551245808938,
      "loss": 5.2973,
      "mean_token_accuracy": 0.1899605467915535,
      "num_tokens": 38939460.0,
      "step": 17445
    },
    {
      "entropy": 5.4758216381073,
      "epoch": 1.9598472510810354,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004623328896355691,
      "loss": 5.2473,
      "mean_token_accuracy": 0.19702684879302979,
      "num_tokens": 38950381.0,
      "step": 17450
    },
    {
      "entropy": 5.425159168243408,
      "epoch": 1.960408827988993,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046231064872549244,
      "loss": 5.2431,
      "mean_token_accuracy": 0.19221047610044478,
      "num_tokens": 38961312.0,
      "step": 17455
    },
    {
      "entropy": 5.555955028533935,
      "epoch": 1.9609704048969507,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046228840185137215,
      "loss": 5.3428,
      "mean_token_accuracy": 0.18744023889303207,
      "num_tokens": 38974030.0,
      "step": 17460
    },
    {
      "entropy": 5.639015769958496,
      "epoch": 1.9615319818049082,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004622661490139165,
      "loss": 5.3969,
      "mean_token_accuracy": 0.18619923293590546,
      "num_tokens": 38984640.0,
      "step": 17465
    },
    {
      "entropy": 5.480964279174804,
      "epoch": 1.9620935587128656,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004622438902138341,
      "loss": 5.2907,
      "mean_token_accuracy": 0.18884125500917434,
      "num_tokens": 38996206.0,
      "step": 17470
    },
    {
      "entropy": 5.4575084209442135,
      "epoch": 1.9626551356208233,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046222162545183357,
      "loss": 5.2196,
      "mean_token_accuracy": 0.2025580108165741,
      "num_tokens": 39007765.0,
      "step": 17475
    },
    {
      "entropy": 5.498712062835693,
      "epoch": 1.963216712528781,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046219935472862386,
      "loss": 5.336,
      "mean_token_accuracy": 0.18791389167308808,
      "num_tokens": 39020691.0,
      "step": 17480
    },
    {
      "entropy": 5.477165222167969,
      "epoch": 1.9637782894367384,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046217707804491404,
      "loss": 5.2596,
      "mean_token_accuracy": 0.19700903445482254,
      "num_tokens": 39032003.0,
      "step": 17485
    },
    {
      "entropy": 5.587220335006714,
      "epoch": 1.9643398663446958,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004621547954014135,
      "loss": 5.3542,
      "mean_token_accuracy": 0.1912691980600357,
      "num_tokens": 39042783.0,
      "step": 17490
    },
    {
      "entropy": 5.501882886886596,
      "epoch": 1.9649014432526535,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004621325067988317,
      "loss": 5.2566,
      "mean_token_accuracy": 0.19201208055019378,
      "num_tokens": 39054395.0,
      "step": 17495
    },
    {
      "entropy": 5.465988636016846,
      "epoch": 1.9654630201606111,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046211021223787827,
      "loss": 5.2412,
      "mean_token_accuracy": 0.19983402639627457,
      "num_tokens": 39064619.0,
      "step": 17500
    },
    {
      "entropy": 5.610533666610718,
      "epoch": 1.9660245970685686,
      "grad_norm": 1.375,
      "learning_rate": 0.00046208791171926315,
      "loss": 5.3812,
      "mean_token_accuracy": 0.18333709388971328,
      "num_tokens": 39076112.0,
      "step": 17505
    },
    {
      "entropy": 5.487556791305542,
      "epoch": 1.966586173976526,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046206560524369636,
      "loss": 5.277,
      "mean_token_accuracy": 0.19122491031885147,
      "num_tokens": 39087669.0,
      "step": 17510
    },
    {
      "entropy": 5.493722581863404,
      "epoch": 1.9671477508844837,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046204329281188816,
      "loss": 5.2796,
      "mean_token_accuracy": 0.19010273963212967,
      "num_tokens": 39098506.0,
      "step": 17515
    },
    {
      "entropy": 5.525453519821167,
      "epoch": 1.9677093277924411,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004620209744245489,
      "loss": 5.3252,
      "mean_token_accuracy": 0.19333317130804062,
      "num_tokens": 39109441.0,
      "step": 17520
    },
    {
      "entropy": 5.5208498477935795,
      "epoch": 1.9682709047003986,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046199865008238937,
      "loss": 5.2342,
      "mean_token_accuracy": 0.19754558205604553,
      "num_tokens": 39119765.0,
      "step": 17525
    },
    {
      "entropy": 5.461196804046631,
      "epoch": 1.9688324816083562,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004619763197861203,
      "loss": 5.2053,
      "mean_token_accuracy": 0.1979791358113289,
      "num_tokens": 39129794.0,
      "step": 17530
    },
    {
      "entropy": 5.445248079299927,
      "epoch": 1.9693940585163139,
      "grad_norm": 1.375,
      "learning_rate": 0.00046195398353645256,
      "loss": 5.2251,
      "mean_token_accuracy": 0.1986946478486061,
      "num_tokens": 39140095.0,
      "step": 17535
    },
    {
      "entropy": 5.523592758178711,
      "epoch": 1.9699556354242713,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004619316413340976,
      "loss": 5.3281,
      "mean_token_accuracy": 0.19205934107303618,
      "num_tokens": 39150855.0,
      "step": 17540
    },
    {
      "entropy": 5.537033891677856,
      "epoch": 1.9705172123322288,
      "grad_norm": 1.25,
      "learning_rate": 0.00046190929317976664,
      "loss": 5.2485,
      "mean_token_accuracy": 0.19780316054821015,
      "num_tokens": 39161455.0,
      "step": 17545
    },
    {
      "entropy": 5.557616186141968,
      "epoch": 1.9710787892401864,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004618869390741713,
      "loss": 5.452,
      "mean_token_accuracy": 0.1836066022515297,
      "num_tokens": 39173199.0,
      "step": 17550
    },
    {
      "entropy": 5.576978015899658,
      "epoch": 1.971640366148144,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004618645790180233,
      "loss": 5.3265,
      "mean_token_accuracy": 0.1939622327685356,
      "num_tokens": 39183145.0,
      "step": 17555
    },
    {
      "entropy": 5.551261377334595,
      "epoch": 1.9722019430561015,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004618422130120347,
      "loss": 5.3381,
      "mean_token_accuracy": 0.1918054074048996,
      "num_tokens": 39194900.0,
      "step": 17560
    },
    {
      "entropy": 5.606948375701904,
      "epoch": 1.972763519964059,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004618198410569175,
      "loss": 5.4062,
      "mean_token_accuracy": 0.1801542103290558,
      "num_tokens": 39206651.0,
      "step": 17565
    },
    {
      "entropy": 5.48220682144165,
      "epoch": 1.9733250968720166,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046179746315338414,
      "loss": 5.3049,
      "mean_token_accuracy": 0.19232531934976577,
      "num_tokens": 39217635.0,
      "step": 17570
    },
    {
      "entropy": 5.553248882293701,
      "epoch": 1.9738866737799743,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00046177507930214714,
      "loss": 5.32,
      "mean_token_accuracy": 0.1951992079615593,
      "num_tokens": 39227937.0,
      "step": 17575
    },
    {
      "entropy": 5.525927400588989,
      "epoch": 1.9744482506879317,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004617526895039191,
      "loss": 5.3167,
      "mean_token_accuracy": 0.19994789361953735,
      "num_tokens": 39238776.0,
      "step": 17580
    },
    {
      "entropy": 5.517746210098267,
      "epoch": 1.9750098275958892,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046173029375941305,
      "loss": 5.3422,
      "mean_token_accuracy": 0.1904160737991333,
      "num_tokens": 39250043.0,
      "step": 17585
    },
    {
      "entropy": 5.452392292022705,
      "epoch": 1.9755714045038468,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000461707892069342,
      "loss": 5.2713,
      "mean_token_accuracy": 0.19367802739143372,
      "num_tokens": 39261440.0,
      "step": 17590
    },
    {
      "entropy": 5.573949670791626,
      "epoch": 1.9761329814118045,
      "grad_norm": 2.046875,
      "learning_rate": 0.00046168548443441926,
      "loss": 5.4926,
      "mean_token_accuracy": 0.1838779166340828,
      "num_tokens": 39272151.0,
      "step": 17595
    },
    {
      "entropy": 5.716833066940308,
      "epoch": 1.976694558319762,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004616630708553583,
      "loss": 5.4842,
      "mean_token_accuracy": 0.17978556901216508,
      "num_tokens": 39284794.0,
      "step": 17600
    },
    {
      "entropy": 5.511546230316162,
      "epoch": 1.9772561352277194,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046164065133287276,
      "loss": 5.3741,
      "mean_token_accuracy": 0.18695790320634842,
      "num_tokens": 39296660.0,
      "step": 17605
    },
    {
      "entropy": 5.5483684062957765,
      "epoch": 1.977817712135677,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004616182258676765,
      "loss": 5.4013,
      "mean_token_accuracy": 0.18527230769395828,
      "num_tokens": 39308609.0,
      "step": 17610
    },
    {
      "entropy": 5.561050224304199,
      "epoch": 1.9783792890436345,
      "grad_norm": 1.25,
      "learning_rate": 0.0004615957944604835,
      "loss": 5.2039,
      "mean_token_accuracy": 0.1979954198002815,
      "num_tokens": 39319295.0,
      "step": 17615
    },
    {
      "entropy": 5.566237926483154,
      "epoch": 1.978940865951592,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004615733571120081,
      "loss": 5.3266,
      "mean_token_accuracy": 0.19095246493816376,
      "num_tokens": 39330874.0,
      "step": 17620
    },
    {
      "entropy": 5.475045394897461,
      "epoch": 1.9795024428595496,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004615509138229646,
      "loss": 5.3228,
      "mean_token_accuracy": 0.18548815399408342,
      "num_tokens": 39343574.0,
      "step": 17625
    },
    {
      "entropy": 5.5390052318573,
      "epoch": 1.9800640197675072,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046152846459406765,
      "loss": 5.3286,
      "mean_token_accuracy": 0.18938498497009276,
      "num_tokens": 39355761.0,
      "step": 17630
    },
    {
      "entropy": 5.59404616355896,
      "epoch": 1.9806255966754647,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004615060094260321,
      "loss": 5.46,
      "mean_token_accuracy": 0.17910851389169694,
      "num_tokens": 39368116.0,
      "step": 17635
    },
    {
      "entropy": 5.5500890731811525,
      "epoch": 1.9811871735834221,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046148354831957285,
      "loss": 5.3128,
      "mean_token_accuracy": 0.2022144928574562,
      "num_tokens": 39378619.0,
      "step": 17640
    },
    {
      "entropy": 5.644384860992432,
      "epoch": 1.9817487504913798,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046146108127540506,
      "loss": 5.4388,
      "mean_token_accuracy": 0.1837463304400444,
      "num_tokens": 39391595.0,
      "step": 17645
    },
    {
      "entropy": 5.555820560455322,
      "epoch": 1.9823103273993374,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004614386082942442,
      "loss": 5.392,
      "mean_token_accuracy": 0.18424100875854493,
      "num_tokens": 39402494.0,
      "step": 17650
    },
    {
      "entropy": 5.5125528335571286,
      "epoch": 1.9828719043072949,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004614161293768057,
      "loss": 5.3069,
      "mean_token_accuracy": 0.18762837648391723,
      "num_tokens": 39413371.0,
      "step": 17655
    },
    {
      "entropy": 5.4820473194122314,
      "epoch": 1.9834334812152523,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046139364452380536,
      "loss": 5.3068,
      "mean_token_accuracy": 0.18936399519443511,
      "num_tokens": 39424170.0,
      "step": 17660
    },
    {
      "entropy": 5.6062427997589115,
      "epoch": 1.98399505812321,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000461371153735959,
      "loss": 5.3933,
      "mean_token_accuracy": 0.18909017294645308,
      "num_tokens": 39434438.0,
      "step": 17665
    },
    {
      "entropy": 5.569085121154785,
      "epoch": 1.9845566350311676,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046134865701398287,
      "loss": 5.3152,
      "mean_token_accuracy": 0.1891639858484268,
      "num_tokens": 39445116.0,
      "step": 17670
    },
    {
      "entropy": 5.559984445571899,
      "epoch": 1.985118211939125,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046132615435859326,
      "loss": 5.3,
      "mean_token_accuracy": 0.1926518663764,
      "num_tokens": 39457296.0,
      "step": 17675
    },
    {
      "entropy": 5.575274229049683,
      "epoch": 1.9856797888470825,
      "grad_norm": 2.09375,
      "learning_rate": 0.00046130364577050655,
      "loss": 5.3801,
      "mean_token_accuracy": 0.19589262902736665,
      "num_tokens": 39468499.0,
      "step": 17680
    },
    {
      "entropy": 5.4970090866088865,
      "epoch": 1.9862413657550402,
      "grad_norm": 2.0,
      "learning_rate": 0.00046128113125043957,
      "loss": 5.2647,
      "mean_token_accuracy": 0.19170113205909728,
      "num_tokens": 39479594.0,
      "step": 17685
    },
    {
      "entropy": 5.50139741897583,
      "epoch": 1.9868029426629978,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046125861079910913,
      "loss": 5.3264,
      "mean_token_accuracy": 0.19295964688062667,
      "num_tokens": 39490048.0,
      "step": 17690
    },
    {
      "entropy": 5.533599758148194,
      "epoch": 1.9873645195709553,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004612360844172322,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18845418840646744,
      "num_tokens": 39501454.0,
      "step": 17695
    },
    {
      "entropy": 5.5037353515625,
      "epoch": 1.9879260964789127,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046121355210552626,
      "loss": 5.2705,
      "mean_token_accuracy": 0.20236016660928727,
      "num_tokens": 39512782.0,
      "step": 17700
    },
    {
      "entropy": 5.522071170806885,
      "epoch": 1.9884876733868704,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004611910138647085,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19241473227739334,
      "num_tokens": 39524342.0,
      "step": 17705
    },
    {
      "entropy": 5.498077297210694,
      "epoch": 1.989049250294828,
      "grad_norm": 1.375,
      "learning_rate": 0.00046116846969549666,
      "loss": 5.2337,
      "mean_token_accuracy": 0.19326062351465226,
      "num_tokens": 39534744.0,
      "step": 17710
    },
    {
      "entropy": 5.5863272190094,
      "epoch": 1.9896108272027853,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004611459195986085,
      "loss": 5.3922,
      "mean_token_accuracy": 0.1856262743473053,
      "num_tokens": 39547729.0,
      "step": 17715
    },
    {
      "entropy": 5.544296026229858,
      "epoch": 1.990172404110743,
      "grad_norm": 1.703125,
      "learning_rate": 0.00046112336357476217,
      "loss": 5.2421,
      "mean_token_accuracy": 0.1931598663330078,
      "num_tokens": 39558194.0,
      "step": 17720
    },
    {
      "entropy": 5.458015441894531,
      "epoch": 1.9907339810187006,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004611008016246757,
      "loss": 5.1788,
      "mean_token_accuracy": 0.19584223181009291,
      "num_tokens": 39568265.0,
      "step": 17725
    },
    {
      "entropy": 5.533123588562011,
      "epoch": 1.991295557926658,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046107823374906754,
      "loss": 5.4039,
      "mean_token_accuracy": 0.18655110001564026,
      "num_tokens": 39578936.0,
      "step": 17730
    },
    {
      "entropy": 5.576171779632569,
      "epoch": 1.9918571348346155,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046105565994865614,
      "loss": 5.3929,
      "mean_token_accuracy": 0.18787167370319366,
      "num_tokens": 39590180.0,
      "step": 17735
    },
    {
      "entropy": 5.543496084213257,
      "epoch": 1.9924187117425731,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004610330802241605,
      "loss": 5.3289,
      "mean_token_accuracy": 0.1881959080696106,
      "num_tokens": 39601226.0,
      "step": 17740
    },
    {
      "entropy": 5.500477027893067,
      "epoch": 1.9929802886505308,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004610104945762994,
      "loss": 5.3614,
      "mean_token_accuracy": 0.19314012080430984,
      "num_tokens": 39613119.0,
      "step": 17745
    },
    {
      "entropy": 5.558902978897095,
      "epoch": 1.9935418655584882,
      "grad_norm": 2.53125,
      "learning_rate": 0.000460987903005792,
      "loss": 5.335,
      "mean_token_accuracy": 0.19106521159410478,
      "num_tokens": 39625269.0,
      "step": 17750
    },
    {
      "entropy": 5.540250253677368,
      "epoch": 1.9941034424664457,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004609653055133576,
      "loss": 5.2332,
      "mean_token_accuracy": 0.1961716890335083,
      "num_tokens": 39635883.0,
      "step": 17755
    },
    {
      "entropy": 5.466345310211182,
      "epoch": 1.9946650193744033,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046094270209971576,
      "loss": 5.2585,
      "mean_token_accuracy": 0.1985495701432228,
      "num_tokens": 39646425.0,
      "step": 17760
    },
    {
      "entropy": 5.542253732681274,
      "epoch": 1.995226596282361,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004609200927655862,
      "loss": 5.3831,
      "mean_token_accuracy": 0.1871304541826248,
      "num_tokens": 39658874.0,
      "step": 17765
    },
    {
      "entropy": 5.542967891693115,
      "epoch": 1.9957881731903184,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004608974775116888,
      "loss": 5.3358,
      "mean_token_accuracy": 0.1885174185037613,
      "num_tokens": 39670401.0,
      "step": 17770
    },
    {
      "entropy": 5.4775111198425295,
      "epoch": 1.9963497500982759,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004608748563387436,
      "loss": 5.2638,
      "mean_token_accuracy": 0.19434180855751038,
      "num_tokens": 39681005.0,
      "step": 17775
    },
    {
      "entropy": 5.652531719207763,
      "epoch": 1.9969113270062335,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004608522292474708,
      "loss": 5.4188,
      "mean_token_accuracy": 0.18446086794137956,
      "num_tokens": 39692311.0,
      "step": 17780
    },
    {
      "entropy": 5.564651441574097,
      "epoch": 1.9974729039141912,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004608295962385911,
      "loss": 5.2163,
      "mean_token_accuracy": 0.20410524904727936,
      "num_tokens": 39702782.0,
      "step": 17785
    },
    {
      "entropy": 5.481700038909912,
      "epoch": 1.9980344808221486,
      "grad_norm": 1.78125,
      "learning_rate": 0.00046080695731282484,
      "loss": 5.3719,
      "mean_token_accuracy": 0.18634167909622193,
      "num_tokens": 39713497.0,
      "step": 17790
    },
    {
      "entropy": 5.501364040374756,
      "epoch": 1.998596057730106,
      "grad_norm": 1.40625,
      "learning_rate": 0.000460784312470893,
      "loss": 5.2912,
      "mean_token_accuracy": 0.1885882779955864,
      "num_tokens": 39724355.0,
      "step": 17795
    },
    {
      "entropy": 5.64681191444397,
      "epoch": 1.9991576346380637,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00046076166171351673,
      "loss": 5.3714,
      "mean_token_accuracy": 0.18769290298223495,
      "num_tokens": 39736523.0,
      "step": 17800
    },
    {
      "entropy": 5.553428077697754,
      "epoch": 1.9997192115460214,
      "grad_norm": 1.25,
      "learning_rate": 0.000460739005041417,
      "loss": 5.3153,
      "mean_token_accuracy": 0.18943432122468948,
      "num_tokens": 39746297.0,
      "step": 17805
    },
    {
      "entropy": 5.368327299753825,
      "epoch": 2.000224630763183,
      "grad_norm": 1.5,
      "learning_rate": 0.0004607163424553153,
      "loss": 5.1191,
      "mean_token_accuracy": 0.20146067440509796,
      "num_tokens": 39755163.0,
      "step": 17810
    },
    {
      "entropy": 5.5035772800445555,
      "epoch": 2.0007862076711405,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046069367395593325,
      "loss": 5.344,
      "mean_token_accuracy": 0.1846245363354683,
      "num_tokens": 39766444.0,
      "step": 17815
    },
    {
      "entropy": 5.594197034835815,
      "epoch": 2.001347784579098,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004606709995439926,
      "loss": 5.27,
      "mean_token_accuracy": 0.19674184620380403,
      "num_tokens": 39778051.0,
      "step": 17820
    },
    {
      "entropy": 5.56218900680542,
      "epoch": 2.001909361487056,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046064831922021534,
      "loss": 5.2984,
      "mean_token_accuracy": 0.18967250138521194,
      "num_tokens": 39787850.0,
      "step": 17825
    },
    {
      "entropy": 5.579576730728149,
      "epoch": 2.002470938395013,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004606256329853235,
      "loss": 5.2277,
      "mean_token_accuracy": 0.1949354574084282,
      "num_tokens": 39799428.0,
      "step": 17830
    },
    {
      "entropy": 5.67228889465332,
      "epoch": 2.0030325153029707,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004606029408400396,
      "loss": 5.4265,
      "mean_token_accuracy": 0.18604052811861038,
      "num_tokens": 39811688.0,
      "step": 17835
    },
    {
      "entropy": 5.519627952575684,
      "epoch": 2.0035940922109283,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004605802427850861,
      "loss": 5.1929,
      "mean_token_accuracy": 0.19501004368066788,
      "num_tokens": 39822243.0,
      "step": 17840
    },
    {
      "entropy": 5.526254272460937,
      "epoch": 2.004155669118886,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046055753882118564,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19255979508161544,
      "num_tokens": 39833699.0,
      "step": 17845
    },
    {
      "entropy": 5.647405385971069,
      "epoch": 2.004717246026843,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004605348289490611,
      "loss": 5.4164,
      "mean_token_accuracy": 0.18332097232341765,
      "num_tokens": 39844177.0,
      "step": 17850
    },
    {
      "entropy": 5.5685715675354,
      "epoch": 2.005278822934801,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004605121131694358,
      "loss": 5.2962,
      "mean_token_accuracy": 0.18893337696790696,
      "num_tokens": 39856008.0,
      "step": 17855
    },
    {
      "entropy": 5.5060914039611815,
      "epoch": 2.0058403998427585,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004604893914830328,
      "loss": 5.186,
      "mean_token_accuracy": 0.1934990867972374,
      "num_tokens": 39866209.0,
      "step": 17860
    },
    {
      "entropy": 5.571491956710815,
      "epoch": 2.006401976750716,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046046666389057563,
      "loss": 5.2418,
      "mean_token_accuracy": 0.19273425042629241,
      "num_tokens": 39876986.0,
      "step": 17865
    },
    {
      "entropy": 5.462259197235108,
      "epoch": 2.0069635536586734,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000460443930392788,
      "loss": 5.2117,
      "mean_token_accuracy": 0.1915942072868347,
      "num_tokens": 39886994.0,
      "step": 17870
    },
    {
      "entropy": 5.534856605529785,
      "epoch": 2.007525130566631,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046042119099039367,
      "loss": 5.2735,
      "mean_token_accuracy": 0.19172825962305068,
      "num_tokens": 39897514.0,
      "step": 17875
    },
    {
      "entropy": 5.6715551853179935,
      "epoch": 2.0080867074745887,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046039844568411667,
      "loss": 5.4847,
      "mean_token_accuracy": 0.18386076986789704,
      "num_tokens": 39909102.0,
      "step": 17880
    },
    {
      "entropy": 5.58957953453064,
      "epoch": 2.0086482843825464,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004603756944746813,
      "loss": 5.1636,
      "mean_token_accuracy": 0.20065847486257554,
      "num_tokens": 39919450.0,
      "step": 17885
    },
    {
      "entropy": 5.45687198638916,
      "epoch": 2.0092098612905036,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046035293736281185,
      "loss": 5.2136,
      "mean_token_accuracy": 0.1846734568476677,
      "num_tokens": 39930362.0,
      "step": 17890
    },
    {
      "entropy": 5.455288505554199,
      "epoch": 2.0097714381984613,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004603301743492331,
      "loss": 5.3014,
      "mean_token_accuracy": 0.19423362314701081,
      "num_tokens": 39942047.0,
      "step": 17895
    },
    {
      "entropy": 5.581796312332154,
      "epoch": 2.010333015106419,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004603074054346696,
      "loss": 5.2827,
      "mean_token_accuracy": 0.192690934240818,
      "num_tokens": 39954870.0,
      "step": 17900
    },
    {
      "entropy": 5.542372322082519,
      "epoch": 2.0108945920143766,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004602846306198465,
      "loss": 5.277,
      "mean_token_accuracy": 0.1974707990884781,
      "num_tokens": 39966345.0,
      "step": 17905
    },
    {
      "entropy": 5.446708822250367,
      "epoch": 2.011456168922334,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004602618499054889,
      "loss": 5.1439,
      "mean_token_accuracy": 0.20305910855531692,
      "num_tokens": 39976453.0,
      "step": 17910
    },
    {
      "entropy": 5.52267370223999,
      "epoch": 2.0120177458302915,
      "grad_norm": 1.375,
      "learning_rate": 0.0004602390632923221,
      "loss": 5.3192,
      "mean_token_accuracy": 0.1889155015349388,
      "num_tokens": 39988990.0,
      "step": 17915
    },
    {
      "entropy": 5.476548767089843,
      "epoch": 2.012579322738249,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046021627078107177,
      "loss": 5.2649,
      "mean_token_accuracy": 0.19363445937633514,
      "num_tokens": 39999131.0,
      "step": 17920
    },
    {
      "entropy": 5.587501525878906,
      "epoch": 2.0131408996462063,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046019347237246343,
      "loss": 5.2734,
      "mean_token_accuracy": 0.1963441863656044,
      "num_tokens": 40010197.0,
      "step": 17925
    },
    {
      "entropy": 5.512665128707885,
      "epoch": 2.013702476554164,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004601706680672232,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20290805995464326,
      "num_tokens": 40021388.0,
      "step": 17930
    },
    {
      "entropy": 5.488842630386353,
      "epoch": 2.0142640534621217,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004601478578660771,
      "loss": 5.249,
      "mean_token_accuracy": 0.19694442749023439,
      "num_tokens": 40032485.0,
      "step": 17935
    },
    {
      "entropy": 5.497480344772339,
      "epoch": 2.0148256303700793,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004601250417697513,
      "loss": 5.3509,
      "mean_token_accuracy": 0.18763996362686158,
      "num_tokens": 40043465.0,
      "step": 17940
    },
    {
      "entropy": 5.541256999969482,
      "epoch": 2.0153872072780366,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004601022197789724,
      "loss": 5.2459,
      "mean_token_accuracy": 0.20102934986352922,
      "num_tokens": 40054934.0,
      "step": 17945
    },
    {
      "entropy": 5.561784648895264,
      "epoch": 2.015948784185994,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004600793918944671,
      "loss": 5.3272,
      "mean_token_accuracy": 0.193603977560997,
      "num_tokens": 40066159.0,
      "step": 17950
    },
    {
      "entropy": 5.434869384765625,
      "epoch": 2.016510361093952,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004600565581169621,
      "loss": 5.1608,
      "mean_token_accuracy": 0.2008107379078865,
      "num_tokens": 40076169.0,
      "step": 17955
    },
    {
      "entropy": 5.544534301757812,
      "epoch": 2.0170719380019095,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046003371844718455,
      "loss": 5.2355,
      "mean_token_accuracy": 0.19955367594957352,
      "num_tokens": 40086852.0,
      "step": 17960
    },
    {
      "entropy": 5.498943424224853,
      "epoch": 2.0176335149098668,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046001087288586164,
      "loss": 5.2003,
      "mean_token_accuracy": 0.19517784118652343,
      "num_tokens": 40097588.0,
      "step": 17965
    },
    {
      "entropy": 5.5377825736999515,
      "epoch": 2.0181950918178244,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004599880214337208,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1989293575286865,
      "num_tokens": 40108797.0,
      "step": 17970
    },
    {
      "entropy": 5.527046203613281,
      "epoch": 2.018756668725782,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004599651640914896,
      "loss": 5.3207,
      "mean_token_accuracy": 0.18557970672845842,
      "num_tokens": 40119700.0,
      "step": 17975
    },
    {
      "entropy": 5.520626354217529,
      "epoch": 2.0193182456337397,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004599423008598957,
      "loss": 5.3341,
      "mean_token_accuracy": 0.18865351974964142,
      "num_tokens": 40131611.0,
      "step": 17980
    },
    {
      "entropy": 5.5383241176605225,
      "epoch": 2.019879822541697,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045991943173966744,
      "loss": 5.2092,
      "mean_token_accuracy": 0.20510743260383607,
      "num_tokens": 40142186.0,
      "step": 17985
    },
    {
      "entropy": 5.493134546279907,
      "epoch": 2.0204413994496546,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004598965567315326,
      "loss": 5.2485,
      "mean_token_accuracy": 0.19158689975738524,
      "num_tokens": 40153856.0,
      "step": 17990
    },
    {
      "entropy": 5.443471145629883,
      "epoch": 2.0210029763576123,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004598736758362198,
      "loss": 5.1661,
      "mean_token_accuracy": 0.20336588323116303,
      "num_tokens": 40164158.0,
      "step": 17995
    },
    {
      "entropy": 5.499774360656739,
      "epoch": 2.02156455326557,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045985078905445734,
      "loss": 5.317,
      "mean_token_accuracy": 0.19301238358020784,
      "num_tokens": 40176751.0,
      "step": 18000
    },
    {
      "epoch": 2.02156455326557,
      "eval_entropy": 5.307580244365906,
      "eval_loss": 5.397276878356934,
      "eval_mean_token_accuracy": 0.1952238485528416,
      "eval_num_tokens": 40176751.0,
      "eval_runtime": 23.8922,
      "eval_samples_per_second": 1298.038,
      "eval_steps_per_second": 162.27,
      "step": 18000
    },
    {
      "entropy": 5.572664880752564,
      "epoch": 2.022126130173527,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045982789638697414,
      "loss": 5.2942,
      "mean_token_accuracy": 0.19219646900892257,
      "num_tokens": 40188000.0,
      "step": 18005
    },
    {
      "entropy": 5.58684229850769,
      "epoch": 2.022687707081485,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045980499783449897,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19214947074651717,
      "num_tokens": 40199177.0,
      "step": 18010
    },
    {
      "entropy": 5.53514928817749,
      "epoch": 2.0232492839894425,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00045978209339776104,
      "loss": 5.2818,
      "mean_token_accuracy": 0.19240461587905883,
      "num_tokens": 40210082.0,
      "step": 18015
    },
    {
      "entropy": 5.528062963485718,
      "epoch": 2.0238108608973997,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045975918307748965,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18675119578838348,
      "num_tokens": 40222503.0,
      "step": 18020
    },
    {
      "entropy": 5.530700254440307,
      "epoch": 2.0243724378053574,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004597362668744141,
      "loss": 5.2483,
      "mean_token_accuracy": 0.194814969599247,
      "num_tokens": 40233690.0,
      "step": 18025
    },
    {
      "entropy": 5.608551025390625,
      "epoch": 2.024934014713315,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004597133447892642,
      "loss": 5.3615,
      "mean_token_accuracy": 0.18277624100446702,
      "num_tokens": 40246274.0,
      "step": 18030
    },
    {
      "entropy": 5.566758012771606,
      "epoch": 2.0254955916212727,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004596904168227698,
      "loss": 5.2937,
      "mean_token_accuracy": 0.1910557433962822,
      "num_tokens": 40257469.0,
      "step": 18035
    },
    {
      "entropy": 5.496110296249389,
      "epoch": 2.02605716852923,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045966748297566085,
      "loss": 5.2682,
      "mean_token_accuracy": 0.19804832339286804,
      "num_tokens": 40268982.0,
      "step": 18040
    },
    {
      "entropy": 5.5884966373443605,
      "epoch": 2.0266187454371876,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045964454324866767,
      "loss": 5.321,
      "mean_token_accuracy": 0.19124726802110673,
      "num_tokens": 40280301.0,
      "step": 18045
    },
    {
      "entropy": 5.555725860595703,
      "epoch": 2.0271803223451452,
      "grad_norm": 1.546875,
      "learning_rate": 0.00045962159764252055,
      "loss": 5.2283,
      "mean_token_accuracy": 0.20217628479003907,
      "num_tokens": 40292058.0,
      "step": 18050
    },
    {
      "entropy": 5.50323371887207,
      "epoch": 2.027741899253103,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004595986461579502,
      "loss": 5.278,
      "mean_token_accuracy": 0.1922963097691536,
      "num_tokens": 40301957.0,
      "step": 18055
    },
    {
      "entropy": 5.54264121055603,
      "epoch": 2.02830347616106,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045957568879568736,
      "loss": 5.2499,
      "mean_token_accuracy": 0.19339917600154877,
      "num_tokens": 40313645.0,
      "step": 18060
    },
    {
      "entropy": 5.562793493270874,
      "epoch": 2.0288650530690178,
      "grad_norm": 1.5,
      "learning_rate": 0.00045955272555646306,
      "loss": 5.324,
      "mean_token_accuracy": 0.18938728421926498,
      "num_tokens": 40324416.0,
      "step": 18065
    },
    {
      "entropy": 5.667379379272461,
      "epoch": 2.0294266299769754,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004595297564410083,
      "loss": 5.4299,
      "mean_token_accuracy": 0.1929930791258812,
      "num_tokens": 40335596.0,
      "step": 18070
    },
    {
      "entropy": 5.568173885345459,
      "epoch": 2.029988206884933,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00045950678145005456,
      "loss": 5.3745,
      "mean_token_accuracy": 0.18585854768753052,
      "num_tokens": 40346426.0,
      "step": 18075
    },
    {
      "entropy": 5.45464940071106,
      "epoch": 2.0305497837928903,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004594838005843333,
      "loss": 5.1245,
      "mean_token_accuracy": 0.20244470238685608,
      "num_tokens": 40357435.0,
      "step": 18080
    },
    {
      "entropy": 5.481536197662353,
      "epoch": 2.031111360700848,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045946081384457633,
      "loss": 5.2141,
      "mean_token_accuracy": 0.19537506103515626,
      "num_tokens": 40367332.0,
      "step": 18085
    },
    {
      "entropy": 5.492844438552856,
      "epoch": 2.0316729376088056,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045943782123151556,
      "loss": 5.202,
      "mean_token_accuracy": 0.18720657825469972,
      "num_tokens": 40378571.0,
      "step": 18090
    },
    {
      "entropy": 5.509575033187867,
      "epoch": 2.0322345145167633,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004594148227458829,
      "loss": 5.2389,
      "mean_token_accuracy": 0.19426094591617585,
      "num_tokens": 40389467.0,
      "step": 18095
    },
    {
      "entropy": 5.525057935714722,
      "epoch": 2.0327960914247205,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004593918183884108,
      "loss": 5.2614,
      "mean_token_accuracy": 0.19487734884023666,
      "num_tokens": 40399785.0,
      "step": 18100
    },
    {
      "entropy": 5.477648210525513,
      "epoch": 2.033357668332678,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004593688081598318,
      "loss": 5.2425,
      "mean_token_accuracy": 0.19481801688671113,
      "num_tokens": 40409862.0,
      "step": 18105
    },
    {
      "entropy": 5.519769477844238,
      "epoch": 2.033919245240636,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00045934579206087837,
      "loss": 5.3306,
      "mean_token_accuracy": 0.18670503050088882,
      "num_tokens": 40421627.0,
      "step": 18110
    },
    {
      "entropy": 5.57765679359436,
      "epoch": 2.034480822148593,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045932277009228336,
      "loss": 5.3041,
      "mean_token_accuracy": 0.1881491169333458,
      "num_tokens": 40433327.0,
      "step": 18115
    },
    {
      "entropy": 5.469289350509643,
      "epoch": 2.0350423990565507,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045929974225477986,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19770171791315078,
      "num_tokens": 40444217.0,
      "step": 18120
    },
    {
      "entropy": 5.471521329879761,
      "epoch": 2.0356039759645084,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045927670854910117,
      "loss": 5.2566,
      "mean_token_accuracy": 0.196864578127861,
      "num_tokens": 40454760.0,
      "step": 18125
    },
    {
      "entropy": 5.540667867660522,
      "epoch": 2.036165552872466,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004592536689759806,
      "loss": 5.2724,
      "mean_token_accuracy": 0.1880270317196846,
      "num_tokens": 40464638.0,
      "step": 18130
    },
    {
      "entropy": 5.55751748085022,
      "epoch": 2.0367271297804233,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045923062353615167,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1920843541622162,
      "num_tokens": 40474911.0,
      "step": 18135
    },
    {
      "entropy": 5.519346380233765,
      "epoch": 2.037288706688381,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045920757223034833,
      "loss": 5.2607,
      "mean_token_accuracy": 0.18908546715974808,
      "num_tokens": 40485369.0,
      "step": 18140
    },
    {
      "entropy": 5.544115161895752,
      "epoch": 2.0378502835963386,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004591845150593044,
      "loss": 5.3029,
      "mean_token_accuracy": 0.18528816252946853,
      "num_tokens": 40495955.0,
      "step": 18145
    },
    {
      "entropy": 5.547362422943115,
      "epoch": 2.0384118605042962,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045916145202375407,
      "loss": 5.3241,
      "mean_token_accuracy": 0.19307520538568496,
      "num_tokens": 40506633.0,
      "step": 18150
    },
    {
      "entropy": 5.591632890701294,
      "epoch": 2.0389734374122535,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004591383831244317,
      "loss": 5.2975,
      "mean_token_accuracy": 0.18528449535369873,
      "num_tokens": 40516868.0,
      "step": 18155
    },
    {
      "entropy": 5.4388343334198,
      "epoch": 2.039535014320211,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00045911530836207173,
      "loss": 5.1988,
      "mean_token_accuracy": 0.1996513158082962,
      "num_tokens": 40528873.0,
      "step": 18160
    },
    {
      "entropy": 5.533275032043457,
      "epoch": 2.0400965912281688,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045909222773740904,
      "loss": 5.2567,
      "mean_token_accuracy": 0.18681711852550506,
      "num_tokens": 40539124.0,
      "step": 18165
    },
    {
      "entropy": 5.480309343338012,
      "epoch": 2.0406581681361264,
      "grad_norm": 1.5,
      "learning_rate": 0.00045906914125117833,
      "loss": 5.2506,
      "mean_token_accuracy": 0.1957709088921547,
      "num_tokens": 40550264.0,
      "step": 18170
    },
    {
      "entropy": 5.414032602310181,
      "epoch": 2.0412197450440837,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045904604890411476,
      "loss": 5.1305,
      "mean_token_accuracy": 0.20195517539978028,
      "num_tokens": 40562319.0,
      "step": 18175
    },
    {
      "entropy": 5.468103647232056,
      "epoch": 2.0417813219520413,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00045902295069695373,
      "loss": 5.3122,
      "mean_token_accuracy": 0.1947186753153801,
      "num_tokens": 40573962.0,
      "step": 18180
    },
    {
      "entropy": 5.465974998474121,
      "epoch": 2.042342898859999,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004589998466304305,
      "loss": 5.1894,
      "mean_token_accuracy": 0.19121768772602082,
      "num_tokens": 40584620.0,
      "step": 18185
    },
    {
      "entropy": 5.559465408325195,
      "epoch": 2.0429044757679566,
      "grad_norm": 1.453125,
      "learning_rate": 0.00045897673670528086,
      "loss": 5.3399,
      "mean_token_accuracy": 0.19162847250699996,
      "num_tokens": 40596670.0,
      "step": 18190
    },
    {
      "entropy": 5.552762269973755,
      "epoch": 2.043466052675914,
      "grad_norm": 2.0,
      "learning_rate": 0.0004589536209222404,
      "loss": 5.2521,
      "mean_token_accuracy": 0.19699085652828216,
      "num_tokens": 40607742.0,
      "step": 18195
    },
    {
      "entropy": 5.483900547027588,
      "epoch": 2.0440276295838715,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004589304992820454,
      "loss": 5.1679,
      "mean_token_accuracy": 0.20076052248477935,
      "num_tokens": 40618873.0,
      "step": 18200
    },
    {
      "entropy": 5.487114143371582,
      "epoch": 2.044589206491829,
      "grad_norm": 1.765625,
      "learning_rate": 0.000458907371785432,
      "loss": 5.2223,
      "mean_token_accuracy": 0.19629937112331391,
      "num_tokens": 40630327.0,
      "step": 18205
    },
    {
      "entropy": 5.441974925994873,
      "epoch": 2.0451507833997864,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004588842384331366,
      "loss": 5.2029,
      "mean_token_accuracy": 0.19794509708881378,
      "num_tokens": 40641515.0,
      "step": 18210
    },
    {
      "entropy": 5.5347236633300785,
      "epoch": 2.045712360307744,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045886109922589564,
      "loss": 5.2387,
      "mean_token_accuracy": 0.1935408443212509,
      "num_tokens": 40652123.0,
      "step": 18215
    },
    {
      "entropy": 5.496812772750855,
      "epoch": 2.0462739372157017,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045883795416444594,
      "loss": 5.2093,
      "mean_token_accuracy": 0.19317083507776261,
      "num_tokens": 40662895.0,
      "step": 18220
    },
    {
      "entropy": 5.437646150588989,
      "epoch": 2.0468355141236594,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045881480324952455,
      "loss": 5.1526,
      "mean_token_accuracy": 0.20172156244516373,
      "num_tokens": 40673243.0,
      "step": 18225
    },
    {
      "entropy": 5.540271282196045,
      "epoch": 2.0473970910316166,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045879164648186853,
      "loss": 5.2218,
      "mean_token_accuracy": 0.20120328664779663,
      "num_tokens": 40683861.0,
      "step": 18230
    },
    {
      "entropy": 5.514056062698364,
      "epoch": 2.0479586679395743,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045876848386221513,
      "loss": 5.1805,
      "mean_token_accuracy": 0.19275143146514892,
      "num_tokens": 40694995.0,
      "step": 18235
    },
    {
      "entropy": 5.487927627563477,
      "epoch": 2.048520244847532,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045874531539130204,
      "loss": 5.2497,
      "mean_token_accuracy": 0.19804618507623672,
      "num_tokens": 40705203.0,
      "step": 18240
    },
    {
      "entropy": 5.519061422348022,
      "epoch": 2.0490818217554896,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045872214106986684,
      "loss": 5.256,
      "mean_token_accuracy": 0.19467736333608626,
      "num_tokens": 40716439.0,
      "step": 18245
    },
    {
      "entropy": 5.4585394859313965,
      "epoch": 2.049643398663447,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045869896089864733,
      "loss": 5.214,
      "mean_token_accuracy": 0.1964782804250717,
      "num_tokens": 40727436.0,
      "step": 18250
    },
    {
      "entropy": 5.493454456329346,
      "epoch": 2.0502049755714045,
      "grad_norm": 1.375,
      "learning_rate": 0.0004586757748783817,
      "loss": 5.1926,
      "mean_token_accuracy": 0.20115760266780852,
      "num_tokens": 40738423.0,
      "step": 18255
    },
    {
      "entropy": 5.658260202407837,
      "epoch": 2.050766552479362,
      "grad_norm": 1.546875,
      "learning_rate": 0.00045865258300980817,
      "loss": 5.4002,
      "mean_token_accuracy": 0.1902313143014908,
      "num_tokens": 40749506.0,
      "step": 18260
    },
    {
      "entropy": 5.581439590454101,
      "epoch": 2.05132812938732,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045862938529366514,
      "loss": 5.2558,
      "mean_token_accuracy": 0.2016041696071625,
      "num_tokens": 40762165.0,
      "step": 18265
    },
    {
      "entropy": 5.5209489345550535,
      "epoch": 2.051889706295277,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004586061817306913,
      "loss": 5.3314,
      "mean_token_accuracy": 0.18586129695177078,
      "num_tokens": 40774043.0,
      "step": 18270
    },
    {
      "entropy": 5.446827602386475,
      "epoch": 2.0524512832032347,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045858297232162536,
      "loss": 5.211,
      "mean_token_accuracy": 0.19647075086832047,
      "num_tokens": 40784758.0,
      "step": 18275
    },
    {
      "entropy": 5.5446666240692135,
      "epoch": 2.0530128601111923,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045855975706720644,
      "loss": 5.2333,
      "mean_token_accuracy": 0.19913334995508195,
      "num_tokens": 40794735.0,
      "step": 18280
    },
    {
      "entropy": 5.438759994506836,
      "epoch": 2.05357443701915,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004585365359681736,
      "loss": 5.1257,
      "mean_token_accuracy": 0.2007010981440544,
      "num_tokens": 40805491.0,
      "step": 18285
    },
    {
      "entropy": 5.546117210388184,
      "epoch": 2.054136013927107,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045851330902526633,
      "loss": 5.3535,
      "mean_token_accuracy": 0.19056230485439302,
      "num_tokens": 40816945.0,
      "step": 18290
    },
    {
      "entropy": 5.613444900512695,
      "epoch": 2.054697590835065,
      "grad_norm": 1.25,
      "learning_rate": 0.00045849007623922406,
      "loss": 5.3182,
      "mean_token_accuracy": 0.19508860260248184,
      "num_tokens": 40828199.0,
      "step": 18295
    },
    {
      "entropy": 5.528549337387085,
      "epoch": 2.0552591677430225,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004584668376107867,
      "loss": 5.3363,
      "mean_token_accuracy": 0.19062856286764146,
      "num_tokens": 40839618.0,
      "step": 18300
    },
    {
      "entropy": 5.464557266235351,
      "epoch": 2.0558207446509797,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000458443593140694,
      "loss": 5.2184,
      "mean_token_accuracy": 0.19472844153642654,
      "num_tokens": 40851078.0,
      "step": 18305
    },
    {
      "entropy": 5.544152736663818,
      "epoch": 2.0563823215589374,
      "grad_norm": 1.25,
      "learning_rate": 0.00045842034282968615,
      "loss": 5.2072,
      "mean_token_accuracy": 0.19696956127882004,
      "num_tokens": 40862397.0,
      "step": 18310
    },
    {
      "entropy": 5.4796758651733395,
      "epoch": 2.056943898466895,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004583970866785035,
      "loss": 5.215,
      "mean_token_accuracy": 0.19224913716316222,
      "num_tokens": 40872739.0,
      "step": 18315
    },
    {
      "entropy": 5.625690078735351,
      "epoch": 2.0575054753748527,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045837382468788643,
      "loss": 5.3098,
      "mean_token_accuracy": 0.19087839126586914,
      "num_tokens": 40884723.0,
      "step": 18320
    },
    {
      "entropy": 5.544408082962036,
      "epoch": 2.05806705228281,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004583505568585757,
      "loss": 5.2939,
      "mean_token_accuracy": 0.19208989590406417,
      "num_tokens": 40895927.0,
      "step": 18325
    },
    {
      "entropy": 5.457180213928223,
      "epoch": 2.0586286291907676,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045832728319131205,
      "loss": 5.258,
      "mean_token_accuracy": 0.1920614942908287,
      "num_tokens": 40906504.0,
      "step": 18330
    },
    {
      "entropy": 5.464088249206543,
      "epoch": 2.0591902060987253,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004583040036868367,
      "loss": 5.1613,
      "mean_token_accuracy": 0.19689573645591735,
      "num_tokens": 40918420.0,
      "step": 18335
    },
    {
      "entropy": 5.5200577735900875,
      "epoch": 2.059751783006683,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004582807183458907,
      "loss": 5.2778,
      "mean_token_accuracy": 0.19521691650152206,
      "num_tokens": 40929665.0,
      "step": 18340
    },
    {
      "entropy": 5.528092098236084,
      "epoch": 2.06031335991464,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00045825742716921553,
      "loss": 5.2535,
      "mean_token_accuracy": 0.18861947506666182,
      "num_tokens": 40941134.0,
      "step": 18345
    },
    {
      "entropy": 5.539303970336914,
      "epoch": 2.060874936822598,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004582341301575529,
      "loss": 5.2767,
      "mean_token_accuracy": 0.1965376242995262,
      "num_tokens": 40953230.0,
      "step": 18350
    },
    {
      "entropy": 5.47193751335144,
      "epoch": 2.0614365137305555,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045821082731164445,
      "loss": 5.2186,
      "mean_token_accuracy": 0.1976081594824791,
      "num_tokens": 40964398.0,
      "step": 18355
    },
    {
      "entropy": 5.57206244468689,
      "epoch": 2.061998090638513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004581875186322322,
      "loss": 5.2787,
      "mean_token_accuracy": 0.19481647461652757,
      "num_tokens": 40975561.0,
      "step": 18360
    },
    {
      "entropy": 5.4444647312164305,
      "epoch": 2.0625596675464704,
      "grad_norm": 1.375,
      "learning_rate": 0.00045816420412005826,
      "loss": 5.1623,
      "mean_token_accuracy": 0.2018464043736458,
      "num_tokens": 40986393.0,
      "step": 18365
    },
    {
      "entropy": 5.467040348052978,
      "epoch": 2.063121244454428,
      "grad_norm": 1.875,
      "learning_rate": 0.00045814088377586505,
      "loss": 5.2849,
      "mean_token_accuracy": 0.19812929183244704,
      "num_tokens": 40998143.0,
      "step": 18370
    },
    {
      "entropy": 5.554757833480835,
      "epoch": 2.0636828213623857,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045811755760039513,
      "loss": 5.295,
      "mean_token_accuracy": 0.1946550041437149,
      "num_tokens": 41009747.0,
      "step": 18375
    },
    {
      "entropy": 5.511597061157227,
      "epoch": 2.0642443982703433,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045809422559439113,
      "loss": 5.2092,
      "mean_token_accuracy": 0.1940724790096283,
      "num_tokens": 41023146.0,
      "step": 18380
    },
    {
      "entropy": 5.526904296875,
      "epoch": 2.0648059751783006,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045807088775859595,
      "loss": 5.3336,
      "mean_token_accuracy": 0.1882246345281601,
      "num_tokens": 41034138.0,
      "step": 18385
    },
    {
      "entropy": 5.50452070236206,
      "epoch": 2.065367552086258,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004580475440937527,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1989658758044243,
      "num_tokens": 41045457.0,
      "step": 18390
    },
    {
      "entropy": 5.465430545806885,
      "epoch": 2.065929128994216,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045802419460060464,
      "loss": 5.2332,
      "mean_token_accuracy": 0.19728562235832214,
      "num_tokens": 41056264.0,
      "step": 18395
    },
    {
      "entropy": 5.546510696411133,
      "epoch": 2.066490705902173,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004580008392798953,
      "loss": 5.3919,
      "mean_token_accuracy": 0.18524643480777742,
      "num_tokens": 41067206.0,
      "step": 18400
    },
    {
      "entropy": 5.422588682174682,
      "epoch": 2.0670522828101308,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045797747813236815,
      "loss": 5.1246,
      "mean_token_accuracy": 0.19666744470596315,
      "num_tokens": 41077897.0,
      "step": 18405
    },
    {
      "entropy": 5.471376037597656,
      "epoch": 2.0676138597180884,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045795411115876725,
      "loss": 5.2099,
      "mean_token_accuracy": 0.19427814334630966,
      "num_tokens": 41088964.0,
      "step": 18410
    },
    {
      "entropy": 5.49049596786499,
      "epoch": 2.068175436626046,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045793073835983644,
      "loss": 5.1571,
      "mean_token_accuracy": 0.2011103078722954,
      "num_tokens": 41098818.0,
      "step": 18415
    },
    {
      "entropy": 5.587615966796875,
      "epoch": 2.0687370135340033,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045790735973631996,
      "loss": 5.364,
      "mean_token_accuracy": 0.18901024609804154,
      "num_tokens": 41111222.0,
      "step": 18420
    },
    {
      "entropy": 5.530643653869629,
      "epoch": 2.069298590441961,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045788397528896224,
      "loss": 5.2167,
      "mean_token_accuracy": 0.19260232746601105,
      "num_tokens": 41122875.0,
      "step": 18425
    },
    {
      "entropy": 5.506813621520996,
      "epoch": 2.0698601673499186,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045786058501850774,
      "loss": 5.2538,
      "mean_token_accuracy": 0.18410073071718216,
      "num_tokens": 41133869.0,
      "step": 18430
    },
    {
      "entropy": 5.508162975311279,
      "epoch": 2.0704217442578763,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045783718892570136,
      "loss": 5.1777,
      "mean_token_accuracy": 0.19572405219078065,
      "num_tokens": 41144482.0,
      "step": 18435
    },
    {
      "entropy": 5.415844249725342,
      "epoch": 2.0709833211658335,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000457813787011288,
      "loss": 5.1835,
      "mean_token_accuracy": 0.19369978755712508,
      "num_tokens": 41154997.0,
      "step": 18440
    },
    {
      "entropy": 5.39900164604187,
      "epoch": 2.071544898073791,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004577903792760127,
      "loss": 5.2225,
      "mean_token_accuracy": 0.19413264840841293,
      "num_tokens": 41167852.0,
      "step": 18445
    },
    {
      "entropy": 5.523342895507812,
      "epoch": 2.072106474981749,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004577669657206209,
      "loss": 5.2132,
      "mean_token_accuracy": 0.19986539632081984,
      "num_tokens": 41178364.0,
      "step": 18450
    },
    {
      "entropy": 5.528639316558838,
      "epoch": 2.0726680518897065,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004577435463458579,
      "loss": 5.3268,
      "mean_token_accuracy": 0.18822480142116546,
      "num_tokens": 41190003.0,
      "step": 18455
    },
    {
      "entropy": 5.46252064704895,
      "epoch": 2.0732296287976637,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004577201211524696,
      "loss": 5.1594,
      "mean_token_accuracy": 0.19574379473924636,
      "num_tokens": 41201574.0,
      "step": 18460
    },
    {
      "entropy": 5.432068252563477,
      "epoch": 2.0737912057056214,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004576966901412018,
      "loss": 5.1589,
      "mean_token_accuracy": 0.20136497765779496,
      "num_tokens": 41211913.0,
      "step": 18465
    },
    {
      "entropy": 5.419327163696289,
      "epoch": 2.074352782613579,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045767325331280044,
      "loss": 5.177,
      "mean_token_accuracy": 0.1994516834616661,
      "num_tokens": 41222626.0,
      "step": 18470
    },
    {
      "entropy": 5.540180206298828,
      "epoch": 2.0749143595215367,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004576498106680119,
      "loss": 5.3444,
      "mean_token_accuracy": 0.1827343687415123,
      "num_tokens": 41234321.0,
      "step": 18475
    },
    {
      "entropy": 5.5479310035705565,
      "epoch": 2.075475936429494,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004576263622075826,
      "loss": 5.1779,
      "mean_token_accuracy": 0.20113203227519988,
      "num_tokens": 41244889.0,
      "step": 18480
    },
    {
      "entropy": 5.574405193328857,
      "epoch": 2.0760375133374516,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000457602907932259,
      "loss": 5.3946,
      "mean_token_accuracy": 0.18477905839681624,
      "num_tokens": 41255710.0,
      "step": 18485
    },
    {
      "entropy": 5.56986894607544,
      "epoch": 2.0765990902454092,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045757944784278803,
      "loss": 5.2966,
      "mean_token_accuracy": 0.19818877577781677,
      "num_tokens": 41265786.0,
      "step": 18490
    },
    {
      "entropy": 5.501178836822509,
      "epoch": 2.0771606671533664,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004575559819399166,
      "loss": 5.2142,
      "mean_token_accuracy": 0.1905228704214096,
      "num_tokens": 41276755.0,
      "step": 18495
    },
    {
      "entropy": 5.527812051773071,
      "epoch": 2.077722244061324,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045753251022439195,
      "loss": 5.1731,
      "mean_token_accuracy": 0.19660303592681885,
      "num_tokens": 41287219.0,
      "step": 18500
    },
    {
      "entropy": 5.547694540023803,
      "epoch": 2.0782838209692818,
      "grad_norm": 1.375,
      "learning_rate": 0.0004575090326969614,
      "loss": 5.2488,
      "mean_token_accuracy": 0.19359165132045747,
      "num_tokens": 41297356.0,
      "step": 18505
    },
    {
      "entropy": 5.397096443176269,
      "epoch": 2.0788453978772394,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004574855493583723,
      "loss": 5.1614,
      "mean_token_accuracy": 0.2053090140223503,
      "num_tokens": 41307219.0,
      "step": 18510
    },
    {
      "entropy": 5.516602182388306,
      "epoch": 2.0794069747851966,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045746206020937276,
      "loss": 5.2944,
      "mean_token_accuracy": 0.19071929454803466,
      "num_tokens": 41318827.0,
      "step": 18515
    },
    {
      "entropy": 5.575522708892822,
      "epoch": 2.0799685516931543,
      "grad_norm": 1.46875,
      "learning_rate": 0.00045743856525071034,
      "loss": 5.3056,
      "mean_token_accuracy": 0.19240287989377974,
      "num_tokens": 41329497.0,
      "step": 18520
    },
    {
      "entropy": 5.531044864654541,
      "epoch": 2.080530128601112,
      "grad_norm": 1.25,
      "learning_rate": 0.00045741506448313327,
      "loss": 5.2407,
      "mean_token_accuracy": 0.1947307512164116,
      "num_tokens": 41340577.0,
      "step": 18525
    },
    {
      "entropy": 5.509425687789917,
      "epoch": 2.0810917055090696,
      "grad_norm": 1.734375,
      "learning_rate": 0.00045739155790738975,
      "loss": 5.2418,
      "mean_token_accuracy": 0.18887036442756652,
      "num_tokens": 41351685.0,
      "step": 18530
    },
    {
      "entropy": 5.650399875640869,
      "epoch": 2.081653282417027,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045736804552422836,
      "loss": 5.3743,
      "mean_token_accuracy": 0.18694644272327424,
      "num_tokens": 41363053.0,
      "step": 18535
    },
    {
      "entropy": 5.678454971313476,
      "epoch": 2.0822148593249845,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004573445273343976,
      "loss": 5.3827,
      "mean_token_accuracy": 0.18889155685901643,
      "num_tokens": 41374847.0,
      "step": 18540
    },
    {
      "entropy": 5.481970024108887,
      "epoch": 2.082776436232942,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004573210033386464,
      "loss": 5.3257,
      "mean_token_accuracy": 0.18636160343885422,
      "num_tokens": 41386114.0,
      "step": 18545
    },
    {
      "entropy": 5.462983226776123,
      "epoch": 2.0833380131409,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004572974735377238,
      "loss": 5.1213,
      "mean_token_accuracy": 0.19758780896663666,
      "num_tokens": 41397842.0,
      "step": 18550
    },
    {
      "entropy": 5.509663248062134,
      "epoch": 2.083899590048857,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004572739379323788,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19286703318357468,
      "num_tokens": 41409824.0,
      "step": 18555
    },
    {
      "entropy": 5.453081274032593,
      "epoch": 2.0844611669568147,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045725039652336105,
      "loss": 5.2259,
      "mean_token_accuracy": 0.18826664984226227,
      "num_tokens": 41422260.0,
      "step": 18560
    },
    {
      "entropy": 5.517183828353882,
      "epoch": 2.0850227438647724,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004572268493114199,
      "loss": 5.2298,
      "mean_token_accuracy": 0.1965215101838112,
      "num_tokens": 41433068.0,
      "step": 18565
    },
    {
      "entropy": 5.508039379119873,
      "epoch": 2.08558432077273,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004572032962973052,
      "loss": 5.2056,
      "mean_token_accuracy": 0.19768486469984053,
      "num_tokens": 41443759.0,
      "step": 18570
    },
    {
      "entropy": 5.49811019897461,
      "epoch": 2.0861458976806873,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045717973748176695,
      "loss": 5.2836,
      "mean_token_accuracy": 0.19462305605411528,
      "num_tokens": 41453876.0,
      "step": 18575
    },
    {
      "entropy": 5.466852378845215,
      "epoch": 2.086707474588645,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045715617286555516,
      "loss": 5.2182,
      "mean_token_accuracy": 0.19392173439264299,
      "num_tokens": 41464731.0,
      "step": 18580
    },
    {
      "entropy": 5.573261833190918,
      "epoch": 2.0872690514966026,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00045713260244942017,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18790953904390334,
      "num_tokens": 41476288.0,
      "step": 18585
    },
    {
      "entropy": 5.573249006271363,
      "epoch": 2.08783062840456,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004571090262341125,
      "loss": 5.2252,
      "mean_token_accuracy": 0.19779866933822632,
      "num_tokens": 41485968.0,
      "step": 18590
    },
    {
      "entropy": 5.522953653335572,
      "epoch": 2.0883922053125175,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045708544422038284,
      "loss": 5.2661,
      "mean_token_accuracy": 0.19240011423826217,
      "num_tokens": 41496393.0,
      "step": 18595
    },
    {
      "entropy": 5.467296504974366,
      "epoch": 2.088953782220475,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004570618564089819,
      "loss": 5.243,
      "mean_token_accuracy": 0.19548484683036804,
      "num_tokens": 41507539.0,
      "step": 18600
    },
    {
      "entropy": 5.596036577224732,
      "epoch": 2.089515359128433,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045703826280066095,
      "loss": 5.3252,
      "mean_token_accuracy": 0.1854417070746422,
      "num_tokens": 41517845.0,
      "step": 18605
    },
    {
      "entropy": 5.524791860580445,
      "epoch": 2.09007693603639,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004570146633961711,
      "loss": 5.2259,
      "mean_token_accuracy": 0.19945240616798401,
      "num_tokens": 41528204.0,
      "step": 18610
    },
    {
      "entropy": 5.528181171417236,
      "epoch": 2.0906385129443477,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045699105819626374,
      "loss": 5.276,
      "mean_token_accuracy": 0.19378704726696014,
      "num_tokens": 41538425.0,
      "step": 18615
    },
    {
      "entropy": 5.5230772495269775,
      "epoch": 2.0912000898523053,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004569674472016904,
      "loss": 5.3237,
      "mean_token_accuracy": 0.19206100851297378,
      "num_tokens": 41548762.0,
      "step": 18620
    },
    {
      "entropy": 5.452923059463501,
      "epoch": 2.091761666760263,
      "grad_norm": 1.375,
      "learning_rate": 0.0004569438304132031,
      "loss": 5.1839,
      "mean_token_accuracy": 0.199401193857193,
      "num_tokens": 41559003.0,
      "step": 18625
    },
    {
      "entropy": 5.472324323654175,
      "epoch": 2.09232324366822,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045692020783155376,
      "loss": 5.1982,
      "mean_token_accuracy": 0.1943460151553154,
      "num_tokens": 41569630.0,
      "step": 18630
    },
    {
      "entropy": 5.567695569992066,
      "epoch": 2.092884820576178,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004568965794574943,
      "loss": 5.3137,
      "mean_token_accuracy": 0.19472360759973525,
      "num_tokens": 41580671.0,
      "step": 18635
    },
    {
      "entropy": 5.5831156253814695,
      "epoch": 2.0934463974841355,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004568729452917772,
      "loss": 5.3094,
      "mean_token_accuracy": 0.18963203132152556,
      "num_tokens": 41592475.0,
      "step": 18640
    },
    {
      "entropy": 5.539744853973389,
      "epoch": 2.094007974392093,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045684930533515506,
      "loss": 5.2669,
      "mean_token_accuracy": 0.18983601927757263,
      "num_tokens": 41603675.0,
      "step": 18645
    },
    {
      "entropy": 5.532234525680542,
      "epoch": 2.0945695513000504,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00045682565958838045,
      "loss": 5.2171,
      "mean_token_accuracy": 0.19250726699829102,
      "num_tokens": 41614684.0,
      "step": 18650
    },
    {
      "entropy": 5.4813155174255375,
      "epoch": 2.095131128208008,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004568020080522063,
      "loss": 5.3144,
      "mean_token_accuracy": 0.19253091961145402,
      "num_tokens": 41625368.0,
      "step": 18655
    },
    {
      "entropy": 5.555939674377441,
      "epoch": 2.0956927051159657,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004567783507273858,
      "loss": 5.2503,
      "mean_token_accuracy": 0.19352529495954512,
      "num_tokens": 41636542.0,
      "step": 18660
    },
    {
      "entropy": 5.597741222381591,
      "epoch": 2.0962542820239234,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045675468761467204,
      "loss": 5.3119,
      "mean_token_accuracy": 0.19495112150907518,
      "num_tokens": 41646654.0,
      "step": 18665
    },
    {
      "entropy": 5.592956924438477,
      "epoch": 2.0968158589318806,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045673101871481853,
      "loss": 5.4201,
      "mean_token_accuracy": 0.1823358029127121,
      "num_tokens": 41659160.0,
      "step": 18670
    },
    {
      "entropy": 5.543447494506836,
      "epoch": 2.0973774358398383,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004567073440285789,
      "loss": 5.2523,
      "mean_token_accuracy": 0.19714139848947526,
      "num_tokens": 41670864.0,
      "step": 18675
    },
    {
      "entropy": 5.494233083724976,
      "epoch": 2.097939012747796,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00045668366355670697,
      "loss": 5.1755,
      "mean_token_accuracy": 0.19407217502593993,
      "num_tokens": 41681947.0,
      "step": 18680
    },
    {
      "entropy": 5.510460996627808,
      "epoch": 2.098500589655753,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004566599772999567,
      "loss": 5.3112,
      "mean_token_accuracy": 0.18885926455259322,
      "num_tokens": 41692625.0,
      "step": 18685
    },
    {
      "entropy": 5.617909145355225,
      "epoch": 2.099062166563711,
      "grad_norm": 2.140625,
      "learning_rate": 0.00045663628525908237,
      "loss": 5.3429,
      "mean_token_accuracy": 0.18358024060726166,
      "num_tokens": 41703118.0,
      "step": 18690
    },
    {
      "entropy": 5.574043846130371,
      "epoch": 2.0996237434716685,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004566125874348382,
      "loss": 5.2765,
      "mean_token_accuracy": 0.19572641253471373,
      "num_tokens": 41713910.0,
      "step": 18695
    },
    {
      "entropy": 5.522626638412476,
      "epoch": 2.100185320379626,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045658888382797886,
      "loss": 5.2512,
      "mean_token_accuracy": 0.19443645179271699,
      "num_tokens": 41725290.0,
      "step": 18700
    },
    {
      "entropy": 5.572107410430908,
      "epoch": 2.1007468972875833,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045656517443925897,
      "loss": 5.2982,
      "mean_token_accuracy": 0.19033897370100022,
      "num_tokens": 41736265.0,
      "step": 18705
    },
    {
      "entropy": 5.57664270401001,
      "epoch": 2.101308474195541,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004565414592694336,
      "loss": 5.3173,
      "mean_token_accuracy": 0.1888023942708969,
      "num_tokens": 41746702.0,
      "step": 18710
    },
    {
      "entropy": 5.643450450897217,
      "epoch": 2.1018700511034987,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004565177383192577,
      "loss": 5.4193,
      "mean_token_accuracy": 0.18916128873825072,
      "num_tokens": 41758570.0,
      "step": 18715
    },
    {
      "entropy": 5.602493619918823,
      "epoch": 2.1024316280114563,
      "grad_norm": 1.5625,
      "learning_rate": 0.00045649401158948664,
      "loss": 5.3028,
      "mean_token_accuracy": 0.20153228789567948,
      "num_tokens": 41769507.0,
      "step": 18720
    },
    {
      "entropy": 5.514243793487549,
      "epoch": 2.1029932049194136,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004564702790808758,
      "loss": 5.2622,
      "mean_token_accuracy": 0.20066044330596924,
      "num_tokens": 41780558.0,
      "step": 18725
    },
    {
      "entropy": 5.454886388778687,
      "epoch": 2.103554781827371,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004564465407941809,
      "loss": 5.1833,
      "mean_token_accuracy": 0.1996917173266411,
      "num_tokens": 41790518.0,
      "step": 18730
    },
    {
      "entropy": 5.53684434890747,
      "epoch": 2.104116358735329,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045642279673015787,
      "loss": 5.301,
      "mean_token_accuracy": 0.19263856112957,
      "num_tokens": 41800329.0,
      "step": 18735
    },
    {
      "entropy": 5.483575105667114,
      "epoch": 2.1046779356432865,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004563990468895625,
      "loss": 5.1409,
      "mean_token_accuracy": 0.19947306662797928,
      "num_tokens": 41811231.0,
      "step": 18740
    },
    {
      "entropy": 5.435629844665527,
      "epoch": 2.1052395125512438,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004563752912731512,
      "loss": 5.1808,
      "mean_token_accuracy": 0.19311793446540831,
      "num_tokens": 41821544.0,
      "step": 18745
    },
    {
      "entropy": 5.447556829452514,
      "epoch": 2.1058010894592014,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004563515298816803,
      "loss": 5.2246,
      "mean_token_accuracy": 0.19453134685754775,
      "num_tokens": 41832371.0,
      "step": 18750
    },
    {
      "entropy": 5.540677452087403,
      "epoch": 2.106362666367159,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045632776271590624,
      "loss": 5.2118,
      "mean_token_accuracy": 0.19812269806861876,
      "num_tokens": 41842828.0,
      "step": 18755
    },
    {
      "entropy": 5.381276988983155,
      "epoch": 2.1069242432751167,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004563039897765859,
      "loss": 5.1851,
      "mean_token_accuracy": 0.19802186340093614,
      "num_tokens": 41855128.0,
      "step": 18760
    },
    {
      "entropy": 5.4975542545318605,
      "epoch": 2.107485820183074,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004562802110644764,
      "loss": 5.2678,
      "mean_token_accuracy": 0.19369828552007676,
      "num_tokens": 41866258.0,
      "step": 18765
    },
    {
      "entropy": 5.563699722290039,
      "epoch": 2.1080473970910316,
      "grad_norm": 1.5625,
      "learning_rate": 0.00045625642658033447,
      "loss": 5.2483,
      "mean_token_accuracy": 0.19284942299127578,
      "num_tokens": 41877713.0,
      "step": 18770
    },
    {
      "entropy": 5.580010843276978,
      "epoch": 2.1086089739989893,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045623263632491766,
      "loss": 5.2598,
      "mean_token_accuracy": 0.19463108032941817,
      "num_tokens": 41888592.0,
      "step": 18775
    },
    {
      "entropy": 5.532656049728393,
      "epoch": 2.1091705509069465,
      "grad_norm": 1.625,
      "learning_rate": 0.00045620884029898344,
      "loss": 5.347,
      "mean_token_accuracy": 0.1923212856054306,
      "num_tokens": 41901024.0,
      "step": 18780
    },
    {
      "entropy": 5.5249697208404545,
      "epoch": 2.109732127814904,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045618503850328954,
      "loss": 5.2645,
      "mean_token_accuracy": 0.19209007024765015,
      "num_tokens": 41911184.0,
      "step": 18785
    },
    {
      "entropy": 5.554409790039062,
      "epoch": 2.110293704722862,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004561612309385936,
      "loss": 5.2922,
      "mean_token_accuracy": 0.1893538936972618,
      "num_tokens": 41922799.0,
      "step": 18790
    },
    {
      "entropy": 5.554865884780884,
      "epoch": 2.1108552816308195,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045613741760565385,
      "loss": 5.2531,
      "mean_token_accuracy": 0.20113448053598404,
      "num_tokens": 41933223.0,
      "step": 18795
    },
    {
      "entropy": 5.484144639968872,
      "epoch": 2.1114168585387767,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004561135985052285,
      "loss": 5.2725,
      "mean_token_accuracy": 0.1906439855694771,
      "num_tokens": 41944288.0,
      "step": 18800
    },
    {
      "entropy": 5.463064861297608,
      "epoch": 2.1119784354467344,
      "grad_norm": 1.71875,
      "learning_rate": 0.000456089773638076,
      "loss": 5.2092,
      "mean_token_accuracy": 0.19603820592164994,
      "num_tokens": 41954312.0,
      "step": 18805
    },
    {
      "entropy": 5.504269123077393,
      "epoch": 2.112540012354692,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004560659430049548,
      "loss": 5.3058,
      "mean_token_accuracy": 0.1928422585129738,
      "num_tokens": 41966871.0,
      "step": 18810
    },
    {
      "entropy": 5.491275882720947,
      "epoch": 2.1131015892626497,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004560421066066237,
      "loss": 5.204,
      "mean_token_accuracy": 0.200209341943264,
      "num_tokens": 41979087.0,
      "step": 18815
    },
    {
      "entropy": 5.499590873718262,
      "epoch": 2.113663166170607,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045601826444384175,
      "loss": 5.2582,
      "mean_token_accuracy": 0.18966676145792008,
      "num_tokens": 41989867.0,
      "step": 18820
    },
    {
      "entropy": 5.482534122467041,
      "epoch": 2.1142247430785646,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045599441651736805,
      "loss": 5.2403,
      "mean_token_accuracy": 0.20110277086496353,
      "num_tokens": 42001205.0,
      "step": 18825
    },
    {
      "entropy": 5.620621919631958,
      "epoch": 2.1147863199865222,
      "grad_norm": 1.453125,
      "learning_rate": 0.00045597056282796196,
      "loss": 5.3692,
      "mean_token_accuracy": 0.188943313062191,
      "num_tokens": 42012466.0,
      "step": 18830
    },
    {
      "entropy": 5.310191583633423,
      "epoch": 2.11534789689448,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004559467033763829,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20320245176553725,
      "num_tokens": 42022517.0,
      "step": 18835
    },
    {
      "entropy": 5.48411374092102,
      "epoch": 2.115909473802437,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004559228381633907,
      "loss": 5.1775,
      "mean_token_accuracy": 0.20534804165363313,
      "num_tokens": 42032253.0,
      "step": 18840
    },
    {
      "entropy": 5.523025321960449,
      "epoch": 2.1164710507103948,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045589896718974517,
      "loss": 5.1944,
      "mean_token_accuracy": 0.20150178372859956,
      "num_tokens": 42042943.0,
      "step": 18845
    },
    {
      "entropy": 5.4963047981262205,
      "epoch": 2.1170326276183524,
      "grad_norm": 1.375,
      "learning_rate": 0.0004558750904562063,
      "loss": 5.1569,
      "mean_token_accuracy": 0.20104045122861863,
      "num_tokens": 42054302.0,
      "step": 18850
    },
    {
      "entropy": 5.498264455795288,
      "epoch": 2.11759420452631,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004558512079635345,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18445347845554352,
      "num_tokens": 42064810.0,
      "step": 18855
    },
    {
      "entropy": 5.427771425247192,
      "epoch": 2.1181557814342673,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004558273197124899,
      "loss": 5.2823,
      "mean_token_accuracy": 0.1854138121008873,
      "num_tokens": 42076468.0,
      "step": 18860
    },
    {
      "entropy": 5.481720781326294,
      "epoch": 2.118717358342225,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045580342570383347,
      "loss": 5.194,
      "mean_token_accuracy": 0.1916309341788292,
      "num_tokens": 42087488.0,
      "step": 18865
    },
    {
      "entropy": 5.539413976669311,
      "epoch": 2.1192789352501826,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004557795259383258,
      "loss": 5.3174,
      "mean_token_accuracy": 0.18633429408073426,
      "num_tokens": 42099992.0,
      "step": 18870
    },
    {
      "entropy": 5.5242226123809814,
      "epoch": 2.11984051215814,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045575562041672794,
      "loss": 5.2672,
      "mean_token_accuracy": 0.1954810306429863,
      "num_tokens": 42111073.0,
      "step": 18875
    },
    {
      "entropy": 5.504447937011719,
      "epoch": 2.1204020890660975,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045573170913980094,
      "loss": 5.2449,
      "mean_token_accuracy": 0.19019717127084732,
      "num_tokens": 42121413.0,
      "step": 18880
    },
    {
      "entropy": 5.624015855789184,
      "epoch": 2.120963665974055,
      "grad_norm": 2.09375,
      "learning_rate": 0.00045570779210830627,
      "loss": 5.3335,
      "mean_token_accuracy": 0.18814467787742614,
      "num_tokens": 42132239.0,
      "step": 18885
    },
    {
      "entropy": 5.500932502746582,
      "epoch": 2.121525242882013,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004556838693230054,
      "loss": 5.2791,
      "mean_token_accuracy": 0.1989013895392418,
      "num_tokens": 42143124.0,
      "step": 18890
    },
    {
      "entropy": 5.544722843170166,
      "epoch": 2.12208681978997,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00045565994078466005,
      "loss": 5.2792,
      "mean_token_accuracy": 0.18843552768230437,
      "num_tokens": 42152653.0,
      "step": 18895
    },
    {
      "entropy": 5.598435783386231,
      "epoch": 2.1226483966979277,
      "grad_norm": 1.9375,
      "learning_rate": 0.00045563600649403213,
      "loss": 5.2902,
      "mean_token_accuracy": 0.19514897912740709,
      "num_tokens": 42162734.0,
      "step": 18900
    },
    {
      "entropy": 5.5806882858276365,
      "epoch": 2.1232099736058854,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004556120664518838,
      "loss": 5.3694,
      "mean_token_accuracy": 0.18368770629167558,
      "num_tokens": 42175312.0,
      "step": 18905
    },
    {
      "entropy": 5.542457532882691,
      "epoch": 2.123771550513843,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004555881206589771,
      "loss": 5.2823,
      "mean_token_accuracy": 0.19374412000179292,
      "num_tokens": 42186512.0,
      "step": 18910
    },
    {
      "entropy": 5.506154775619507,
      "epoch": 2.1243331274218002,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004555641691160746,
      "loss": 5.2942,
      "mean_token_accuracy": 0.19333634078502654,
      "num_tokens": 42197698.0,
      "step": 18915
    },
    {
      "entropy": 5.461877727508545,
      "epoch": 2.124894704329758,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000455540211823939,
      "loss": 5.2573,
      "mean_token_accuracy": 0.19873272776603698,
      "num_tokens": 42208761.0,
      "step": 18920
    },
    {
      "entropy": 5.528551912307739,
      "epoch": 2.1254562812377156,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045551624878333296,
      "loss": 5.2581,
      "mean_token_accuracy": 0.19103594720363617,
      "num_tokens": 42218920.0,
      "step": 18925
    },
    {
      "entropy": 5.556779098510742,
      "epoch": 2.1260178581456732,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045549227999501957,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19957176297903062,
      "num_tokens": 42229819.0,
      "step": 18930
    },
    {
      "entropy": 5.454245948791504,
      "epoch": 2.1265794350536305,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045546830545976203,
      "loss": 5.2067,
      "mean_token_accuracy": 0.1932344540953636,
      "num_tokens": 42241189.0,
      "step": 18935
    },
    {
      "entropy": 5.484561920166016,
      "epoch": 2.127141011961588,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045544432517832356,
      "loss": 5.1614,
      "mean_token_accuracy": 0.198034605383873,
      "num_tokens": 42252665.0,
      "step": 18940
    },
    {
      "entropy": 5.482719039916992,
      "epoch": 2.1277025888695458,
      "grad_norm": 1.375,
      "learning_rate": 0.00045542033915146787,
      "loss": 5.2222,
      "mean_token_accuracy": 0.1934773027896881,
      "num_tokens": 42264514.0,
      "step": 18945
    },
    {
      "entropy": 5.532171821594238,
      "epoch": 2.1282641657775034,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045539634737995857,
      "loss": 5.263,
      "mean_token_accuracy": 0.19248943775892258,
      "num_tokens": 42275229.0,
      "step": 18950
    },
    {
      "entropy": 5.535907793045044,
      "epoch": 2.1288257426854607,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045537234986455964,
      "loss": 5.2618,
      "mean_token_accuracy": 0.19488944113254547,
      "num_tokens": 42286819.0,
      "step": 18955
    },
    {
      "entropy": 5.54831371307373,
      "epoch": 2.1293873195934183,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004553483466060351,
      "loss": 5.2815,
      "mean_token_accuracy": 0.19568655043840408,
      "num_tokens": 42298329.0,
      "step": 18960
    },
    {
      "entropy": 5.376667451858521,
      "epoch": 2.129948896501376,
      "grad_norm": 1.25,
      "learning_rate": 0.00045532433760514925,
      "loss": 5.1636,
      "mean_token_accuracy": 0.19899864345788956,
      "num_tokens": 42308813.0,
      "step": 18965
    },
    {
      "entropy": 5.546373701095581,
      "epoch": 2.130510473409333,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004553003228626665,
      "loss": 5.3032,
      "mean_token_accuracy": 0.19323157519102097,
      "num_tokens": 42321264.0,
      "step": 18970
    },
    {
      "entropy": 5.470273780822754,
      "epoch": 2.131072050317291,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004552763023793517,
      "loss": 5.1835,
      "mean_token_accuracy": 0.19296262860298158,
      "num_tokens": 42332301.0,
      "step": 18975
    },
    {
      "entropy": 5.4789141654968265,
      "epoch": 2.1316336272252485,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004552522761559694,
      "loss": 5.1546,
      "mean_token_accuracy": 0.19569229185581208,
      "num_tokens": 42343424.0,
      "step": 18980
    },
    {
      "entropy": 5.499362564086914,
      "epoch": 2.132195204133206,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045522824419328475,
      "loss": 5.2499,
      "mean_token_accuracy": 0.19062476903200148,
      "num_tokens": 42354350.0,
      "step": 18985
    },
    {
      "entropy": 5.484207105636597,
      "epoch": 2.1327567810411634,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045520420649206283,
      "loss": 5.1596,
      "mean_token_accuracy": 0.19860177040100097,
      "num_tokens": 42365110.0,
      "step": 18990
    },
    {
      "entropy": 5.5182812213897705,
      "epoch": 2.133318357949121,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004551801630530692,
      "loss": 5.285,
      "mean_token_accuracy": 0.1936671495437622,
      "num_tokens": 42376345.0,
      "step": 18995
    },
    {
      "entropy": 5.475411128997803,
      "epoch": 2.1338799348570787,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004551561138770693,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19906097054481506,
      "num_tokens": 42387161.0,
      "step": 19000
    },
    {
      "entropy": 5.461827754974365,
      "epoch": 2.1344415117650364,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045513205896482875,
      "loss": 5.205,
      "mean_token_accuracy": 0.19788686484098433,
      "num_tokens": 42397468.0,
      "step": 19005
    },
    {
      "entropy": 5.456885766983032,
      "epoch": 2.1350030886729936,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045510799831711366,
      "loss": 5.2461,
      "mean_token_accuracy": 0.19609155505895615,
      "num_tokens": 42408547.0,
      "step": 19010
    },
    {
      "entropy": 5.428559732437134,
      "epoch": 2.1355646655809513,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045508393193469,
      "loss": 5.1474,
      "mean_token_accuracy": 0.20036545246839524,
      "num_tokens": 42420383.0,
      "step": 19015
    },
    {
      "entropy": 5.536728191375732,
      "epoch": 2.136126242488909,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004550598598183241,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19071185141801833,
      "num_tokens": 42431723.0,
      "step": 19020
    },
    {
      "entropy": 5.604503440856933,
      "epoch": 2.1366878193968666,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004550357819687825,
      "loss": 5.2857,
      "mean_token_accuracy": 0.19198835790157318,
      "num_tokens": 42443740.0,
      "step": 19025
    },
    {
      "entropy": 5.447312259674073,
      "epoch": 2.137249396304824,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004550116983868317,
      "loss": 5.2151,
      "mean_token_accuracy": 0.20047879517078399,
      "num_tokens": 42454288.0,
      "step": 19030
    },
    {
      "entropy": 5.507414150238037,
      "epoch": 2.1378109732127815,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004549876090732386,
      "loss": 5.2465,
      "mean_token_accuracy": 0.19338610917329788,
      "num_tokens": 42467576.0,
      "step": 19035
    },
    {
      "entropy": 5.53527603149414,
      "epoch": 2.138372550120739,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004549635140287703,
      "loss": 5.2209,
      "mean_token_accuracy": 0.19276407808065416,
      "num_tokens": 42477839.0,
      "step": 19040
    },
    {
      "entropy": 5.437886476516724,
      "epoch": 2.138934127028697,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045493941325419385,
      "loss": 5.1649,
      "mean_token_accuracy": 0.19419862031936647,
      "num_tokens": 42489875.0,
      "step": 19045
    },
    {
      "entropy": 5.536298608779907,
      "epoch": 2.139495703936654,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004549153067502768,
      "loss": 5.3179,
      "mean_token_accuracy": 0.18805031925439836,
      "num_tokens": 42500946.0,
      "step": 19050
    },
    {
      "entropy": 5.467712688446045,
      "epoch": 2.1400572808446117,
      "grad_norm": 1.25,
      "learning_rate": 0.00045489119451778645,
      "loss": 5.1996,
      "mean_token_accuracy": 0.19779159426689147,
      "num_tokens": 42511503.0,
      "step": 19055
    },
    {
      "entropy": 5.585913610458374,
      "epoch": 2.1406188577525693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004548670765574908,
      "loss": 5.3816,
      "mean_token_accuracy": 0.18621216118335723,
      "num_tokens": 42523436.0,
      "step": 19060
    },
    {
      "entropy": 5.525981807708741,
      "epoch": 2.1411804346605265,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00045484295287015757,
      "loss": 5.2927,
      "mean_token_accuracy": 0.1896200805902481,
      "num_tokens": 42534777.0,
      "step": 19065
    },
    {
      "entropy": 5.5501199722290036,
      "epoch": 2.141742011568484,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004548188234565551,
      "loss": 5.2539,
      "mean_token_accuracy": 0.19365346133708955,
      "num_tokens": 42545803.0,
      "step": 19070
    },
    {
      "entropy": 5.571218347549438,
      "epoch": 2.142303588476442,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045479468831745144,
      "loss": 5.2811,
      "mean_token_accuracy": 0.18831623047590257,
      "num_tokens": 42556803.0,
      "step": 19075
    },
    {
      "entropy": 5.602970314025879,
      "epoch": 2.1428651653843995,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004547705474536152,
      "loss": 5.3413,
      "mean_token_accuracy": 0.18605442792177201,
      "num_tokens": 42568357.0,
      "step": 19080
    },
    {
      "entropy": 5.487061882019043,
      "epoch": 2.1434267422923567,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045474640086581506,
      "loss": 5.2097,
      "mean_token_accuracy": 0.19671832770109177,
      "num_tokens": 42578990.0,
      "step": 19085
    },
    {
      "entropy": 5.529238843917847,
      "epoch": 2.1439883192003144,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045472224855481975,
      "loss": 5.2819,
      "mean_token_accuracy": 0.19358625262975693,
      "num_tokens": 42589823.0,
      "step": 19090
    },
    {
      "entropy": 5.5032336711883545,
      "epoch": 2.144549896108272,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045469809052139844,
      "loss": 5.2068,
      "mean_token_accuracy": 0.1942850321531296,
      "num_tokens": 42600508.0,
      "step": 19095
    },
    {
      "entropy": 5.481950139999389,
      "epoch": 2.1451114730162297,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045467392676632016,
      "loss": 5.2492,
      "mean_token_accuracy": 0.19504590183496476,
      "num_tokens": 42611061.0,
      "step": 19100
    },
    {
      "entropy": 5.525798988342285,
      "epoch": 2.145673049924187,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004546497572903543,
      "loss": 5.2237,
      "mean_token_accuracy": 0.19559953808784486,
      "num_tokens": 42621936.0,
      "step": 19105
    },
    {
      "entropy": 5.520177984237671,
      "epoch": 2.1462346268321446,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004546255820942706,
      "loss": 5.2898,
      "mean_token_accuracy": 0.19541528075933456,
      "num_tokens": 42632880.0,
      "step": 19110
    },
    {
      "entropy": 5.521885728836059,
      "epoch": 2.1467962037401023,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004546014011788386,
      "loss": 5.174,
      "mean_token_accuracy": 0.2015230044722557,
      "num_tokens": 42642819.0,
      "step": 19115
    },
    {
      "entropy": 5.482599973678589,
      "epoch": 2.14735778064806,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004545772145448284,
      "loss": 5.2632,
      "mean_token_accuracy": 0.19289801716804506,
      "num_tokens": 42654581.0,
      "step": 19120
    },
    {
      "entropy": 5.525862836837769,
      "epoch": 2.147919357556017,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045455302219301,
      "loss": 5.3143,
      "mean_token_accuracy": 0.19050929099321365,
      "num_tokens": 42665323.0,
      "step": 19125
    },
    {
      "entropy": 5.602271509170532,
      "epoch": 2.148480934463975,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004545288241241538,
      "loss": 5.3037,
      "mean_token_accuracy": 0.19070056080818176,
      "num_tokens": 42676465.0,
      "step": 19130
    },
    {
      "entropy": 5.487083911895752,
      "epoch": 2.1490425113719325,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004545046203390302,
      "loss": 5.2911,
      "mean_token_accuracy": 0.19485869854688645,
      "num_tokens": 42688242.0,
      "step": 19135
    },
    {
      "entropy": 5.505398273468018,
      "epoch": 2.14960408827989,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045448041083840973,
      "loss": 5.2802,
      "mean_token_accuracy": 0.19949985444545745,
      "num_tokens": 42699498.0,
      "step": 19140
    },
    {
      "entropy": 5.571474504470825,
      "epoch": 2.1501656651878474,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045445619562306354,
      "loss": 5.2712,
      "mean_token_accuracy": 0.19473840445280075,
      "num_tokens": 42710330.0,
      "step": 19145
    },
    {
      "entropy": 5.4964793682098385,
      "epoch": 2.150727242095805,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045443197469376243,
      "loss": 5.2784,
      "mean_token_accuracy": 0.19105039834976195,
      "num_tokens": 42721210.0,
      "step": 19150
    },
    {
      "entropy": 5.545171403884888,
      "epoch": 2.1512888190037627,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004544077480512776,
      "loss": 5.2683,
      "mean_token_accuracy": 0.1932425916194916,
      "num_tokens": 42732272.0,
      "step": 19155
    },
    {
      "entropy": 5.666682529449463,
      "epoch": 2.15185039591172,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045438351569638055,
      "loss": 5.3999,
      "mean_token_accuracy": 0.18227856308221818,
      "num_tokens": 42745737.0,
      "step": 19160
    },
    {
      "entropy": 5.538751029968262,
      "epoch": 2.1524119728196776,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045435927762984277,
      "loss": 5.2754,
      "mean_token_accuracy": 0.19352304339408874,
      "num_tokens": 42756405.0,
      "step": 19165
    },
    {
      "entropy": 5.544000720977783,
      "epoch": 2.152973549727635,
      "grad_norm": 1.25,
      "learning_rate": 0.000454335033852436,
      "loss": 5.2795,
      "mean_token_accuracy": 0.1869244635105133,
      "num_tokens": 42767991.0,
      "step": 19170
    },
    {
      "entropy": 5.5490223407745365,
      "epoch": 2.153535126635593,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004543107843649322,
      "loss": 5.3411,
      "mean_token_accuracy": 0.1835654318332672,
      "num_tokens": 42779177.0,
      "step": 19175
    },
    {
      "entropy": 5.48776159286499,
      "epoch": 2.15409670354355,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004542865291681035,
      "loss": 5.2396,
      "mean_token_accuracy": 0.2039228767156601,
      "num_tokens": 42789748.0,
      "step": 19180
    },
    {
      "entropy": 5.585640907287598,
      "epoch": 2.1546582804515078,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045426226826272213,
      "loss": 5.3021,
      "mean_token_accuracy": 0.19389602839946746,
      "num_tokens": 42802741.0,
      "step": 19185
    },
    {
      "entropy": 5.556554841995239,
      "epoch": 2.1552198573594654,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004542380016495607,
      "loss": 5.2857,
      "mean_token_accuracy": 0.18895589113235473,
      "num_tokens": 42813091.0,
      "step": 19190
    },
    {
      "entropy": 5.544634866714477,
      "epoch": 2.155781434267423,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045421372932939177,
      "loss": 5.2909,
      "mean_token_accuracy": 0.1943572133779526,
      "num_tokens": 42823711.0,
      "step": 19195
    },
    {
      "entropy": 5.430270099639893,
      "epoch": 2.1563430111753803,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004541894513029881,
      "loss": 5.2306,
      "mean_token_accuracy": 0.17984076887369155,
      "num_tokens": 42835057.0,
      "step": 19200
    },
    {
      "entropy": 5.62523398399353,
      "epoch": 2.156904588083338,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045416516757112286,
      "loss": 5.3007,
      "mean_token_accuracy": 0.19330391734838487,
      "num_tokens": 42846784.0,
      "step": 19205
    },
    {
      "entropy": 5.54594054222107,
      "epoch": 2.1574661649912956,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045414087813456925,
      "loss": 5.2205,
      "mean_token_accuracy": 0.19698640555143357,
      "num_tokens": 42857618.0,
      "step": 19210
    },
    {
      "entropy": 5.509330224990845,
      "epoch": 2.1580277418992533,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004541165829941006,
      "loss": 5.2704,
      "mean_token_accuracy": 0.1926967903971672,
      "num_tokens": 42869076.0,
      "step": 19215
    },
    {
      "entropy": 5.469746065139771,
      "epoch": 2.1585893188072105,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004540922821504904,
      "loss": 5.2445,
      "mean_token_accuracy": 0.19599961191415788,
      "num_tokens": 42879714.0,
      "step": 19220
    },
    {
      "entropy": 5.562300205230713,
      "epoch": 2.159150895715168,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004540679756045125,
      "loss": 5.2632,
      "mean_token_accuracy": 0.19898033887147903,
      "num_tokens": 42890187.0,
      "step": 19225
    },
    {
      "entropy": 5.476594686508179,
      "epoch": 2.159712472623126,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004540436633569407,
      "loss": 5.212,
      "mean_token_accuracy": 0.195633165538311,
      "num_tokens": 42900713.0,
      "step": 19230
    },
    {
      "entropy": 5.4860940933227536,
      "epoch": 2.1602740495310835,
      "grad_norm": 1.546875,
      "learning_rate": 0.00045401934540854937,
      "loss": 5.2614,
      "mean_token_accuracy": 0.19223775267601012,
      "num_tokens": 42911703.0,
      "step": 19235
    },
    {
      "entropy": 5.465535259246826,
      "epoch": 2.1608356264390407,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004539950217601127,
      "loss": 5.2205,
      "mean_token_accuracy": 0.19699688851833344,
      "num_tokens": 42922684.0,
      "step": 19240
    },
    {
      "entropy": 5.531664657592773,
      "epoch": 2.1613972033469984,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045397069241240497,
      "loss": 5.2297,
      "mean_token_accuracy": 0.19130341410636903,
      "num_tokens": 42933095.0,
      "step": 19245
    },
    {
      "entropy": 5.508335542678833,
      "epoch": 2.161958780254956,
      "grad_norm": 1.078125,
      "learning_rate": 0.000453946357366201,
      "loss": 5.3344,
      "mean_token_accuracy": 0.19552526026964187,
      "num_tokens": 42944508.0,
      "step": 19250
    },
    {
      "entropy": 5.432960367202758,
      "epoch": 2.1625203571629132,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045392201662227565,
      "loss": 5.1937,
      "mean_token_accuracy": 0.20042930394411088,
      "num_tokens": 42954902.0,
      "step": 19255
    },
    {
      "entropy": 5.507016992568969,
      "epoch": 2.163081934070871,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045389767018140395,
      "loss": 5.1887,
      "mean_token_accuracy": 0.19380445778369904,
      "num_tokens": 42965165.0,
      "step": 19260
    },
    {
      "entropy": 5.495782041549683,
      "epoch": 2.1636435109788286,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00045387331804436096,
      "loss": 5.2073,
      "mean_token_accuracy": 0.1978279322385788,
      "num_tokens": 42975025.0,
      "step": 19265
    },
    {
      "entropy": 5.5999737739562985,
      "epoch": 2.1642050878867862,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004538489602119222,
      "loss": 5.3737,
      "mean_token_accuracy": 0.1898595005273819,
      "num_tokens": 42986666.0,
      "step": 19270
    },
    {
      "entropy": 5.559444284439087,
      "epoch": 2.1647666647947434,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004538245966848632,
      "loss": 5.3116,
      "mean_token_accuracy": 0.1925888329744339,
      "num_tokens": 42998607.0,
      "step": 19275
    },
    {
      "entropy": 5.489432144165039,
      "epoch": 2.165328241702701,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045380022746395953,
      "loss": 5.2562,
      "mean_token_accuracy": 0.1940259099006653,
      "num_tokens": 43009969.0,
      "step": 19280
    },
    {
      "entropy": 5.631491947174072,
      "epoch": 2.1658898186106588,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045377585254998743,
      "loss": 5.4061,
      "mean_token_accuracy": 0.18729043006896973,
      "num_tokens": 43023255.0,
      "step": 19285
    },
    {
      "entropy": 5.515974521636963,
      "epoch": 2.1664513955186164,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004537514719437228,
      "loss": 5.2342,
      "mean_token_accuracy": 0.19376095682382583,
      "num_tokens": 43033995.0,
      "step": 19290
    },
    {
      "entropy": 5.524006271362305,
      "epoch": 2.1670129724265736,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045372708564594196,
      "loss": 5.3422,
      "mean_token_accuracy": 0.18745407909154893,
      "num_tokens": 43046239.0,
      "step": 19295
    },
    {
      "entropy": 5.482912874221801,
      "epoch": 2.1675745493345313,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045370269365742143,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19853470623493194,
      "num_tokens": 43057262.0,
      "step": 19300
    },
    {
      "entropy": 5.552854442596436,
      "epoch": 2.168136126242489,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045367829597893776,
      "loss": 5.3334,
      "mean_token_accuracy": 0.1878657042980194,
      "num_tokens": 43068495.0,
      "step": 19305
    },
    {
      "entropy": 5.47726035118103,
      "epoch": 2.1686977031504466,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045365389261126793,
      "loss": 5.2281,
      "mean_token_accuracy": 0.19548221826553344,
      "num_tokens": 43080706.0,
      "step": 19310
    },
    {
      "entropy": 5.40993070602417,
      "epoch": 2.169259280058404,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045362948355518883,
      "loss": 5.2152,
      "mean_token_accuracy": 0.2017691984772682,
      "num_tokens": 43091444.0,
      "step": 19315
    },
    {
      "entropy": 5.591124582290649,
      "epoch": 2.1698208569663615,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004536050688114777,
      "loss": 5.3168,
      "mean_token_accuracy": 0.18960983604192733,
      "num_tokens": 43103648.0,
      "step": 19320
    },
    {
      "entropy": 5.622720718383789,
      "epoch": 2.170382433874319,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045358064838091187,
      "loss": 5.3027,
      "mean_token_accuracy": 0.1937421292066574,
      "num_tokens": 43116183.0,
      "step": 19325
    },
    {
      "entropy": 5.513739442825317,
      "epoch": 2.170944010782277,
      "grad_norm": 1.25,
      "learning_rate": 0.00045355622226426903,
      "loss": 5.2599,
      "mean_token_accuracy": 0.19330281019210815,
      "num_tokens": 43126557.0,
      "step": 19330
    },
    {
      "entropy": 5.492511558532715,
      "epoch": 2.171505587690234,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045353179046232674,
      "loss": 5.2923,
      "mean_token_accuracy": 0.18933189362287522,
      "num_tokens": 43137753.0,
      "step": 19335
    },
    {
      "entropy": 5.576199960708618,
      "epoch": 2.1720671645981917,
      "grad_norm": 1.34375,
      "learning_rate": 0.000453507352975863,
      "loss": 5.3037,
      "mean_token_accuracy": 0.19113630056381226,
      "num_tokens": 43148461.0,
      "step": 19340
    },
    {
      "entropy": 5.485734653472901,
      "epoch": 2.1726287415061494,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045348290980565595,
      "loss": 5.2074,
      "mean_token_accuracy": 0.19547132551670074,
      "num_tokens": 43161174.0,
      "step": 19345
    },
    {
      "entropy": 5.534107446670532,
      "epoch": 2.1731903184141066,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045345846095248376,
      "loss": 5.3086,
      "mean_token_accuracy": 0.18765278458595275,
      "num_tokens": 43172844.0,
      "step": 19350
    },
    {
      "entropy": 5.435306215286255,
      "epoch": 2.1737518953220643,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045343400641712503,
      "loss": 5.1853,
      "mean_token_accuracy": 0.19814585745334626,
      "num_tokens": 43183476.0,
      "step": 19355
    },
    {
      "entropy": 5.5760307788848875,
      "epoch": 2.174313472230022,
      "grad_norm": 1.25,
      "learning_rate": 0.00045340954620035827,
      "loss": 5.2708,
      "mean_token_accuracy": 0.18906071782112122,
      "num_tokens": 43195563.0,
      "step": 19360
    },
    {
      "entropy": 5.560254907608032,
      "epoch": 2.1748750491379796,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045338508030296246,
      "loss": 5.3118,
      "mean_token_accuracy": 0.19427053928375243,
      "num_tokens": 43206067.0,
      "step": 19365
    },
    {
      "entropy": 5.569857263565064,
      "epoch": 2.175436626045937,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045336060872571646,
      "loss": 5.2062,
      "mean_token_accuracy": 0.20014894902706146,
      "num_tokens": 43216912.0,
      "step": 19370
    },
    {
      "entropy": 5.59420075416565,
      "epoch": 2.1759982029538945,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004533361314693994,
      "loss": 5.262,
      "mean_token_accuracy": 0.19414740800857544,
      "num_tokens": 43228746.0,
      "step": 19375
    },
    {
      "entropy": 5.496267700195313,
      "epoch": 2.176559779861852,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045331164853479084,
      "loss": 5.2727,
      "mean_token_accuracy": 0.18966237902641297,
      "num_tokens": 43241228.0,
      "step": 19380
    },
    {
      "entropy": 5.47652006149292,
      "epoch": 2.17712135676981,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045328715992267024,
      "loss": 5.3169,
      "mean_token_accuracy": 0.19196716099977493,
      "num_tokens": 43252916.0,
      "step": 19385
    },
    {
      "entropy": 5.528986120223999,
      "epoch": 2.177682933677767,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045326266563381734,
      "loss": 5.1475,
      "mean_token_accuracy": 0.20205250829458238,
      "num_tokens": 43264357.0,
      "step": 19390
    },
    {
      "entropy": 5.483857774734497,
      "epoch": 2.1782445105857247,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045323816566901205,
      "loss": 5.2905,
      "mean_token_accuracy": 0.19976740777492524,
      "num_tokens": 43274547.0,
      "step": 19395
    },
    {
      "entropy": 5.432673501968384,
      "epoch": 2.1788060874936823,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004532136600290343,
      "loss": 5.1778,
      "mean_token_accuracy": 0.1993179976940155,
      "num_tokens": 43284225.0,
      "step": 19400
    },
    {
      "entropy": 5.515079498291016,
      "epoch": 2.17936766440164,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004531891487146646,
      "loss": 5.2657,
      "mean_token_accuracy": 0.19107242226600646,
      "num_tokens": 43295218.0,
      "step": 19405
    },
    {
      "entropy": 5.449270248413086,
      "epoch": 2.179929241309597,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004531646317266833,
      "loss": 5.2055,
      "mean_token_accuracy": 0.19958023428916932,
      "num_tokens": 43305169.0,
      "step": 19410
    },
    {
      "entropy": 5.487324333190918,
      "epoch": 2.180490818217555,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045314010906587096,
      "loss": 5.1992,
      "mean_token_accuracy": 0.19760687202215194,
      "num_tokens": 43317119.0,
      "step": 19415
    },
    {
      "entropy": 5.540035152435303,
      "epoch": 2.1810523951255125,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045311558073300853,
      "loss": 5.2468,
      "mean_token_accuracy": 0.19819342941045762,
      "num_tokens": 43327800.0,
      "step": 19420
    },
    {
      "entropy": 5.427812194824218,
      "epoch": 2.18161397203347,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004530910467288769,
      "loss": 5.167,
      "mean_token_accuracy": 0.19073342978954316,
      "num_tokens": 43338607.0,
      "step": 19425
    },
    {
      "entropy": 5.5063619136810305,
      "epoch": 2.1821755489414274,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00045306650705425727,
      "loss": 5.2748,
      "mean_token_accuracy": 0.2007783278822899,
      "num_tokens": 43350621.0,
      "step": 19430
    },
    {
      "entropy": 5.451795053482056,
      "epoch": 2.182737125849385,
      "grad_norm": 1.375,
      "learning_rate": 0.00045304196170993093,
      "loss": 5.2293,
      "mean_token_accuracy": 0.19908606112003327,
      "num_tokens": 43360327.0,
      "step": 19435
    },
    {
      "entropy": 5.433092975616455,
      "epoch": 2.1832987027573427,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004530174106966795,
      "loss": 5.2496,
      "mean_token_accuracy": 0.19363509863615036,
      "num_tokens": 43371538.0,
      "step": 19440
    },
    {
      "entropy": 5.579623603820801,
      "epoch": 2.1838602796653,
      "grad_norm": 2.09375,
      "learning_rate": 0.00045299285401528475,
      "loss": 5.2938,
      "mean_token_accuracy": 0.1886764571070671,
      "num_tokens": 43381818.0,
      "step": 19445
    },
    {
      "entropy": 5.508121538162231,
      "epoch": 2.1844218565732576,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004529682916665285,
      "loss": 5.2413,
      "mean_token_accuracy": 0.18868270218372346,
      "num_tokens": 43391977.0,
      "step": 19450
    },
    {
      "entropy": 5.557366466522216,
      "epoch": 2.1849834334812153,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045294372365119276,
      "loss": 5.2948,
      "mean_token_accuracy": 0.19008123278617858,
      "num_tokens": 43403657.0,
      "step": 19455
    },
    {
      "entropy": 5.4605669498443605,
      "epoch": 2.185545010389173,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004529191499700598,
      "loss": 5.1872,
      "mean_token_accuracy": 0.1950461596250534,
      "num_tokens": 43414325.0,
      "step": 19460
    },
    {
      "entropy": 5.612634515762329,
      "epoch": 2.18610658729713,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004528945706239122,
      "loss": 5.3737,
      "mean_token_accuracy": 0.1871746674180031,
      "num_tokens": 43425864.0,
      "step": 19465
    },
    {
      "entropy": 5.569682264328003,
      "epoch": 2.186668164205088,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045286998561353236,
      "loss": 5.2945,
      "mean_token_accuracy": 0.19014448523521424,
      "num_tokens": 43437135.0,
      "step": 19470
    },
    {
      "entropy": 5.616008567810058,
      "epoch": 2.1872297411130455,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004528453949397033,
      "loss": 5.3293,
      "mean_token_accuracy": 0.19414877742528916,
      "num_tokens": 43448603.0,
      "step": 19475
    },
    {
      "entropy": 5.4768781661987305,
      "epoch": 2.187791318021003,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045282079860320786,
      "loss": 5.1723,
      "mean_token_accuracy": 0.20090740621089936,
      "num_tokens": 43460316.0,
      "step": 19480
    },
    {
      "entropy": 5.444617891311646,
      "epoch": 2.1883528949289603,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045279619660482925,
      "loss": 5.24,
      "mean_token_accuracy": 0.19947486370801926,
      "num_tokens": 43471869.0,
      "step": 19485
    },
    {
      "entropy": 5.498412704467773,
      "epoch": 2.188914471836918,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045277158894535075,
      "loss": 5.3088,
      "mean_token_accuracy": 0.19135208129882814,
      "num_tokens": 43484004.0,
      "step": 19490
    },
    {
      "entropy": 5.5440808773040775,
      "epoch": 2.1894760487448757,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045274697562555584,
      "loss": 5.2708,
      "mean_token_accuracy": 0.19465432912111283,
      "num_tokens": 43494401.0,
      "step": 19495
    },
    {
      "entropy": 5.435019683837891,
      "epoch": 2.1900376256528333,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004527223566462284,
      "loss": 5.2078,
      "mean_token_accuracy": 0.19536787867546082,
      "num_tokens": 43505437.0,
      "step": 19500
    },
    {
      "entropy": 5.445728874206543,
      "epoch": 2.1905992025607905,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004526977320081522,
      "loss": 5.1652,
      "mean_token_accuracy": 0.2000671371817589,
      "num_tokens": 43516177.0,
      "step": 19505
    },
    {
      "entropy": 5.536247062683105,
      "epoch": 2.191160779468748,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004526731017121111,
      "loss": 5.249,
      "mean_token_accuracy": 0.19590997099876403,
      "num_tokens": 43527070.0,
      "step": 19510
    },
    {
      "entropy": 5.512525987625122,
      "epoch": 2.191722356376706,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045264846575888973,
      "loss": 5.2887,
      "mean_token_accuracy": 0.19077699780464172,
      "num_tokens": 43538247.0,
      "step": 19515
    },
    {
      "entropy": 5.5083921432495115,
      "epoch": 2.1922839332846635,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045262382414927235,
      "loss": 5.2159,
      "mean_token_accuracy": 0.19949048310518264,
      "num_tokens": 43549994.0,
      "step": 19520
    },
    {
      "entropy": 5.587218761444092,
      "epoch": 2.1928455101926208,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045259917688404336,
      "loss": 5.3478,
      "mean_token_accuracy": 0.18835411220788956,
      "num_tokens": 43562171.0,
      "step": 19525
    },
    {
      "entropy": 5.538046073913574,
      "epoch": 2.1934070871005784,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004525745239639877,
      "loss": 5.2338,
      "mean_token_accuracy": 0.1900772526860237,
      "num_tokens": 43572308.0,
      "step": 19530
    },
    {
      "entropy": 5.439638900756836,
      "epoch": 2.193968664008536,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045254986538989045,
      "loss": 5.2285,
      "mean_token_accuracy": 0.1969294786453247,
      "num_tokens": 43583565.0,
      "step": 19535
    },
    {
      "entropy": 5.52717399597168,
      "epoch": 2.1945302409164933,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004525252011625366,
      "loss": 5.3539,
      "mean_token_accuracy": 0.19046106040477753,
      "num_tokens": 43595921.0,
      "step": 19540
    },
    {
      "entropy": 5.534983587265015,
      "epoch": 2.195091817824451,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004525005312827114,
      "loss": 5.2135,
      "mean_token_accuracy": 0.1939015194773674,
      "num_tokens": 43606957.0,
      "step": 19545
    },
    {
      "entropy": 5.460388612747193,
      "epoch": 2.1956533947324086,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004524758557512005,
      "loss": 5.1451,
      "mean_token_accuracy": 0.1989664301276207,
      "num_tokens": 43616983.0,
      "step": 19550
    },
    {
      "entropy": 5.371473550796509,
      "epoch": 2.1962149716403663,
      "grad_norm": 1.25,
      "learning_rate": 0.0004524511745687895,
      "loss": 5.0985,
      "mean_token_accuracy": 0.20401592552661896,
      "num_tokens": 43626951.0,
      "step": 19555
    },
    {
      "entropy": 5.467970275878907,
      "epoch": 2.1967765485483235,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004524264877362643,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19011069536209108,
      "num_tokens": 43638446.0,
      "step": 19560
    },
    {
      "entropy": 5.563729953765869,
      "epoch": 2.197338125456281,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045240179525441094,
      "loss": 5.298,
      "mean_token_accuracy": 0.19505397528409957,
      "num_tokens": 43649646.0,
      "step": 19565
    },
    {
      "entropy": 5.522436189651489,
      "epoch": 2.197899702364239,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004523770971240156,
      "loss": 5.3053,
      "mean_token_accuracy": 0.1905547097325325,
      "num_tokens": 43661633.0,
      "step": 19570
    },
    {
      "entropy": 5.519252109527588,
      "epoch": 2.1984612792721965,
      "grad_norm": 1.25,
      "learning_rate": 0.0004523523933458647,
      "loss": 5.2681,
      "mean_token_accuracy": 0.19578531235456467,
      "num_tokens": 43672671.0,
      "step": 19575
    },
    {
      "entropy": 5.544933891296386,
      "epoch": 2.1990228561801537,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004523276839207448,
      "loss": 5.2544,
      "mean_token_accuracy": 0.1938066765666008,
      "num_tokens": 43683426.0,
      "step": 19580
    },
    {
      "entropy": 5.501959133148193,
      "epoch": 2.1995844330881114,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045230296884944264,
      "loss": 5.3057,
      "mean_token_accuracy": 0.19192762821912765,
      "num_tokens": 43694460.0,
      "step": 19585
    },
    {
      "entropy": 5.433472204208374,
      "epoch": 2.200146009996069,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045227824813274517,
      "loss": 5.1042,
      "mean_token_accuracy": 0.19947481900453568,
      "num_tokens": 43706047.0,
      "step": 19590
    },
    {
      "entropy": 5.42898736000061,
      "epoch": 2.2007075869040267,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004522535217714396,
      "loss": 5.2102,
      "mean_token_accuracy": 0.19789150953292847,
      "num_tokens": 43715879.0,
      "step": 19595
    },
    {
      "entropy": 5.480247116088867,
      "epoch": 2.201269163811984,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004522287897663131,
      "loss": 5.2518,
      "mean_token_accuracy": 0.1936460942029953,
      "num_tokens": 43726870.0,
      "step": 19600
    },
    {
      "entropy": 5.516616773605347,
      "epoch": 2.2018307407199416,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045220405211815327,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18914685398340225,
      "num_tokens": 43739116.0,
      "step": 19605
    },
    {
      "entropy": 5.528266954421997,
      "epoch": 2.202392317627899,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045217930882774754,
      "loss": 5.2623,
      "mean_token_accuracy": 0.19535401314496995,
      "num_tokens": 43750239.0,
      "step": 19610
    },
    {
      "entropy": 5.587613439559936,
      "epoch": 2.202953894535857,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045215455989588396,
      "loss": 5.3587,
      "mean_token_accuracy": 0.19037794768810273,
      "num_tokens": 43762391.0,
      "step": 19615
    },
    {
      "entropy": 5.461773538589478,
      "epoch": 2.203515471443814,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045212980532335055,
      "loss": 5.2609,
      "mean_token_accuracy": 0.19787550270557402,
      "num_tokens": 43773514.0,
      "step": 19620
    },
    {
      "entropy": 5.482375383377075,
      "epoch": 2.2040770483517718,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045210504511093534,
      "loss": 5.2409,
      "mean_token_accuracy": 0.18991211652755738,
      "num_tokens": 43784617.0,
      "step": 19625
    },
    {
      "entropy": 5.592576026916504,
      "epoch": 2.2046386252597294,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004520802792594269,
      "loss": 5.3333,
      "mean_token_accuracy": 0.19061177223920822,
      "num_tokens": 43795396.0,
      "step": 19630
    },
    {
      "entropy": 5.5489280223846436,
      "epoch": 2.2052002021676866,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045205550776961366,
      "loss": 5.2761,
      "mean_token_accuracy": 0.18839634358882903,
      "num_tokens": 43805512.0,
      "step": 19635
    },
    {
      "entropy": 5.564304447174072,
      "epoch": 2.2057617790756443,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004520307306422844,
      "loss": 5.3972,
      "mean_token_accuracy": 0.1884940579533577,
      "num_tokens": 43815874.0,
      "step": 19640
    },
    {
      "entropy": 5.5941956520080565,
      "epoch": 2.206323355983602,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045200594787822795,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18378949761390687,
      "num_tokens": 43827948.0,
      "step": 19645
    },
    {
      "entropy": 5.535559368133545,
      "epoch": 2.2068849328915596,
      "grad_norm": 1.25,
      "learning_rate": 0.0004519811594782335,
      "loss": 5.1839,
      "mean_token_accuracy": 0.19542222768068312,
      "num_tokens": 43838306.0,
      "step": 19650
    },
    {
      "entropy": 5.485494756698609,
      "epoch": 2.207446509799517,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045195636544309033,
      "loss": 5.2243,
      "mean_token_accuracy": 0.19710462540388107,
      "num_tokens": 43848716.0,
      "step": 19655
    },
    {
      "entropy": 5.551632356643677,
      "epoch": 2.2080080867074745,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045193156577358784,
      "loss": 5.3156,
      "mean_token_accuracy": 0.19411621540784835,
      "num_tokens": 43859707.0,
      "step": 19660
    },
    {
      "entropy": 5.48183274269104,
      "epoch": 2.208569663615432,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004519067604705157,
      "loss": 5.1787,
      "mean_token_accuracy": 0.20226583778858184,
      "num_tokens": 43870599.0,
      "step": 19665
    },
    {
      "entropy": 5.541427850723267,
      "epoch": 2.20913124052339,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045188194953466363,
      "loss": 5.2715,
      "mean_token_accuracy": 0.19070045202970504,
      "num_tokens": 43881843.0,
      "step": 19670
    },
    {
      "entropy": 5.483605813980103,
      "epoch": 2.209692817431347,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045185713296682175,
      "loss": 5.164,
      "mean_token_accuracy": 0.1961521476507187,
      "num_tokens": 43892206.0,
      "step": 19675
    },
    {
      "entropy": 5.513283586502075,
      "epoch": 2.2102543943393047,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045183231076778005,
      "loss": 5.2411,
      "mean_token_accuracy": 0.19523411244153976,
      "num_tokens": 43905255.0,
      "step": 19680
    },
    {
      "entropy": 5.473288345336914,
      "epoch": 2.2108159712472624,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045180748293832906,
      "loss": 5.3078,
      "mean_token_accuracy": 0.18868002742528917,
      "num_tokens": 43917774.0,
      "step": 19685
    },
    {
      "entropy": 5.468965673446656,
      "epoch": 2.21137754815522,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045178264947925924,
      "loss": 5.2187,
      "mean_token_accuracy": 0.20265576839447022,
      "num_tokens": 43929127.0,
      "step": 19690
    },
    {
      "entropy": 5.583199739456177,
      "epoch": 2.2119391250631772,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045175781039136133,
      "loss": 5.2709,
      "mean_token_accuracy": 0.18844543993473054,
      "num_tokens": 43942558.0,
      "step": 19695
    },
    {
      "entropy": 5.471088981628418,
      "epoch": 2.212500701971135,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004517329656754261,
      "loss": 5.2716,
      "mean_token_accuracy": 0.1902107387781143,
      "num_tokens": 43952966.0,
      "step": 19700
    },
    {
      "entropy": 5.564641714096069,
      "epoch": 2.2130622788790926,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004517081153322448,
      "loss": 5.2755,
      "mean_token_accuracy": 0.19167562574148178,
      "num_tokens": 43964725.0,
      "step": 19705
    },
    {
      "entropy": 5.51898226737976,
      "epoch": 2.2136238557870502,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004516832593626085,
      "loss": 5.1585,
      "mean_token_accuracy": 0.2001033678650856,
      "num_tokens": 43976356.0,
      "step": 19710
    },
    {
      "entropy": 5.399066019058227,
      "epoch": 2.2141854326950074,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004516583977673087,
      "loss": 5.2039,
      "mean_token_accuracy": 0.1964070677757263,
      "num_tokens": 43987732.0,
      "step": 19715
    },
    {
      "entropy": 5.535548686981201,
      "epoch": 2.214747009602965,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045163353054713704,
      "loss": 5.3434,
      "mean_token_accuracy": 0.18696456402540207,
      "num_tokens": 43998701.0,
      "step": 19720
    },
    {
      "entropy": 5.634602832794189,
      "epoch": 2.2153085865109228,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004516086577028852,
      "loss": 5.2706,
      "mean_token_accuracy": 0.1942870184779167,
      "num_tokens": 44008978.0,
      "step": 19725
    },
    {
      "entropy": 5.499393081665039,
      "epoch": 2.21587016341888,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004515837792353453,
      "loss": 5.2563,
      "mean_token_accuracy": 0.19361717849969864,
      "num_tokens": 44018928.0,
      "step": 19730
    },
    {
      "entropy": 5.452484560012818,
      "epoch": 2.2164317403268377,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004515588951453093,
      "loss": 5.2428,
      "mean_token_accuracy": 0.1916105180978775,
      "num_tokens": 44030044.0,
      "step": 19735
    },
    {
      "entropy": 5.396527433395386,
      "epoch": 2.2169933172347953,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004515340054335696,
      "loss": 5.1809,
      "mean_token_accuracy": 0.19517827033996582,
      "num_tokens": 44041851.0,
      "step": 19740
    },
    {
      "entropy": 5.490589284896851,
      "epoch": 2.217554894142753,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004515091101009188,
      "loss": 5.1613,
      "mean_token_accuracy": 0.19913836121559142,
      "num_tokens": 44051871.0,
      "step": 19745
    },
    {
      "entropy": 5.524836778640747,
      "epoch": 2.21811647105071,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004514842091481493,
      "loss": 5.3633,
      "mean_token_accuracy": 0.18144014328718186,
      "num_tokens": 44063222.0,
      "step": 19750
    },
    {
      "entropy": 5.620134353637695,
      "epoch": 2.218678047958668,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004514593025760543,
      "loss": 5.3185,
      "mean_token_accuracy": 0.19031717330217363,
      "num_tokens": 44075020.0,
      "step": 19755
    },
    {
      "entropy": 5.52335376739502,
      "epoch": 2.2192396248666255,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045143439038542653,
      "loss": 5.2453,
      "mean_token_accuracy": 0.1905816301703453,
      "num_tokens": 44086696.0,
      "step": 19760
    },
    {
      "entropy": 5.5227399349212645,
      "epoch": 2.219801201774583,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004514094725770594,
      "loss": 5.2048,
      "mean_token_accuracy": 0.1949351578950882,
      "num_tokens": 44096768.0,
      "step": 19765
    },
    {
      "entropy": 5.469168615341187,
      "epoch": 2.2203627786825404,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045138454915174625,
      "loss": 5.2209,
      "mean_token_accuracy": 0.2001909077167511,
      "num_tokens": 44108008.0,
      "step": 19770
    },
    {
      "entropy": 5.481593942642212,
      "epoch": 2.220924355590498,
      "grad_norm": 1.8125,
      "learning_rate": 0.00045135962011028064,
      "loss": 5.179,
      "mean_token_accuracy": 0.1917398080229759,
      "num_tokens": 44119231.0,
      "step": 19775
    },
    {
      "entropy": 5.4186042785644535,
      "epoch": 2.2214859324984557,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004513346854534564,
      "loss": 5.1678,
      "mean_token_accuracy": 0.19441763311624527,
      "num_tokens": 44129264.0,
      "step": 19780
    },
    {
      "entropy": 5.50796217918396,
      "epoch": 2.2220475094064134,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004513097451820673,
      "loss": 5.3098,
      "mean_token_accuracy": 0.19406169652938843,
      "num_tokens": 44141234.0,
      "step": 19785
    },
    {
      "entropy": 5.545570611953735,
      "epoch": 2.2226090863143706,
      "grad_norm": 1.734375,
      "learning_rate": 0.00045128479929690756,
      "loss": 5.1761,
      "mean_token_accuracy": 0.20019723176956178,
      "num_tokens": 44151751.0,
      "step": 19790
    },
    {
      "entropy": 5.376328468322754,
      "epoch": 2.2231706632223283,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004512598477987714,
      "loss": 5.1131,
      "mean_token_accuracy": 0.19938686341047288,
      "num_tokens": 44162687.0,
      "step": 19795
    },
    {
      "entropy": 5.472272872924805,
      "epoch": 2.223732240130286,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045123489068845334,
      "loss": 5.1861,
      "mean_token_accuracy": 0.19867129176855086,
      "num_tokens": 44173758.0,
      "step": 19800
    },
    {
      "entropy": 5.488106441497803,
      "epoch": 2.2242938170382436,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045120992796674804,
      "loss": 5.2227,
      "mean_token_accuracy": 0.1970738038420677,
      "num_tokens": 44184176.0,
      "step": 19805
    },
    {
      "entropy": 5.5556807041168215,
      "epoch": 2.224855393946201,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004511849596344503,
      "loss": 5.2982,
      "mean_token_accuracy": 0.1883039206266403,
      "num_tokens": 44195155.0,
      "step": 19810
    },
    {
      "entropy": 5.601989316940307,
      "epoch": 2.2254169708541585,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045115998569235507,
      "loss": 5.3327,
      "mean_token_accuracy": 0.18982167840003966,
      "num_tokens": 44207139.0,
      "step": 19815
    },
    {
      "entropy": 5.55784125328064,
      "epoch": 2.225978547762116,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045113500614125756,
      "loss": 5.2924,
      "mean_token_accuracy": 0.19552985280752183,
      "num_tokens": 44218317.0,
      "step": 19820
    },
    {
      "entropy": 5.513452816009521,
      "epoch": 2.2265401246700733,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004511100209819532,
      "loss": 5.2196,
      "mean_token_accuracy": 0.20117057114839554,
      "num_tokens": 44228700.0,
      "step": 19825
    },
    {
      "entropy": 5.381744909286499,
      "epoch": 2.227101701578031,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045108503021523737,
      "loss": 5.1642,
      "mean_token_accuracy": 0.2060288295149803,
      "num_tokens": 44239991.0,
      "step": 19830
    },
    {
      "entropy": 5.341881990432739,
      "epoch": 2.2276632784859887,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045106003384190604,
      "loss": 5.0701,
      "mean_token_accuracy": 0.20321475267410277,
      "num_tokens": 44250224.0,
      "step": 19835
    },
    {
      "entropy": 5.462254285812378,
      "epoch": 2.2282248553939463,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045103503186275486,
      "loss": 5.2653,
      "mean_token_accuracy": 0.1943323627114296,
      "num_tokens": 44261403.0,
      "step": 19840
    },
    {
      "entropy": 5.443727731704712,
      "epoch": 2.2287864323019035,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045101002427857997,
      "loss": 5.1813,
      "mean_token_accuracy": 0.19533726274967195,
      "num_tokens": 44272935.0,
      "step": 19845
    },
    {
      "entropy": 5.4540385723114015,
      "epoch": 2.229348009209861,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045098501109017765,
      "loss": 5.1805,
      "mean_token_accuracy": 0.20066781789064408,
      "num_tokens": 44283458.0,
      "step": 19850
    },
    {
      "entropy": 5.448845720291137,
      "epoch": 2.229909586117819,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00045095999229834437,
      "loss": 5.194,
      "mean_token_accuracy": 0.19300126135349274,
      "num_tokens": 44294847.0,
      "step": 19855
    },
    {
      "entropy": 5.503479719161987,
      "epoch": 2.2304711630257765,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004509349679038767,
      "loss": 5.3141,
      "mean_token_accuracy": 0.1934979274868965,
      "num_tokens": 44306395.0,
      "step": 19860
    },
    {
      "entropy": 5.52719554901123,
      "epoch": 2.2310327399337337,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045090993790757137,
      "loss": 5.2291,
      "mean_token_accuracy": 0.19407544285058975,
      "num_tokens": 44317452.0,
      "step": 19865
    },
    {
      "entropy": 5.47685546875,
      "epoch": 2.2315943168416914,
      "grad_norm": 1.375,
      "learning_rate": 0.0004508849023102253,
      "loss": 5.2109,
      "mean_token_accuracy": 0.20183431953191758,
      "num_tokens": 44329862.0,
      "step": 19870
    },
    {
      "entropy": 5.507463693618774,
      "epoch": 2.232155893749649,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004508598611126359,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19494223594665527,
      "num_tokens": 44340334.0,
      "step": 19875
    },
    {
      "entropy": 5.570824003219604,
      "epoch": 2.2327174706576067,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004508348143156002,
      "loss": 5.3593,
      "mean_token_accuracy": 0.18573096096515657,
      "num_tokens": 44351810.0,
      "step": 19880
    },
    {
      "entropy": 5.6180867671966555,
      "epoch": 2.233279047565564,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045080976191991587,
      "loss": 5.4113,
      "mean_token_accuracy": 0.18101848810911178,
      "num_tokens": 44365143.0,
      "step": 19885
    },
    {
      "entropy": 5.504892826080322,
      "epoch": 2.2338406244735216,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004507847039263805,
      "loss": 5.29,
      "mean_token_accuracy": 0.19741635918617248,
      "num_tokens": 44375895.0,
      "step": 19890
    },
    {
      "entropy": 5.499074125289917,
      "epoch": 2.2344022013814793,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000450759640335792,
      "loss": 5.188,
      "mean_token_accuracy": 0.19528525769710542,
      "num_tokens": 44387378.0,
      "step": 19895
    },
    {
      "entropy": 5.442280435562134,
      "epoch": 2.234963778289437,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004507345711489484,
      "loss": 5.1585,
      "mean_token_accuracy": 0.20831980258226396,
      "num_tokens": 44398460.0,
      "step": 19900
    },
    {
      "entropy": 5.422170114517212,
      "epoch": 2.235525355197394,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045070949636664785,
      "loss": 5.1875,
      "mean_token_accuracy": 0.19626613408327104,
      "num_tokens": 44409885.0,
      "step": 19905
    },
    {
      "entropy": 5.440979242324829,
      "epoch": 2.236086932105352,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004506844159896888,
      "loss": 5.1271,
      "mean_token_accuracy": 0.20281644463539122,
      "num_tokens": 44420897.0,
      "step": 19910
    },
    {
      "entropy": 5.553892469406128,
      "epoch": 2.2366485090133095,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004506593300188698,
      "loss": 5.3214,
      "mean_token_accuracy": 0.188603737950325,
      "num_tokens": 44432691.0,
      "step": 19915
    },
    {
      "entropy": 5.532892465591431,
      "epoch": 2.2372100859212667,
      "grad_norm": 1.25,
      "learning_rate": 0.0004506342384549897,
      "loss": 5.2373,
      "mean_token_accuracy": 0.19359887689352034,
      "num_tokens": 44444731.0,
      "step": 19920
    },
    {
      "entropy": 5.584154844284058,
      "epoch": 2.2377716628292244,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004506091412988472,
      "loss": 5.2339,
      "mean_token_accuracy": 0.20170113146305085,
      "num_tokens": 44456959.0,
      "step": 19925
    },
    {
      "entropy": 5.5804445266723635,
      "epoch": 2.238333239737182,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045058403855124165,
      "loss": 5.2942,
      "mean_token_accuracy": 0.19498781263828277,
      "num_tokens": 44468970.0,
      "step": 19930
    },
    {
      "entropy": 5.452075576782226,
      "epoch": 2.2388948166451397,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004505589302129721,
      "loss": 5.1701,
      "mean_token_accuracy": 0.20246293842792512,
      "num_tokens": 44480485.0,
      "step": 19935
    },
    {
      "entropy": 5.51645679473877,
      "epoch": 2.239456393553097,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004505338162848382,
      "loss": 5.338,
      "mean_token_accuracy": 0.19500317871570588,
      "num_tokens": 44492890.0,
      "step": 19940
    },
    {
      "entropy": 5.5942668437957765,
      "epoch": 2.2400179704610546,
      "grad_norm": 1.25,
      "learning_rate": 0.0004505086967676395,
      "loss": 5.4195,
      "mean_token_accuracy": 0.18069773763418198,
      "num_tokens": 44504743.0,
      "step": 19945
    },
    {
      "entropy": 5.570518827438354,
      "epoch": 2.240579547369012,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004504835716621757,
      "loss": 5.2083,
      "mean_token_accuracy": 0.19530122280120848,
      "num_tokens": 44514998.0,
      "step": 19950
    },
    {
      "entropy": 5.438938236236572,
      "epoch": 2.24114112427697,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004504584409692471,
      "loss": 5.1848,
      "mean_token_accuracy": 0.20347164571285248,
      "num_tokens": 44527530.0,
      "step": 19955
    },
    {
      "entropy": 5.556153011322022,
      "epoch": 2.241702701184927,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045043330468965364,
      "loss": 5.3963,
      "mean_token_accuracy": 0.1878069669008255,
      "num_tokens": 44540939.0,
      "step": 19960
    },
    {
      "entropy": 5.589362144470215,
      "epoch": 2.2422642780928848,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045040816282419566,
      "loss": 5.2659,
      "mean_token_accuracy": 0.19184372276067735,
      "num_tokens": 44551557.0,
      "step": 19965
    },
    {
      "entropy": 5.439628791809082,
      "epoch": 2.2428258550008424,
      "grad_norm": 1.25,
      "learning_rate": 0.00045038301537367374,
      "loss": 5.1595,
      "mean_token_accuracy": 0.19913299679756163,
      "num_tokens": 44561969.0,
      "step": 19970
    },
    {
      "entropy": 5.372342681884765,
      "epoch": 2.2433874319088,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045035786233888864,
      "loss": 5.1541,
      "mean_token_accuracy": 0.19579712003469468,
      "num_tokens": 44573542.0,
      "step": 19975
    },
    {
      "entropy": 5.559537076950074,
      "epoch": 2.2439490088167573,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045033270372064123,
      "loss": 5.2257,
      "mean_token_accuracy": 0.204598106443882,
      "num_tokens": 44583720.0,
      "step": 19980
    },
    {
      "entropy": 5.485935354232788,
      "epoch": 2.244510585724715,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045030753951973243,
      "loss": 5.3023,
      "mean_token_accuracy": 0.1947477087378502,
      "num_tokens": 44595314.0,
      "step": 19985
    },
    {
      "entropy": 5.6282354354858395,
      "epoch": 2.2450721626326726,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045028236973696363,
      "loss": 5.4322,
      "mean_token_accuracy": 0.18839803487062454,
      "num_tokens": 44608569.0,
      "step": 19990
    },
    {
      "entropy": 5.493777847290039,
      "epoch": 2.2456337395406303,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004502571943731362,
      "loss": 5.2371,
      "mean_token_accuracy": 0.1939469173550606,
      "num_tokens": 44619389.0,
      "step": 19995
    },
    {
      "entropy": 5.514086818695068,
      "epoch": 2.2461953164485875,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004502320134290517,
      "loss": 5.32,
      "mean_token_accuracy": 0.19063089042901993,
      "num_tokens": 44629945.0,
      "step": 20000
    },
    {
      "entropy": 5.556463289260864,
      "epoch": 2.246756893356545,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045020682690551203,
      "loss": 5.3412,
      "mean_token_accuracy": 0.18668145984411239,
      "num_tokens": 44640741.0,
      "step": 20005
    },
    {
      "entropy": 5.524778699874878,
      "epoch": 2.247318470264503,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045018163480331887,
      "loss": 5.1792,
      "mean_token_accuracy": 0.2004580393433571,
      "num_tokens": 44651312.0,
      "step": 20010
    },
    {
      "entropy": 5.449014711380005,
      "epoch": 2.24788004717246,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045015643712327465,
      "loss": 5.2182,
      "mean_token_accuracy": 0.19613993614912034,
      "num_tokens": 44663156.0,
      "step": 20015
    },
    {
      "entropy": 5.4309642791748045,
      "epoch": 2.2484416240804177,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004501312338661815,
      "loss": 5.2239,
      "mean_token_accuracy": 0.20006002634763717,
      "num_tokens": 44673479.0,
      "step": 20020
    },
    {
      "entropy": 5.492065954208374,
      "epoch": 2.2490032009883754,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004501060250328419,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19600303024053572,
      "num_tokens": 44684207.0,
      "step": 20025
    },
    {
      "entropy": 5.522979259490967,
      "epoch": 2.249564777896333,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004500808106240586,
      "loss": 5.2219,
      "mean_token_accuracy": 0.19826527833938598,
      "num_tokens": 44695611.0,
      "step": 20030
    },
    {
      "entropy": 5.499130344390869,
      "epoch": 2.2501263548042907,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004500555906406343,
      "loss": 5.2581,
      "mean_token_accuracy": 0.19427683651447297,
      "num_tokens": 44707061.0,
      "step": 20035
    },
    {
      "entropy": 5.58782844543457,
      "epoch": 2.250687931712248,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004500303650833722,
      "loss": 5.3577,
      "mean_token_accuracy": 0.1854730948805809,
      "num_tokens": 44719366.0,
      "step": 20040
    },
    {
      "entropy": 5.556455802917481,
      "epoch": 2.2512495086202056,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004500051339530754,
      "loss": 5.2256,
      "mean_token_accuracy": 0.20091187506914138,
      "num_tokens": 44730578.0,
      "step": 20045
    },
    {
      "entropy": 5.457146549224854,
      "epoch": 2.2518110855281632,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004499798972505472,
      "loss": 5.1848,
      "mean_token_accuracy": 0.19817287474870682,
      "num_tokens": 44740620.0,
      "step": 20050
    },
    {
      "entropy": 5.484510374069214,
      "epoch": 2.2523726624361204,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044995465497659127,
      "loss": 5.2389,
      "mean_token_accuracy": 0.20124588906764984,
      "num_tokens": 44751518.0,
      "step": 20055
    },
    {
      "entropy": 5.510428333282471,
      "epoch": 2.252934239344078,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004499294071320112,
      "loss": 5.2686,
      "mean_token_accuracy": 0.19703642874956132,
      "num_tokens": 44761563.0,
      "step": 20060
    },
    {
      "entropy": 5.5353888988494875,
      "epoch": 2.2534958162520358,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000449904153717611,
      "loss": 5.2805,
      "mean_token_accuracy": 0.1933363363146782,
      "num_tokens": 44773330.0,
      "step": 20065
    },
    {
      "entropy": 5.53394832611084,
      "epoch": 2.2540573931599934,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004498788947341947,
      "loss": 5.2463,
      "mean_token_accuracy": 0.19813215285539626,
      "num_tokens": 44786198.0,
      "step": 20070
    },
    {
      "entropy": 5.387781095504761,
      "epoch": 2.2546189700679506,
      "grad_norm": 1.125,
      "learning_rate": 0.0004498536301825665,
      "loss": 5.1404,
      "mean_token_accuracy": 0.2007957324385643,
      "num_tokens": 44797643.0,
      "step": 20075
    },
    {
      "entropy": 5.407376337051391,
      "epoch": 2.2551805469759083,
      "grad_norm": 1.171875,
      "learning_rate": 0.000449828360063531,
      "loss": 5.168,
      "mean_token_accuracy": 0.2013155400753021,
      "num_tokens": 44808283.0,
      "step": 20080
    },
    {
      "entropy": 5.447269535064697,
      "epoch": 2.255742123883866,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044980308437789267,
      "loss": 5.0732,
      "mean_token_accuracy": 0.20297736674547195,
      "num_tokens": 44817699.0,
      "step": 20085
    },
    {
      "entropy": 5.486398029327392,
      "epoch": 2.2563037007918236,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004497778031264564,
      "loss": 5.2906,
      "mean_token_accuracy": 0.19047022014856338,
      "num_tokens": 44828428.0,
      "step": 20090
    },
    {
      "entropy": 5.40444598197937,
      "epoch": 2.256865277699781,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044975251631002694,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19405728727579116,
      "num_tokens": 44839288.0,
      "step": 20095
    },
    {
      "entropy": 5.518749046325683,
      "epoch": 2.2574268546077385,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004497272239294097,
      "loss": 5.4182,
      "mean_token_accuracy": 0.1873346209526062,
      "num_tokens": 44850250.0,
      "step": 20100
    },
    {
      "entropy": 5.575524854660034,
      "epoch": 2.257988431515696,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004497019259854099,
      "loss": 5.255,
      "mean_token_accuracy": 0.1953403204679489,
      "num_tokens": 44862186.0,
      "step": 20105
    },
    {
      "entropy": 5.54745135307312,
      "epoch": 2.2585500084236534,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044967662247883295,
      "loss": 5.3404,
      "mean_token_accuracy": 0.19116792529821397,
      "num_tokens": 44874266.0,
      "step": 20110
    },
    {
      "entropy": 5.553813028335571,
      "epoch": 2.259111585331611,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044965131341048454,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19943919479846955,
      "num_tokens": 44885783.0,
      "step": 20115
    },
    {
      "entropy": 5.480351400375366,
      "epoch": 2.2596731622395687,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004496259987811707,
      "loss": 5.1375,
      "mean_token_accuracy": 0.2034609317779541,
      "num_tokens": 44896972.0,
      "step": 20120
    },
    {
      "entropy": 5.466098499298096,
      "epoch": 2.2602347391475264,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004496006785916973,
      "loss": 5.246,
      "mean_token_accuracy": 0.19787174910306932,
      "num_tokens": 44907414.0,
      "step": 20125
    },
    {
      "entropy": 5.498031806945801,
      "epoch": 2.260796316055484,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044957535284287055,
      "loss": 5.1722,
      "mean_token_accuracy": 0.19434682428836822,
      "num_tokens": 44917150.0,
      "step": 20130
    },
    {
      "entropy": 5.469934034347534,
      "epoch": 2.2613578929634413,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044955002153549686,
      "loss": 5.211,
      "mean_token_accuracy": 0.19735645353794098,
      "num_tokens": 44928927.0,
      "step": 20135
    },
    {
      "entropy": 5.467767143249512,
      "epoch": 2.261919469871399,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004495246846703828,
      "loss": 5.2256,
      "mean_token_accuracy": 0.19621748328208924,
      "num_tokens": 44938987.0,
      "step": 20140
    },
    {
      "entropy": 5.498761415481567,
      "epoch": 2.2624810467793566,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044949934224833505,
      "loss": 5.1625,
      "mean_token_accuracy": 0.1975456029176712,
      "num_tokens": 44950395.0,
      "step": 20145
    },
    {
      "entropy": 5.505730009078979,
      "epoch": 2.263042623687314,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004494739942701606,
      "loss": 5.1744,
      "mean_token_accuracy": 0.20360083431005477,
      "num_tokens": 44961389.0,
      "step": 20150
    },
    {
      "entropy": 5.532986211776733,
      "epoch": 2.2636042005952715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044944864073666655,
      "loss": 5.3595,
      "mean_token_accuracy": 0.19310897290706636,
      "num_tokens": 44973142.0,
      "step": 20155
    },
    {
      "entropy": 5.401855993270874,
      "epoch": 2.264165777503229,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044942328164866,
      "loss": 5.1518,
      "mean_token_accuracy": 0.20104912370443345,
      "num_tokens": 44983831.0,
      "step": 20160
    },
    {
      "entropy": 5.500086736679077,
      "epoch": 2.2647273544111868,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004493979170069486,
      "loss": 5.2509,
      "mean_token_accuracy": 0.2013973131775856,
      "num_tokens": 44996386.0,
      "step": 20165
    },
    {
      "entropy": 5.498330402374267,
      "epoch": 2.265288931319144,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044937254681233984,
      "loss": 5.1929,
      "mean_token_accuracy": 0.19184104055166246,
      "num_tokens": 45007607.0,
      "step": 20170
    },
    {
      "entropy": 5.427362871170044,
      "epoch": 2.2658505082271017,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044934717106564164,
      "loss": 5.2193,
      "mean_token_accuracy": 0.19729479551315307,
      "num_tokens": 45019637.0,
      "step": 20175
    },
    {
      "entropy": 5.422962856292725,
      "epoch": 2.2664120851350593,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044932178976766183,
      "loss": 5.2172,
      "mean_token_accuracy": 0.19794851690530776,
      "num_tokens": 45031329.0,
      "step": 20180
    },
    {
      "entropy": 5.463313817977905,
      "epoch": 2.266973662043017,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044929640291920865,
      "loss": 5.2061,
      "mean_token_accuracy": 0.19512274116277695,
      "num_tokens": 45042539.0,
      "step": 20185
    },
    {
      "entropy": 5.391692638397217,
      "epoch": 2.267535238950974,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004492710105210904,
      "loss": 5.098,
      "mean_token_accuracy": 0.20303086936473846,
      "num_tokens": 45054021.0,
      "step": 20190
    },
    {
      "entropy": 5.570929288864136,
      "epoch": 2.268096815858932,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044924561257411554,
      "loss": 5.2995,
      "mean_token_accuracy": 0.19303088784217834,
      "num_tokens": 45065912.0,
      "step": 20195
    },
    {
      "entropy": 5.4894567966461185,
      "epoch": 2.2686583927668895,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004492202090790928,
      "loss": 5.1844,
      "mean_token_accuracy": 0.20377012491226196,
      "num_tokens": 45075965.0,
      "step": 20200
    },
    {
      "entropy": 5.381528711318969,
      "epoch": 2.2692199696748467,
      "grad_norm": 1.25,
      "learning_rate": 0.0004491948000368311,
      "loss": 5.1851,
      "mean_token_accuracy": 0.1979663237929344,
      "num_tokens": 45086676.0,
      "step": 20205
    },
    {
      "entropy": 5.445681858062744,
      "epoch": 2.2697815465828044,
      "grad_norm": 1.25,
      "learning_rate": 0.00044916938544813936,
      "loss": 5.2159,
      "mean_token_accuracy": 0.20002556443214417,
      "num_tokens": 45097747.0,
      "step": 20210
    },
    {
      "entropy": 5.54331488609314,
      "epoch": 2.270343123490762,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044914396531382686,
      "loss": 5.1724,
      "mean_token_accuracy": 0.20359463840723038,
      "num_tokens": 45107506.0,
      "step": 20215
    },
    {
      "entropy": 5.4871217727661135,
      "epoch": 2.2709047003987197,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044911853963470295,
      "loss": 5.2335,
      "mean_token_accuracy": 0.20210439264774321,
      "num_tokens": 45119939.0,
      "step": 20220
    },
    {
      "entropy": 5.48626184463501,
      "epoch": 2.2714662773066774,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004490931084115772,
      "loss": 5.3068,
      "mean_token_accuracy": 0.18659437596797943,
      "num_tokens": 45133749.0,
      "step": 20225
    },
    {
      "entropy": 5.516075468063354,
      "epoch": 2.2720278542146346,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044906767164525936,
      "loss": 5.2831,
      "mean_token_accuracy": 0.19292282611131667,
      "num_tokens": 45144696.0,
      "step": 20230
    },
    {
      "entropy": 5.484831094741821,
      "epoch": 2.2725894311225923,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004490422293365594,
      "loss": 5.1476,
      "mean_token_accuracy": 0.19563103020191192,
      "num_tokens": 45154921.0,
      "step": 20235
    },
    {
      "entropy": 5.409498834609986,
      "epoch": 2.27315100803055,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004490167814862872,
      "loss": 5.1423,
      "mean_token_accuracy": 0.20068973451852798,
      "num_tokens": 45164769.0,
      "step": 20240
    },
    {
      "entropy": 5.550851106643677,
      "epoch": 2.273712584938507,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004489913280952533,
      "loss": 5.3287,
      "mean_token_accuracy": 0.19520364105701446,
      "num_tokens": 45176619.0,
      "step": 20245
    },
    {
      "entropy": 5.460875177383423,
      "epoch": 2.274274161846465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004489658691642681,
      "loss": 5.2376,
      "mean_token_accuracy": 0.1926131084561348,
      "num_tokens": 45186508.0,
      "step": 20250
    },
    {
      "entropy": 5.578034257888794,
      "epoch": 2.2748357387544225,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044894040469414207,
      "loss": 5.3442,
      "mean_token_accuracy": 0.1901059105992317,
      "num_tokens": 45198213.0,
      "step": 20255
    },
    {
      "entropy": 5.486166191101074,
      "epoch": 2.27539731566238,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044891493468568607,
      "loss": 5.2244,
      "mean_token_accuracy": 0.18832046240568162,
      "num_tokens": 45209328.0,
      "step": 20260
    },
    {
      "entropy": 5.503150701522827,
      "epoch": 2.2759588925703373,
      "grad_norm": 1.25,
      "learning_rate": 0.00044888945913971106,
      "loss": 5.2118,
      "mean_token_accuracy": 0.19665393978357315,
      "num_tokens": 45219722.0,
      "step": 20265
    },
    {
      "entropy": 5.591525936126709,
      "epoch": 2.276520469478295,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044886397805702835,
      "loss": 5.2907,
      "mean_token_accuracy": 0.18618952333927155,
      "num_tokens": 45230377.0,
      "step": 20270
    },
    {
      "entropy": 5.579563331604004,
      "epoch": 2.2770820463862527,
      "grad_norm": 1.25,
      "learning_rate": 0.00044883849143844914,
      "loss": 5.2096,
      "mean_token_accuracy": 0.19831510931253432,
      "num_tokens": 45241598.0,
      "step": 20275
    },
    {
      "entropy": 5.5415668964385985,
      "epoch": 2.2776436232942103,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004488129992847848,
      "loss": 5.4183,
      "mean_token_accuracy": 0.18950254023075103,
      "num_tokens": 45253153.0,
      "step": 20280
    },
    {
      "entropy": 5.513988065719604,
      "epoch": 2.2782052002021675,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004487875015968473,
      "loss": 5.2859,
      "mean_token_accuracy": 0.1942953109741211,
      "num_tokens": 45264523.0,
      "step": 20285
    },
    {
      "entropy": 5.477509784698486,
      "epoch": 2.278766777110125,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004487619983754483,
      "loss": 5.2018,
      "mean_token_accuracy": 0.1925478085875511,
      "num_tokens": 45275320.0,
      "step": 20290
    },
    {
      "entropy": 5.522681856155396,
      "epoch": 2.279328354018083,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044873648962139983,
      "loss": 5.2232,
      "mean_token_accuracy": 0.19909849315881728,
      "num_tokens": 45286704.0,
      "step": 20295
    },
    {
      "entropy": 5.540631628036499,
      "epoch": 2.27988993092604,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004487109753355142,
      "loss": 5.259,
      "mean_token_accuracy": 0.19319239258766174,
      "num_tokens": 45299098.0,
      "step": 20300
    },
    {
      "entropy": 5.49213342666626,
      "epoch": 2.2804515078339977,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004486854555186037,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19423681497573853,
      "num_tokens": 45309995.0,
      "step": 20305
    },
    {
      "entropy": 5.580655097961426,
      "epoch": 2.2810130847419554,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004486599301714809,
      "loss": 5.377,
      "mean_token_accuracy": 0.19141939878463746,
      "num_tokens": 45321711.0,
      "step": 20310
    },
    {
      "entropy": 5.499113750457764,
      "epoch": 2.281574661649913,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004486343992949587,
      "loss": 5.2829,
      "mean_token_accuracy": 0.19541093707084656,
      "num_tokens": 45332401.0,
      "step": 20315
    },
    {
      "entropy": 5.573700761795044,
      "epoch": 2.2821362385578707,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044860886288984977,
      "loss": 5.3328,
      "mean_token_accuracy": 0.1919061630964279,
      "num_tokens": 45344138.0,
      "step": 20320
    },
    {
      "entropy": 5.442095041275024,
      "epoch": 2.282697815465828,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004485833209569674,
      "loss": 5.2385,
      "mean_token_accuracy": 0.19264374673366547,
      "num_tokens": 45355913.0,
      "step": 20325
    },
    {
      "entropy": 5.498832941055298,
      "epoch": 2.2832593923737856,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004485577734971247,
      "loss": 5.1837,
      "mean_token_accuracy": 0.1980019763112068,
      "num_tokens": 45367342.0,
      "step": 20330
    },
    {
      "entropy": 5.50741286277771,
      "epoch": 2.2838209692817433,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044853222051113525,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19548344910144805,
      "num_tokens": 45377866.0,
      "step": 20335
    },
    {
      "entropy": 5.472334480285644,
      "epoch": 2.2843825461897005,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044850666199981244,
      "loss": 5.2634,
      "mean_token_accuracy": 0.19072347432374953,
      "num_tokens": 45389346.0,
      "step": 20340
    },
    {
      "entropy": 5.437552738189697,
      "epoch": 2.284944123097658,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044848109796397035,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19399820417165756,
      "num_tokens": 45400112.0,
      "step": 20345
    },
    {
      "entropy": 5.4474702835083,
      "epoch": 2.285505700005616,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044845552840442276,
      "loss": 5.1763,
      "mean_token_accuracy": 0.19980261623859405,
      "num_tokens": 45411666.0,
      "step": 20350
    },
    {
      "entropy": 5.428030920028687,
      "epoch": 2.2860672769135735,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044842995332198393,
      "loss": 5.1573,
      "mean_token_accuracy": 0.20993824154138566,
      "num_tokens": 45423631.0,
      "step": 20355
    },
    {
      "entropy": 5.345173454284668,
      "epoch": 2.2866288538215307,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044840437271746807,
      "loss": 5.1262,
      "mean_token_accuracy": 0.198094242811203,
      "num_tokens": 45434697.0,
      "step": 20360
    },
    {
      "entropy": 5.534241247177124,
      "epoch": 2.2871904307294884,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004483787865916898,
      "loss": 5.2774,
      "mean_token_accuracy": 0.18845348060131073,
      "num_tokens": 45446637.0,
      "step": 20365
    },
    {
      "entropy": 5.472633218765258,
      "epoch": 2.287752007637446,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044835319494546366,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19916970729827882,
      "num_tokens": 45456750.0,
      "step": 20370
    },
    {
      "entropy": 5.430306768417358,
      "epoch": 2.2883135845454037,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00044832759777960456,
      "loss": 5.2257,
      "mean_token_accuracy": 0.19966684579849242,
      "num_tokens": 45469383.0,
      "step": 20375
    },
    {
      "entropy": 5.4830526351928714,
      "epoch": 2.288875161453361,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004483019950949276,
      "loss": 5.1962,
      "mean_token_accuracy": 0.2003369987010956,
      "num_tokens": 45479668.0,
      "step": 20380
    },
    {
      "entropy": 5.445210313796997,
      "epoch": 2.2894367383613186,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004482763868922478,
      "loss": 5.2717,
      "mean_token_accuracy": 0.19506213515996934,
      "num_tokens": 45490888.0,
      "step": 20385
    },
    {
      "entropy": 5.512267637252807,
      "epoch": 2.289998315269276,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004482507731723807,
      "loss": 5.2628,
      "mean_token_accuracy": 0.19120286852121354,
      "num_tokens": 45501973.0,
      "step": 20390
    },
    {
      "entropy": 5.498227834701538,
      "epoch": 2.2905598921772334,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004482251539361418,
      "loss": 5.2486,
      "mean_token_accuracy": 0.19746894091367723,
      "num_tokens": 45513171.0,
      "step": 20395
    },
    {
      "entropy": 5.418797159194947,
      "epoch": 2.291121469085191,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004481995291843467,
      "loss": 5.1761,
      "mean_token_accuracy": 0.19927647709846497,
      "num_tokens": 45523642.0,
      "step": 20400
    },
    {
      "entropy": 5.523365068435669,
      "epoch": 2.2916830459931488,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004481738989178115,
      "loss": 5.2533,
      "mean_token_accuracy": 0.19765454083681105,
      "num_tokens": 45534467.0,
      "step": 20405
    },
    {
      "entropy": 5.52879753112793,
      "epoch": 2.2922446229011064,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004481482631373522,
      "loss": 5.2405,
      "mean_token_accuracy": 0.18979344069957732,
      "num_tokens": 45545902.0,
      "step": 20410
    },
    {
      "entropy": 5.439551973342896,
      "epoch": 2.292806199809064,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044812262184378503,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20488597005605697,
      "num_tokens": 45556270.0,
      "step": 20415
    },
    {
      "entropy": 5.3579466342926025,
      "epoch": 2.2933677767170213,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044809697503792647,
      "loss": 5.175,
      "mean_token_accuracy": 0.20420384109020234,
      "num_tokens": 45566773.0,
      "step": 20420
    },
    {
      "entropy": 5.464023017883301,
      "epoch": 2.293929353624979,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044807132272059306,
      "loss": 5.2818,
      "mean_token_accuracy": 0.1917489230632782,
      "num_tokens": 45578819.0,
      "step": 20425
    },
    {
      "entropy": 5.527934646606445,
      "epoch": 2.2944909305329366,
      "grad_norm": 1.25,
      "learning_rate": 0.0004480456648926016,
      "loss": 5.233,
      "mean_token_accuracy": 0.19213209748268129,
      "num_tokens": 45589128.0,
      "step": 20430
    },
    {
      "entropy": 5.528298330307007,
      "epoch": 2.295052507440894,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004480200015547691,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1943219929933548,
      "num_tokens": 45600253.0,
      "step": 20435
    },
    {
      "entropy": 5.560419750213623,
      "epoch": 2.2956140843488515,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044799433270791255,
      "loss": 5.2608,
      "mean_token_accuracy": 0.19483513832092286,
      "num_tokens": 45611258.0,
      "step": 20440
    },
    {
      "entropy": 5.508506965637207,
      "epoch": 2.296175661256809,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004479686583528494,
      "loss": 5.2134,
      "mean_token_accuracy": 0.20238776803016661,
      "num_tokens": 45622720.0,
      "step": 20445
    },
    {
      "entropy": 5.422779130935669,
      "epoch": 2.296737238164767,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004479429784903971,
      "loss": 5.2001,
      "mean_token_accuracy": 0.19343502968549728,
      "num_tokens": 45634748.0,
      "step": 20450
    },
    {
      "entropy": 5.4938685417175295,
      "epoch": 2.297298815072724,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004479172931213733,
      "loss": 5.1436,
      "mean_token_accuracy": 0.19243053793907167,
      "num_tokens": 45645735.0,
      "step": 20455
    },
    {
      "entropy": 5.469387865066528,
      "epoch": 2.2978603919806817,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004478916022465958,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20273699313402177,
      "num_tokens": 45656309.0,
      "step": 20460
    },
    {
      "entropy": 5.423810958862305,
      "epoch": 2.2984219688886394,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004478659058668827,
      "loss": 5.2987,
      "mean_token_accuracy": 0.19772691130638123,
      "num_tokens": 45667925.0,
      "step": 20465
    },
    {
      "entropy": 5.490616941452027,
      "epoch": 2.298983545796597,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044784020398305204,
      "loss": 5.2781,
      "mean_token_accuracy": 0.19499016106128692,
      "num_tokens": 45679352.0,
      "step": 20470
    },
    {
      "entropy": 5.454271745681763,
      "epoch": 2.2995451227045542,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004478144965959222,
      "loss": 5.2407,
      "mean_token_accuracy": 0.19642347991466522,
      "num_tokens": 45689723.0,
      "step": 20475
    },
    {
      "entropy": 5.563471889495849,
      "epoch": 2.300106699612512,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044778878370631183,
      "loss": 5.3727,
      "mean_token_accuracy": 0.18927737176418305,
      "num_tokens": 45701270.0,
      "step": 20480
    },
    {
      "entropy": 5.473050928115844,
      "epoch": 2.3006682765204696,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044776306531503957,
      "loss": 5.1355,
      "mean_token_accuracy": 0.2100411131978035,
      "num_tokens": 45711409.0,
      "step": 20485
    },
    {
      "entropy": 5.419238758087158,
      "epoch": 2.301229853428427,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044773734142292425,
      "loss": 5.2152,
      "mean_token_accuracy": 0.19952069371938705,
      "num_tokens": 45722032.0,
      "step": 20490
    },
    {
      "entropy": 5.48419713973999,
      "epoch": 2.3017914303363844,
      "grad_norm": 1.25,
      "learning_rate": 0.0004477116120307851,
      "loss": 5.212,
      "mean_token_accuracy": 0.20220494419336318,
      "num_tokens": 45733211.0,
      "step": 20495
    },
    {
      "entropy": 5.574658441543579,
      "epoch": 2.302353007244342,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004476858771394412,
      "loss": 5.3393,
      "mean_token_accuracy": 0.19307047575712205,
      "num_tokens": 45744585.0,
      "step": 20500
    },
    {
      "entropy": 5.520409202575683,
      "epoch": 2.3029145841522998,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004476601367497119,
      "loss": 5.2771,
      "mean_token_accuracy": 0.19734428375959395,
      "num_tokens": 45756040.0,
      "step": 20505
    },
    {
      "entropy": 5.503814029693603,
      "epoch": 2.3034761610602574,
      "grad_norm": 1.1875,
      "learning_rate": 0.000447634390862417,
      "loss": 5.2667,
      "mean_token_accuracy": 0.19268766641616822,
      "num_tokens": 45767717.0,
      "step": 20510
    },
    {
      "entropy": 5.559252071380615,
      "epoch": 2.3040377379682146,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044760863947837607,
      "loss": 5.2832,
      "mean_token_accuracy": 0.1924678087234497,
      "num_tokens": 45778326.0,
      "step": 20515
    },
    {
      "entropy": 5.476704788208008,
      "epoch": 2.3045993148761723,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044758288259840904,
      "loss": 5.1972,
      "mean_token_accuracy": 0.20741473585367204,
      "num_tokens": 45789780.0,
      "step": 20520
    },
    {
      "entropy": 5.437363147735596,
      "epoch": 2.30516089178413,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004475571202233363,
      "loss": 5.3431,
      "mean_token_accuracy": 0.18793452829122542,
      "num_tokens": 45801686.0,
      "step": 20525
    },
    {
      "entropy": 5.4736144065856935,
      "epoch": 2.305722468692087,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004475313523539778,
      "loss": 5.1862,
      "mean_token_accuracy": 0.1942782148718834,
      "num_tokens": 45812123.0,
      "step": 20530
    },
    {
      "entropy": 5.545000171661377,
      "epoch": 2.306284045600045,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044750557899115407,
      "loss": 5.27,
      "mean_token_accuracy": 0.1891551986336708,
      "num_tokens": 45822863.0,
      "step": 20535
    },
    {
      "entropy": 5.529285907745361,
      "epoch": 2.3068456225080025,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004474798001356859,
      "loss": 5.2805,
      "mean_token_accuracy": 0.1944286823272705,
      "num_tokens": 45833585.0,
      "step": 20540
    },
    {
      "entropy": 5.366938209533691,
      "epoch": 2.30740719941596,
      "grad_norm": 1.4375,
      "learning_rate": 0.000447454015788394,
      "loss": 5.0982,
      "mean_token_accuracy": 0.20143962204456328,
      "num_tokens": 45844745.0,
      "step": 20545
    },
    {
      "entropy": 5.422857475280762,
      "epoch": 2.3079687763239174,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004474282259500992,
      "loss": 5.185,
      "mean_token_accuracy": 0.20190047919750215,
      "num_tokens": 45856018.0,
      "step": 20550
    },
    {
      "entropy": 5.541861200332642,
      "epoch": 2.308530353231875,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000447402430621623,
      "loss": 5.2822,
      "mean_token_accuracy": 0.19415438324213027,
      "num_tokens": 45867182.0,
      "step": 20555
    },
    {
      "entropy": 5.525685834884643,
      "epoch": 2.3090919301398327,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004473766298037865,
      "loss": 5.3643,
      "mean_token_accuracy": 0.18285060822963714,
      "num_tokens": 45880116.0,
      "step": 20560
    },
    {
      "entropy": 5.571984338760376,
      "epoch": 2.3096535070477904,
      "grad_norm": 1.40625,
      "learning_rate": 0.00044735082349741116,
      "loss": 5.3002,
      "mean_token_accuracy": 0.19150526374578475,
      "num_tokens": 45891143.0,
      "step": 20565
    },
    {
      "entropy": 5.514363384246826,
      "epoch": 2.3102150839557476,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044732501170331886,
      "loss": 5.1928,
      "mean_token_accuracy": 0.19794349521398544,
      "num_tokens": 45901443.0,
      "step": 20570
    },
    {
      "entropy": 5.568791723251342,
      "epoch": 2.3107766608637053,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004472991944223313,
      "loss": 5.2944,
      "mean_token_accuracy": 0.18966622203588485,
      "num_tokens": 45913313.0,
      "step": 20575
    },
    {
      "entropy": 5.533223104476929,
      "epoch": 2.311338237771663,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044727337165527056,
      "loss": 5.3339,
      "mean_token_accuracy": 0.18738346099853515,
      "num_tokens": 45925248.0,
      "step": 20580
    },
    {
      "entropy": 5.617816352844239,
      "epoch": 2.31189981467962,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044724754340295887,
      "loss": 5.3318,
      "mean_token_accuracy": 0.19027050882577895,
      "num_tokens": 45936366.0,
      "step": 20585
    },
    {
      "entropy": 5.450673580169678,
      "epoch": 2.312461391587578,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004472217096662186,
      "loss": 5.1696,
      "mean_token_accuracy": 0.20805188417434692,
      "num_tokens": 45947732.0,
      "step": 20590
    },
    {
      "entropy": 5.46936731338501,
      "epoch": 2.3130229684955355,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004471958704458723,
      "loss": 5.2308,
      "mean_token_accuracy": 0.19785091131925583,
      "num_tokens": 45958655.0,
      "step": 20595
    },
    {
      "entropy": 5.523629999160766,
      "epoch": 2.313584545403493,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044717002574274267,
      "loss": 5.3072,
      "mean_token_accuracy": 0.19829923659563065,
      "num_tokens": 45970767.0,
      "step": 20600
    },
    {
      "entropy": 5.422164011001587,
      "epoch": 2.314146122311451,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004471441755576527,
      "loss": 5.1213,
      "mean_token_accuracy": 0.20740276277065278,
      "num_tokens": 45981063.0,
      "step": 20605
    },
    {
      "entropy": 5.392886590957642,
      "epoch": 2.314707699219408,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044711831989142534,
      "loss": 5.0888,
      "mean_token_accuracy": 0.20422577857971191,
      "num_tokens": 45991963.0,
      "step": 20610
    },
    {
      "entropy": 5.4641026020050045,
      "epoch": 2.3152692761273657,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044709245874488405,
      "loss": 5.3477,
      "mean_token_accuracy": 0.18637771755456925,
      "num_tokens": 46003460.0,
      "step": 20615
    },
    {
      "entropy": 5.521129322052002,
      "epoch": 2.3158308530353233,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004470665921188519,
      "loss": 5.2284,
      "mean_token_accuracy": 0.19587883800268174,
      "num_tokens": 46014338.0,
      "step": 20620
    },
    {
      "entropy": 5.4204026222229,
      "epoch": 2.3163924299432805,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044704072001415285,
      "loss": 5.1001,
      "mean_token_accuracy": 0.20640003085136413,
      "num_tokens": 46025282.0,
      "step": 20625
    },
    {
      "entropy": 5.452847051620483,
      "epoch": 2.316954006851238,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044701484243161055,
      "loss": 5.2719,
      "mean_token_accuracy": 0.1958751544356346,
      "num_tokens": 46035303.0,
      "step": 20630
    },
    {
      "entropy": 5.46298303604126,
      "epoch": 2.317515583759196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000446988959372049,
      "loss": 5.2433,
      "mean_token_accuracy": 0.2012130707502365,
      "num_tokens": 46046336.0,
      "step": 20635
    },
    {
      "entropy": 5.400800323486328,
      "epoch": 2.3180771606671535,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004469630708362922,
      "loss": 5.1206,
      "mean_token_accuracy": 0.19708970487117766,
      "num_tokens": 46056862.0,
      "step": 20640
    },
    {
      "entropy": 5.591737842559814,
      "epoch": 2.3186387375751107,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004469371768251646,
      "loss": 5.2662,
      "mean_token_accuracy": 0.19652716517448426,
      "num_tokens": 46067785.0,
      "step": 20645
    },
    {
      "entropy": 5.46607027053833,
      "epoch": 2.3192003144830684,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004469112773394906,
      "loss": 5.2525,
      "mean_token_accuracy": 0.18587084263563156,
      "num_tokens": 46080553.0,
      "step": 20650
    },
    {
      "entropy": 5.50519323348999,
      "epoch": 2.319761891391026,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004468853723800948,
      "loss": 5.2042,
      "mean_token_accuracy": 0.19355346262454987,
      "num_tokens": 46090985.0,
      "step": 20655
    },
    {
      "entropy": 5.560759830474853,
      "epoch": 2.3203234682989837,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004468594619478021,
      "loss": 5.2913,
      "mean_token_accuracy": 0.1927972510457039,
      "num_tokens": 46101990.0,
      "step": 20660
    },
    {
      "entropy": 5.394636392593384,
      "epoch": 2.320885045206941,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004468335460434375,
      "loss": 5.1439,
      "mean_token_accuracy": 0.203435218334198,
      "num_tokens": 46112853.0,
      "step": 20665
    },
    {
      "entropy": 5.463116931915283,
      "epoch": 2.3214466221148986,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004468076246678261,
      "loss": 5.2555,
      "mean_token_accuracy": 0.19200120866298676,
      "num_tokens": 46123508.0,
      "step": 20670
    },
    {
      "entropy": 5.627512788772583,
      "epoch": 2.3220081990228563,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044678169782179334,
      "loss": 5.3533,
      "mean_token_accuracy": 0.19093358218669892,
      "num_tokens": 46135394.0,
      "step": 20675
    },
    {
      "entropy": 5.581290245056152,
      "epoch": 2.3225697759308135,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044675576550616466,
      "loss": 5.3046,
      "mean_token_accuracy": 0.1954210266470909,
      "num_tokens": 46146838.0,
      "step": 20680
    },
    {
      "entropy": 5.542792224884034,
      "epoch": 2.323131352838771,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044672982772176584,
      "loss": 5.1875,
      "mean_token_accuracy": 0.20086237639188767,
      "num_tokens": 46157116.0,
      "step": 20685
    },
    {
      "entropy": 5.413563823699951,
      "epoch": 2.323692929746729,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004467038844694227,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19913049936294555,
      "num_tokens": 46168405.0,
      "step": 20690
    },
    {
      "entropy": 5.4855122566223145,
      "epoch": 2.3242545066546865,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004466779357499612,
      "loss": 5.2279,
      "mean_token_accuracy": 0.19788960069417955,
      "num_tokens": 46179907.0,
      "step": 20695
    },
    {
      "entropy": 5.473715448379517,
      "epoch": 2.324816083562644,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044665198156420774,
      "loss": 5.1498,
      "mean_token_accuracy": 0.1955687463283539,
      "num_tokens": 46190750.0,
      "step": 20700
    },
    {
      "entropy": 5.458232975006103,
      "epoch": 2.3253776604706013,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044662602191298856,
      "loss": 5.1499,
      "mean_token_accuracy": 0.20274998545646666,
      "num_tokens": 46201364.0,
      "step": 20705
    },
    {
      "entropy": 5.477144384384156,
      "epoch": 2.325939237378559,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044660005679713027,
      "loss": 5.1724,
      "mean_token_accuracy": 0.19556164145469665,
      "num_tokens": 46212341.0,
      "step": 20710
    },
    {
      "entropy": 5.454949378967285,
      "epoch": 2.3265008142865167,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044657408621745964,
      "loss": 5.2605,
      "mean_token_accuracy": 0.19336785972118378,
      "num_tokens": 46223286.0,
      "step": 20715
    },
    {
      "entropy": 5.395768880844116,
      "epoch": 2.327062391194474,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004465481101748036,
      "loss": 5.2775,
      "mean_token_accuracy": 0.19591585844755172,
      "num_tokens": 46234787.0,
      "step": 20720
    },
    {
      "entropy": 5.513534641265869,
      "epoch": 2.3276239681024316,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044652212866998914,
      "loss": 5.1267,
      "mean_token_accuracy": 0.20190000236034394,
      "num_tokens": 46244845.0,
      "step": 20725
    },
    {
      "entropy": 5.5209887504577635,
      "epoch": 2.328185545010389,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044649614170384354,
      "loss": 5.2636,
      "mean_token_accuracy": 0.1971845492720604,
      "num_tokens": 46256352.0,
      "step": 20730
    },
    {
      "entropy": 5.631619215011597,
      "epoch": 2.328747121918347,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044647014927719436,
      "loss": 5.4191,
      "mean_token_accuracy": 0.18905749171972275,
      "num_tokens": 46268535.0,
      "step": 20735
    },
    {
      "entropy": 5.476808261871338,
      "epoch": 2.329308698826304,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044644415139086913,
      "loss": 5.213,
      "mean_token_accuracy": 0.20166843831539155,
      "num_tokens": 46279164.0,
      "step": 20740
    },
    {
      "entropy": 5.431314468383789,
      "epoch": 2.3298702757342618,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004464181480456955,
      "loss": 5.0825,
      "mean_token_accuracy": 0.2080667272210121,
      "num_tokens": 46289379.0,
      "step": 20745
    },
    {
      "entropy": 5.424643087387085,
      "epoch": 2.3304318526422194,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004463921392425017,
      "loss": 5.1219,
      "mean_token_accuracy": 0.2021471932530403,
      "num_tokens": 46301023.0,
      "step": 20750
    },
    {
      "entropy": 5.475365257263183,
      "epoch": 2.330993429550177,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044636612498211566,
      "loss": 5.2373,
      "mean_token_accuracy": 0.2012315347790718,
      "num_tokens": 46312043.0,
      "step": 20755
    },
    {
      "entropy": 5.406970405578614,
      "epoch": 2.3315550064581343,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044634010526536575,
      "loss": 5.1932,
      "mean_token_accuracy": 0.19140066504478453,
      "num_tokens": 46322808.0,
      "step": 20760
    },
    {
      "entropy": 5.508573865890503,
      "epoch": 2.332116583366092,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004463140800930804,
      "loss": 5.2605,
      "mean_token_accuracy": 0.19064713418483734,
      "num_tokens": 46334260.0,
      "step": 20765
    },
    {
      "entropy": 5.501081562042236,
      "epoch": 2.3326781602740496,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004462880494660883,
      "loss": 5.2287,
      "mean_token_accuracy": 0.1961723119020462,
      "num_tokens": 46344245.0,
      "step": 20770
    },
    {
      "entropy": 5.452253627777099,
      "epoch": 2.333239737182007,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044626201338521834,
      "loss": 5.1576,
      "mean_token_accuracy": 0.1990344777703285,
      "num_tokens": 46356105.0,
      "step": 20775
    },
    {
      "entropy": 5.375404214859008,
      "epoch": 2.3338013140899645,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004462359718512993,
      "loss": 5.1464,
      "mean_token_accuracy": 0.20020316839218139,
      "num_tokens": 46367591.0,
      "step": 20780
    },
    {
      "entropy": 5.444856309890747,
      "epoch": 2.334362890997922,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044620992486516073,
      "loss": 5.1699,
      "mean_token_accuracy": 0.19651600420475007,
      "num_tokens": 46379022.0,
      "step": 20785
    },
    {
      "entropy": 5.448221921920776,
      "epoch": 2.33492446790588,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004461838724276316,
      "loss": 5.1906,
      "mean_token_accuracy": 0.20681209713220597,
      "num_tokens": 46390988.0,
      "step": 20790
    },
    {
      "entropy": 5.493107652664184,
      "epoch": 2.3354860448138375,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044615781453954164,
      "loss": 5.2515,
      "mean_token_accuracy": 0.1984211578965187,
      "num_tokens": 46401576.0,
      "step": 20795
    },
    {
      "entropy": 5.46716947555542,
      "epoch": 2.3360476217217947,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044613175120172035,
      "loss": 5.1302,
      "mean_token_accuracy": 0.2037460535764694,
      "num_tokens": 46413114.0,
      "step": 20800
    },
    {
      "entropy": 5.38899416923523,
      "epoch": 2.3366091986297524,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004461056824149979,
      "loss": 5.1383,
      "mean_token_accuracy": 0.20543973594903947,
      "num_tokens": 46424000.0,
      "step": 20805
    },
    {
      "entropy": 5.492922019958496,
      "epoch": 2.33717077553771,
      "grad_norm": 1.453125,
      "learning_rate": 0.000446079608180204,
      "loss": 5.1985,
      "mean_token_accuracy": 0.19603900611400604,
      "num_tokens": 46435392.0,
      "step": 20810
    },
    {
      "entropy": 5.388095331192017,
      "epoch": 2.3377323524456672,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004460535284981692,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19755102396011354,
      "num_tokens": 46445876.0,
      "step": 20815
    },
    {
      "entropy": 5.492770433425903,
      "epoch": 2.338293929353625,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004460274433697235,
      "loss": 5.1972,
      "mean_token_accuracy": 0.19560093581676483,
      "num_tokens": 46457175.0,
      "step": 20820
    },
    {
      "entropy": 5.529646205902099,
      "epoch": 2.3388555062615826,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044600135279569784,
      "loss": 5.3371,
      "mean_token_accuracy": 0.19093599766492844,
      "num_tokens": 46469100.0,
      "step": 20825
    },
    {
      "entropy": 5.451491498947144,
      "epoch": 2.3394170831695402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044597525677692273,
      "loss": 5.1585,
      "mean_token_accuracy": 0.19938063323497773,
      "num_tokens": 46481221.0,
      "step": 20830
    },
    {
      "entropy": 5.414064264297485,
      "epoch": 2.3399786600774974,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004459491553142291,
      "loss": 5.1088,
      "mean_token_accuracy": 0.2074011042714119,
      "num_tokens": 46492127.0,
      "step": 20835
    },
    {
      "entropy": 5.471398162841797,
      "epoch": 2.340540236985455,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044592304840844813,
      "loss": 5.2045,
      "mean_token_accuracy": 0.19230275005102157,
      "num_tokens": 46502880.0,
      "step": 20840
    },
    {
      "entropy": 5.4528186321258545,
      "epoch": 2.3411018138934128,
      "grad_norm": 1.21875,
      "learning_rate": 0.000445896936060411,
      "loss": 5.1434,
      "mean_token_accuracy": 0.20249173939228057,
      "num_tokens": 46513483.0,
      "step": 20845
    },
    {
      "entropy": 5.447299814224243,
      "epoch": 2.3416633908013704,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044587081827094906,
      "loss": 5.2349,
      "mean_token_accuracy": 0.19579770565032958,
      "num_tokens": 46526110.0,
      "step": 20850
    },
    {
      "entropy": 5.451260757446289,
      "epoch": 2.3422249677093276,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044584469504089407,
      "loss": 5.2106,
      "mean_token_accuracy": 0.19846130758523942,
      "num_tokens": 46537832.0,
      "step": 20855
    },
    {
      "entropy": 5.39276089668274,
      "epoch": 2.3427865446172853,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004458185663710777,
      "loss": 5.1322,
      "mean_token_accuracy": 0.19853869676589966,
      "num_tokens": 46548553.0,
      "step": 20860
    },
    {
      "entropy": 5.538682842254639,
      "epoch": 2.343348121525243,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004457924322623319,
      "loss": 5.3082,
      "mean_token_accuracy": 0.19690529853105546,
      "num_tokens": 46561095.0,
      "step": 20865
    },
    {
      "entropy": 5.572622776031494,
      "epoch": 2.3439096984332,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044576629271548884,
      "loss": 5.3001,
      "mean_token_accuracy": 0.19458558559417724,
      "num_tokens": 46572070.0,
      "step": 20870
    },
    {
      "entropy": 5.58575325012207,
      "epoch": 2.344471275341158,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004457401477313808,
      "loss": 5.324,
      "mean_token_accuracy": 0.19038349390029907,
      "num_tokens": 46583355.0,
      "step": 20875
    },
    {
      "entropy": 5.484343814849853,
      "epoch": 2.3450328522491155,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004457139973108403,
      "loss": 5.2034,
      "mean_token_accuracy": 0.1940722569823265,
      "num_tokens": 46594909.0,
      "step": 20880
    },
    {
      "entropy": 5.560077762603759,
      "epoch": 2.345594429157073,
      "grad_norm": 1.125,
      "learning_rate": 0.0004456878414546998,
      "loss": 5.2618,
      "mean_token_accuracy": 0.19874756783246994,
      "num_tokens": 46606245.0,
      "step": 20885
    },
    {
      "entropy": 5.407064533233642,
      "epoch": 2.346156006065031,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044566168016379234,
      "loss": 5.1625,
      "mean_token_accuracy": 0.19909542053937912,
      "num_tokens": 46617761.0,
      "step": 20890
    },
    {
      "entropy": 5.4622338771820065,
      "epoch": 2.346717582972988,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004456355134389507,
      "loss": 5.1911,
      "mean_token_accuracy": 0.19062627702951432,
      "num_tokens": 46628947.0,
      "step": 20895
    },
    {
      "entropy": 5.466204071044922,
      "epoch": 2.3472791598809457,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044560934128100814,
      "loss": 5.2053,
      "mean_token_accuracy": 0.1994910404086113,
      "num_tokens": 46640243.0,
      "step": 20900
    },
    {
      "entropy": 5.38988094329834,
      "epoch": 2.3478407367889034,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044558316369079805,
      "loss": 5.1334,
      "mean_token_accuracy": 0.19735938757658006,
      "num_tokens": 46650328.0,
      "step": 20905
    },
    {
      "entropy": 5.369638967514038,
      "epoch": 2.3484023136968606,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044555698066915385,
      "loss": 5.1271,
      "mean_token_accuracy": 0.19721096605062485,
      "num_tokens": 46661770.0,
      "step": 20910
    },
    {
      "entropy": 5.419118976593017,
      "epoch": 2.3489638906048182,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004455307922169092,
      "loss": 5.0799,
      "mean_token_accuracy": 0.2022887259721756,
      "num_tokens": 46673075.0,
      "step": 20915
    },
    {
      "entropy": 5.484280824661255,
      "epoch": 2.349525467512776,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004455045983348979,
      "loss": 5.2202,
      "mean_token_accuracy": 0.1938351273536682,
      "num_tokens": 46684040.0,
      "step": 20920
    },
    {
      "entropy": 5.423160791397095,
      "epoch": 2.3500870444207336,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004454783990239542,
      "loss": 5.1123,
      "mean_token_accuracy": 0.2010540097951889,
      "num_tokens": 46696295.0,
      "step": 20925
    },
    {
      "entropy": 5.433575105667114,
      "epoch": 2.350648621328691,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044545219428491215,
      "loss": 5.1991,
      "mean_token_accuracy": 0.200055393576622,
      "num_tokens": 46706399.0,
      "step": 20930
    },
    {
      "entropy": 5.505274534225464,
      "epoch": 2.3512101982366485,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044542598411860605,
      "loss": 5.3113,
      "mean_token_accuracy": 0.1915813311934471,
      "num_tokens": 46717644.0,
      "step": 20935
    },
    {
      "entropy": 5.613769245147705,
      "epoch": 2.351771775144606,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004453997685258706,
      "loss": 5.2798,
      "mean_token_accuracy": 0.19733862429857255,
      "num_tokens": 46728556.0,
      "step": 20940
    },
    {
      "entropy": 5.487457799911499,
      "epoch": 2.3523333520525638,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044537354750754033,
      "loss": 5.19,
      "mean_token_accuracy": 0.20072409212589265,
      "num_tokens": 46739730.0,
      "step": 20945
    },
    {
      "entropy": 5.520910835266113,
      "epoch": 2.352894928960521,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004453473210644503,
      "loss": 5.286,
      "mean_token_accuracy": 0.19550357908010482,
      "num_tokens": 46752349.0,
      "step": 20950
    },
    {
      "entropy": 5.52295446395874,
      "epoch": 2.3534565058684787,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004453210891974355,
      "loss": 5.2533,
      "mean_token_accuracy": 0.19241360276937486,
      "num_tokens": 46763771.0,
      "step": 20955
    },
    {
      "entropy": 5.5087450504302975,
      "epoch": 2.3540180827764363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044529485190733113,
      "loss": 5.1847,
      "mean_token_accuracy": 0.19992031306028366,
      "num_tokens": 46773371.0,
      "step": 20960
    },
    {
      "entropy": 5.435989713668823,
      "epoch": 2.3545796596843935,
      "grad_norm": 1.125,
      "learning_rate": 0.0004452686091949727,
      "loss": 5.1631,
      "mean_token_accuracy": 0.20077840387821197,
      "num_tokens": 46784244.0,
      "step": 20965
    },
    {
      "entropy": 5.4940653324127195,
      "epoch": 2.355141236592351,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044524236106119563,
      "loss": 5.2489,
      "mean_token_accuracy": 0.1923815131187439,
      "num_tokens": 46795373.0,
      "step": 20970
    },
    {
      "entropy": 5.559467124938965,
      "epoch": 2.355702813500309,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044521610750683586,
      "loss": 5.2519,
      "mean_token_accuracy": 0.1975223660469055,
      "num_tokens": 46806342.0,
      "step": 20975
    },
    {
      "entropy": 5.41952109336853,
      "epoch": 2.3562643904082665,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004451898485327291,
      "loss": 5.1612,
      "mean_token_accuracy": 0.20110985189676284,
      "num_tokens": 46817576.0,
      "step": 20980
    },
    {
      "entropy": 5.420933961868286,
      "epoch": 2.356825967316224,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004451635841397116,
      "loss": 5.2282,
      "mean_token_accuracy": 0.19303133636713027,
      "num_tokens": 46829335.0,
      "step": 20985
    },
    {
      "entropy": 5.519724988937378,
      "epoch": 2.3573875442241814,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004451373143286195,
      "loss": 5.2745,
      "mean_token_accuracy": 0.19118888974189757,
      "num_tokens": 46841334.0,
      "step": 20990
    },
    {
      "entropy": 5.57857232093811,
      "epoch": 2.357949121132139,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044511103910028945,
      "loss": 5.4278,
      "mean_token_accuracy": 0.19080803245306016,
      "num_tokens": 46852418.0,
      "step": 20995
    },
    {
      "entropy": 5.577915334701538,
      "epoch": 2.3585106980400967,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004450847584555579,
      "loss": 5.2815,
      "mean_token_accuracy": 0.19626296162605286,
      "num_tokens": 46863743.0,
      "step": 21000
    },
    {
      "epoch": 2.3585106980400967,
      "eval_entropy": 5.332942704118833,
      "eval_loss": 5.344221591949463,
      "eval_mean_token_accuracy": 0.19913381254139528,
      "eval_num_tokens": 46863743.0,
      "eval_runtime": 23.697,
      "eval_samples_per_second": 1308.728,
      "eval_steps_per_second": 163.607,
      "step": 21000
    },
    {
      "entropy": 5.425669097900391,
      "epoch": 2.359072274948054,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004450584723952617,
      "loss": 5.1023,
      "mean_token_accuracy": 0.19782320261001587,
      "num_tokens": 46873628.0,
      "step": 21005
    },
    {
      "entropy": 5.528709173202515,
      "epoch": 2.3596338518560116,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004450321809202377,
      "loss": 5.2874,
      "mean_token_accuracy": 0.18882786184549333,
      "num_tokens": 46885288.0,
      "step": 21010
    },
    {
      "entropy": 5.531344079971314,
      "epoch": 2.3601954287639693,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004450058840313232,
      "loss": 5.2575,
      "mean_token_accuracy": 0.19361186623573304,
      "num_tokens": 46897027.0,
      "step": 21015
    },
    {
      "entropy": 5.409613800048828,
      "epoch": 2.360757005671927,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044497958172935536,
      "loss": 5.1471,
      "mean_token_accuracy": 0.19863545894622803,
      "num_tokens": 46907947.0,
      "step": 21020
    },
    {
      "entropy": 5.421298170089722,
      "epoch": 2.361318582579884,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004449532740151717,
      "loss": 5.1952,
      "mean_token_accuracy": 0.20064647346735,
      "num_tokens": 46918334.0,
      "step": 21025
    },
    {
      "entropy": 5.468659114837647,
      "epoch": 2.361880159487842,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044492696088960985,
      "loss": 5.2638,
      "mean_token_accuracy": 0.20052294433116913,
      "num_tokens": 46929710.0,
      "step": 21030
    },
    {
      "entropy": 5.478493356704712,
      "epoch": 2.3624417363957995,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004449006423535078,
      "loss": 5.2107,
      "mean_token_accuracy": 0.1907900795340538,
      "num_tokens": 46941436.0,
      "step": 21035
    },
    {
      "entropy": 5.561452436447143,
      "epoch": 2.363003313303757,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044487431840770325,
      "loss": 5.2492,
      "mean_token_accuracy": 0.19540422409772873,
      "num_tokens": 46953673.0,
      "step": 21040
    },
    {
      "entropy": 5.4241691589355465,
      "epoch": 2.3635648902117143,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004448479890530347,
      "loss": 5.1262,
      "mean_token_accuracy": 0.20031379014253617,
      "num_tokens": 46964412.0,
      "step": 21045
    },
    {
      "entropy": 5.394551086425781,
      "epoch": 2.364126467119672,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044482165429034014,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19813333153724672,
      "num_tokens": 46975159.0,
      "step": 21050
    },
    {
      "entropy": 5.475469446182251,
      "epoch": 2.3646880440276297,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004447953141204583,
      "loss": 5.1862,
      "mean_token_accuracy": 0.2049855351448059,
      "num_tokens": 46987165.0,
      "step": 21055
    },
    {
      "entropy": 5.560744094848633,
      "epoch": 2.365249620935587,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044476896854422785,
      "loss": 5.2877,
      "mean_token_accuracy": 0.18687691390514374,
      "num_tokens": 46999085.0,
      "step": 21060
    },
    {
      "entropy": 5.496333026885987,
      "epoch": 2.3658111978435445,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044474261756248745,
      "loss": 5.2659,
      "mean_token_accuracy": 0.19189661294221877,
      "num_tokens": 47010970.0,
      "step": 21065
    },
    {
      "entropy": 5.507092428207398,
      "epoch": 2.366372774751502,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004447162611760764,
      "loss": 5.1938,
      "mean_token_accuracy": 0.20141583383083345,
      "num_tokens": 47021396.0,
      "step": 21070
    },
    {
      "entropy": 5.526192331314087,
      "epoch": 2.36693435165946,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004446898993858338,
      "loss": 5.2316,
      "mean_token_accuracy": 0.1941898286342621,
      "num_tokens": 47033394.0,
      "step": 21075
    },
    {
      "entropy": 5.485520839691162,
      "epoch": 2.3674959285674175,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044466353219259884,
      "loss": 5.2322,
      "mean_token_accuracy": 0.19840211272239686,
      "num_tokens": 47044688.0,
      "step": 21080
    },
    {
      "entropy": 5.578689193725586,
      "epoch": 2.3680575054753747,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004446371595972114,
      "loss": 5.2939,
      "mean_token_accuracy": 0.19130343198776245,
      "num_tokens": 47057483.0,
      "step": 21085
    },
    {
      "entropy": 5.521618032455445,
      "epoch": 2.3686190823833324,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004446107816005109,
      "loss": 5.2024,
      "mean_token_accuracy": 0.1980918064713478,
      "num_tokens": 47069031.0,
      "step": 21090
    },
    {
      "entropy": 5.469498348236084,
      "epoch": 2.36918065929129,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004445843982033373,
      "loss": 5.229,
      "mean_token_accuracy": 0.19595300406217575,
      "num_tokens": 47080998.0,
      "step": 21095
    },
    {
      "entropy": 5.411626386642456,
      "epoch": 2.3697422361992473,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004445580094065307,
      "loss": 5.2143,
      "mean_token_accuracy": 0.19779896587133408,
      "num_tokens": 47092065.0,
      "step": 21100
    },
    {
      "entropy": 5.495804166793823,
      "epoch": 2.370303813107205,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004445316152109313,
      "loss": 5.2135,
      "mean_token_accuracy": 0.1879710540175438,
      "num_tokens": 47102207.0,
      "step": 21105
    },
    {
      "entropy": 5.464909505844116,
      "epoch": 2.3708653900151626,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004445052156173796,
      "loss": 5.1925,
      "mean_token_accuracy": 0.19657965153455734,
      "num_tokens": 47112675.0,
      "step": 21110
    },
    {
      "entropy": 5.494316959381104,
      "epoch": 2.3714269669231203,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004444788106267161,
      "loss": 5.2422,
      "mean_token_accuracy": 0.20091370046138762,
      "num_tokens": 47123728.0,
      "step": 21115
    },
    {
      "entropy": 5.499920082092285,
      "epoch": 2.3719885438310775,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004444524002397815,
      "loss": 5.2291,
      "mean_token_accuracy": 0.1928659752011299,
      "num_tokens": 47136194.0,
      "step": 21120
    },
    {
      "entropy": 5.387128734588623,
      "epoch": 2.372550120739035,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004444259844574167,
      "loss": 5.0844,
      "mean_token_accuracy": 0.19661763608455657,
      "num_tokens": 47146539.0,
      "step": 21125
    },
    {
      "entropy": 5.5167817115783695,
      "epoch": 2.373111697646993,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044439956328046296,
      "loss": 5.2679,
      "mean_token_accuracy": 0.19259125292301177,
      "num_tokens": 47158449.0,
      "step": 21130
    },
    {
      "entropy": 5.467296409606933,
      "epoch": 2.3736732745549505,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004443731367097613,
      "loss": 5.1154,
      "mean_token_accuracy": 0.20548777729272844,
      "num_tokens": 47169687.0,
      "step": 21135
    },
    {
      "entropy": 5.304896688461303,
      "epoch": 2.3742348514629077,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044434670474615334,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20976399332284928,
      "num_tokens": 47179731.0,
      "step": 21140
    },
    {
      "entropy": 5.37644329071045,
      "epoch": 2.3747964283708654,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044432026739048064,
      "loss": 5.1509,
      "mean_token_accuracy": 0.20831314772367476,
      "num_tokens": 47189925.0,
      "step": 21145
    },
    {
      "entropy": 5.468923187255859,
      "epoch": 2.375358005278823,
      "grad_norm": 1.171875,
      "learning_rate": 0.000444293824643585,
      "loss": 5.1981,
      "mean_token_accuracy": 0.20327981561422348,
      "num_tokens": 47200637.0,
      "step": 21150
    },
    {
      "entropy": 5.496252775192261,
      "epoch": 2.3759195821867802,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004442673765063082,
      "loss": 5.1965,
      "mean_token_accuracy": 0.19572321623563765,
      "num_tokens": 47211969.0,
      "step": 21155
    },
    {
      "entropy": 5.51967978477478,
      "epoch": 2.376481159094738,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004442409229794925,
      "loss": 5.3411,
      "mean_token_accuracy": 0.19481445103883743,
      "num_tokens": 47223150.0,
      "step": 21160
    },
    {
      "entropy": 5.558747005462647,
      "epoch": 2.3770427360026956,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004442144640639803,
      "loss": 5.2928,
      "mean_token_accuracy": 0.1921680375933647,
      "num_tokens": 47234403.0,
      "step": 21165
    },
    {
      "entropy": 5.467461252212525,
      "epoch": 2.377604312910653,
      "grad_norm": 1.125,
      "learning_rate": 0.0004441879997606138,
      "loss": 5.1148,
      "mean_token_accuracy": 0.19939623326063155,
      "num_tokens": 47245800.0,
      "step": 21170
    },
    {
      "entropy": 5.438232564926148,
      "epoch": 2.378165889818611,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004441615300702358,
      "loss": 5.1934,
      "mean_token_accuracy": 0.1920433059334755,
      "num_tokens": 47256109.0,
      "step": 21175
    },
    {
      "entropy": 5.431833076477051,
      "epoch": 2.378727466726568,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004441350549936891,
      "loss": 5.1239,
      "mean_token_accuracy": 0.20218961238861083,
      "num_tokens": 47266502.0,
      "step": 21180
    },
    {
      "entropy": 5.517393159866333,
      "epoch": 2.3792890436345258,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004441085745318167,
      "loss": 5.2107,
      "mean_token_accuracy": 0.20080445408821107,
      "num_tokens": 47278866.0,
      "step": 21185
    },
    {
      "entropy": 5.371696949005127,
      "epoch": 2.3798506205424834,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044408208868546164,
      "loss": 5.1539,
      "mean_token_accuracy": 0.20703070163726806,
      "num_tokens": 47289449.0,
      "step": 21190
    },
    {
      "entropy": 5.440367364883423,
      "epoch": 2.3804121974504406,
      "grad_norm": 1.125,
      "learning_rate": 0.0004440555974554673,
      "loss": 5.2964,
      "mean_token_accuracy": 0.2001742973923683,
      "num_tokens": 47301026.0,
      "step": 21195
    },
    {
      "entropy": 5.483521270751953,
      "epoch": 2.3809737743583983,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044402910084267717,
      "loss": 5.1938,
      "mean_token_accuracy": 0.1970280796289444,
      "num_tokens": 47311579.0,
      "step": 21200
    },
    {
      "entropy": 5.457720613479614,
      "epoch": 2.381535351266356,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004440025988479349,
      "loss": 5.1977,
      "mean_token_accuracy": 0.2033959746360779,
      "num_tokens": 47322042.0,
      "step": 21205
    },
    {
      "entropy": 5.470384311676026,
      "epoch": 2.3820969281743136,
      "grad_norm": 1.125,
      "learning_rate": 0.00044397609147208436,
      "loss": 5.156,
      "mean_token_accuracy": 0.2049620494246483,
      "num_tokens": 47332821.0,
      "step": 21210
    },
    {
      "entropy": 5.454410362243652,
      "epoch": 2.382658505082271,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004439495787159695,
      "loss": 5.2077,
      "mean_token_accuracy": 0.1985888212919235,
      "num_tokens": 47344221.0,
      "step": 21215
    },
    {
      "entropy": 5.586800050735474,
      "epoch": 2.3832200819902285,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044392306058043453,
      "loss": 5.3565,
      "mean_token_accuracy": 0.19308149218559265,
      "num_tokens": 47356424.0,
      "step": 21220
    },
    {
      "entropy": 5.4759235858917235,
      "epoch": 2.383781658898186,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004438965370663238,
      "loss": 5.1769,
      "mean_token_accuracy": 0.20478518456220626,
      "num_tokens": 47367546.0,
      "step": 21225
    },
    {
      "entropy": 5.512812709808349,
      "epoch": 2.384343235806144,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004438700081744819,
      "loss": 5.2764,
      "mean_token_accuracy": 0.1924913927912712,
      "num_tokens": 47378281.0,
      "step": 21230
    },
    {
      "entropy": 5.460228967666626,
      "epoch": 2.384904812714101,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044384347390575337,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20389873534440994,
      "num_tokens": 47389233.0,
      "step": 21235
    },
    {
      "entropy": 5.42513484954834,
      "epoch": 2.3854663896220587,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004438169342609831,
      "loss": 5.1965,
      "mean_token_accuracy": 0.19487940967082978,
      "num_tokens": 47400114.0,
      "step": 21240
    },
    {
      "entropy": 5.627143526077271,
      "epoch": 2.3860279665300164,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044379038924101627,
      "loss": 5.374,
      "mean_token_accuracy": 0.19131412953138352,
      "num_tokens": 47411926.0,
      "step": 21245
    },
    {
      "entropy": 5.578207778930664,
      "epoch": 2.3865895434379736,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004437638388466979,
      "loss": 5.2383,
      "mean_token_accuracy": 0.1985018417239189,
      "num_tokens": 47424048.0,
      "step": 21250
    },
    {
      "entropy": 5.4164690494537355,
      "epoch": 2.3871511203459312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044373728307887353,
      "loss": 5.148,
      "mean_token_accuracy": 0.19993937909603118,
      "num_tokens": 47435581.0,
      "step": 21255
    },
    {
      "entropy": 5.447855997085571,
      "epoch": 2.387712697253889,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044371072193838857,
      "loss": 5.1636,
      "mean_token_accuracy": 0.20491148233413697,
      "num_tokens": 47446831.0,
      "step": 21260
    },
    {
      "entropy": 5.3792437553405765,
      "epoch": 2.3882742741618466,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044368415542608874,
      "loss": 5.1374,
      "mean_token_accuracy": 0.19673147648572922,
      "num_tokens": 47457885.0,
      "step": 21265
    },
    {
      "entropy": 5.497529602050781,
      "epoch": 2.3888358510698042,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044365758354282005,
      "loss": 5.2463,
      "mean_token_accuracy": 0.19050194919109345,
      "num_tokens": 47470195.0,
      "step": 21270
    },
    {
      "entropy": 5.492781734466552,
      "epoch": 2.3893974279777614,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004436310062894285,
      "loss": 5.1939,
      "mean_token_accuracy": 0.20272512584924698,
      "num_tokens": 47481068.0,
      "step": 21275
    },
    {
      "entropy": 5.52130765914917,
      "epoch": 2.389959004885719,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004436044236667603,
      "loss": 5.1542,
      "mean_token_accuracy": 0.2017817109823227,
      "num_tokens": 47491409.0,
      "step": 21280
    },
    {
      "entropy": 5.424414110183716,
      "epoch": 2.3905205817936768,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004435778356756618,
      "loss": 5.2363,
      "mean_token_accuracy": 0.19426699727773666,
      "num_tokens": 47504440.0,
      "step": 21285
    },
    {
      "entropy": 5.434523010253907,
      "epoch": 2.391082158701634,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044355124231697967,
      "loss": 5.124,
      "mean_token_accuracy": 0.20635359287261962,
      "num_tokens": 47514370.0,
      "step": 21290
    },
    {
      "entropy": 5.462828874588013,
      "epoch": 2.3916437356095916,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004435246435915607,
      "loss": 5.1682,
      "mean_token_accuracy": 0.2015223890542984,
      "num_tokens": 47525415.0,
      "step": 21295
    },
    {
      "entropy": 5.521757555007935,
      "epoch": 2.3922053125175493,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004434980395002516,
      "loss": 5.2496,
      "mean_token_accuracy": 0.19774934351444245,
      "num_tokens": 47537029.0,
      "step": 21300
    },
    {
      "entropy": 5.470596027374268,
      "epoch": 2.392766889425507,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044347143004389955,
      "loss": 5.1567,
      "mean_token_accuracy": 0.19732739329338073,
      "num_tokens": 47550434.0,
      "step": 21305
    },
    {
      "entropy": 5.438879728317261,
      "epoch": 2.393328466333464,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004434448152233519,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20404388010501862,
      "num_tokens": 47560985.0,
      "step": 21310
    },
    {
      "entropy": 5.51512770652771,
      "epoch": 2.393890043241422,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044341819503945594,
      "loss": 5.2735,
      "mean_token_accuracy": 0.19304731190204621,
      "num_tokens": 47573207.0,
      "step": 21315
    },
    {
      "entropy": 5.487081050872803,
      "epoch": 2.3944516201493795,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004433915694930594,
      "loss": 5.0995,
      "mean_token_accuracy": 0.21096139401197433,
      "num_tokens": 47584666.0,
      "step": 21320
    },
    {
      "entropy": 5.474404954910279,
      "epoch": 2.395013197057337,
      "grad_norm": 1.25,
      "learning_rate": 0.0004433649385850099,
      "loss": 5.1937,
      "mean_token_accuracy": 0.19833858609199523,
      "num_tokens": 47598017.0,
      "step": 21325
    },
    {
      "entropy": 5.507979106903076,
      "epoch": 2.3955747739652944,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004433383023161555,
      "loss": 5.3571,
      "mean_token_accuracy": 0.18930508196353912,
      "num_tokens": 47610947.0,
      "step": 21330
    },
    {
      "entropy": 5.5602569580078125,
      "epoch": 2.396136350873252,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004433116606873443,
      "loss": 5.2842,
      "mean_token_accuracy": 0.1920780673623085,
      "num_tokens": 47621393.0,
      "step": 21335
    },
    {
      "entropy": 5.562190246582031,
      "epoch": 2.3966979277812097,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044328501369942454,
      "loss": 5.246,
      "mean_token_accuracy": 0.19772695749998093,
      "num_tokens": 47631615.0,
      "step": 21340
    },
    {
      "entropy": 5.461170768737793,
      "epoch": 2.3972595046891674,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004432583613532446,
      "loss": 5.215,
      "mean_token_accuracy": 0.19658462107181549,
      "num_tokens": 47642813.0,
      "step": 21345
    },
    {
      "entropy": 5.640759611129761,
      "epoch": 2.3978210815971246,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004432317036496533,
      "loss": 5.2835,
      "mean_token_accuracy": 0.19452656507492067,
      "num_tokens": 47654012.0,
      "step": 21350
    },
    {
      "entropy": 5.476614236831665,
      "epoch": 2.3983826585050823,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044320504058949916,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20887311398983002,
      "num_tokens": 47664575.0,
      "step": 21355
    },
    {
      "entropy": 5.422935676574707,
      "epoch": 2.39894423541304,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004431783721736314,
      "loss": 5.2253,
      "mean_token_accuracy": 0.19494512677192688,
      "num_tokens": 47674763.0,
      "step": 21360
    },
    {
      "entropy": 5.421364879608154,
      "epoch": 2.3995058123209976,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044315169840289896,
      "loss": 5.1746,
      "mean_token_accuracy": 0.20039768964052201,
      "num_tokens": 47685435.0,
      "step": 21365
    },
    {
      "entropy": 5.609927320480347,
      "epoch": 2.400067389228955,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004431250192781513,
      "loss": 5.3662,
      "mean_token_accuracy": 0.19823579341173173,
      "num_tokens": 47696634.0,
      "step": 21370
    },
    {
      "entropy": 5.501060819625854,
      "epoch": 2.4006289661369125,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004430983348002378,
      "loss": 5.2362,
      "mean_token_accuracy": 0.19303722083568572,
      "num_tokens": 47707737.0,
      "step": 21375
    },
    {
      "entropy": 5.535626077651978,
      "epoch": 2.40119054304487,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044307164497000813,
      "loss": 5.2156,
      "mean_token_accuracy": 0.19697881042957305,
      "num_tokens": 47718131.0,
      "step": 21380
    },
    {
      "entropy": 5.419151163101196,
      "epoch": 2.4017521199528273,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044304494978831204,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19738378822803498,
      "num_tokens": 47730154.0,
      "step": 21385
    },
    {
      "entropy": 5.4863170146942135,
      "epoch": 2.402313696860785,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004430182492559997,
      "loss": 5.1709,
      "mean_token_accuracy": 0.19812498688697816,
      "num_tokens": 47741893.0,
      "step": 21390
    },
    {
      "entropy": 5.505048179626465,
      "epoch": 2.4028752737687427,
      "grad_norm": 1.203125,
      "learning_rate": 0.000442991543373921,
      "loss": 5.2499,
      "mean_token_accuracy": 0.19859177023172378,
      "num_tokens": 47753430.0,
      "step": 21395
    },
    {
      "entropy": 5.3799785614013675,
      "epoch": 2.4034368506767003,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004429648321429264,
      "loss": 5.159,
      "mean_token_accuracy": 0.2043390080332756,
      "num_tokens": 47765181.0,
      "step": 21400
    },
    {
      "entropy": 5.588997793197632,
      "epoch": 2.4039984275846575,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044293811556386656,
      "loss": 5.3652,
      "mean_token_accuracy": 0.18540471643209458,
      "num_tokens": 47777211.0,
      "step": 21405
    },
    {
      "entropy": 5.46024489402771,
      "epoch": 2.404560004492615,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004429113936375918,
      "loss": 5.1676,
      "mean_token_accuracy": 0.2016657367348671,
      "num_tokens": 47788397.0,
      "step": 21410
    },
    {
      "entropy": 5.524742555618286,
      "epoch": 2.405121581400573,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044288466636495323,
      "loss": 5.2796,
      "mean_token_accuracy": 0.19366027265787125,
      "num_tokens": 47799870.0,
      "step": 21415
    },
    {
      "entropy": 5.476950454711914,
      "epoch": 2.4056831583085305,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044285793374680184,
      "loss": 5.2725,
      "mean_token_accuracy": 0.19704962372779847,
      "num_tokens": 47810696.0,
      "step": 21420
    },
    {
      "entropy": 5.512076807022095,
      "epoch": 2.4062447352164877,
      "grad_norm": 1.125,
      "learning_rate": 0.00044283119578398866,
      "loss": 5.2361,
      "mean_token_accuracy": 0.1989618495106697,
      "num_tokens": 47822579.0,
      "step": 21425
    },
    {
      "entropy": 5.459765291213989,
      "epoch": 2.4068063121244454,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004428044524773651,
      "loss": 5.1445,
      "mean_token_accuracy": 0.19508033692836763,
      "num_tokens": 47832556.0,
      "step": 21430
    },
    {
      "entropy": 5.459570598602295,
      "epoch": 2.407367889032403,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004427777038277827,
      "loss": 5.2222,
      "mean_token_accuracy": 0.19387545585632324,
      "num_tokens": 47844409.0,
      "step": 21435
    },
    {
      "entropy": 5.482930707931518,
      "epoch": 2.4079294659403607,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044275094983609313,
      "loss": 5.1683,
      "mean_token_accuracy": 0.20039867460727692,
      "num_tokens": 47855666.0,
      "step": 21440
    },
    {
      "entropy": 5.3964927196502686,
      "epoch": 2.408491042848318,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044272419050314833,
      "loss": 5.0914,
      "mean_token_accuracy": 0.2102293461561203,
      "num_tokens": 47866261.0,
      "step": 21445
    },
    {
      "entropy": 5.434059476852417,
      "epoch": 2.4090526197562756,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044269742582980017,
      "loss": 5.3075,
      "mean_token_accuracy": 0.19146720468997955,
      "num_tokens": 47878807.0,
      "step": 21450
    },
    {
      "entropy": 5.467469930648804,
      "epoch": 2.4096141966642333,
      "grad_norm": 1.484375,
      "learning_rate": 0.000442670655816901,
      "loss": 5.1535,
      "mean_token_accuracy": 0.20308677107095718,
      "num_tokens": 47888935.0,
      "step": 21455
    },
    {
      "entropy": 5.539342975616455,
      "epoch": 2.410175773572191,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044264388046530304,
      "loss": 5.1845,
      "mean_token_accuracy": 0.20222682803869246,
      "num_tokens": 47900208.0,
      "step": 21460
    },
    {
      "entropy": 5.5797394752502445,
      "epoch": 2.410737350480148,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044261709977585893,
      "loss": 5.3258,
      "mean_token_accuracy": 0.1934771865606308,
      "num_tokens": 47911131.0,
      "step": 21465
    },
    {
      "entropy": 5.493864965438843,
      "epoch": 2.411298927388106,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004425903137494214,
      "loss": 5.2858,
      "mean_token_accuracy": 0.19310020953416823,
      "num_tokens": 47922601.0,
      "step": 21470
    },
    {
      "entropy": 5.562084770202636,
      "epoch": 2.4118605042960635,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004425635223868433,
      "loss": 5.2483,
      "mean_token_accuracy": 0.19078425019979478,
      "num_tokens": 47933168.0,
      "step": 21475
    },
    {
      "entropy": 5.567868947982788,
      "epoch": 2.4124220812040207,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00044253672568897754,
      "loss": 5.326,
      "mean_token_accuracy": 0.18399229794740676,
      "num_tokens": 47945433.0,
      "step": 21480
    },
    {
      "entropy": 5.510581064224243,
      "epoch": 2.4129836581119783,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004425099236566775,
      "loss": 5.253,
      "mean_token_accuracy": 0.19290145188570024,
      "num_tokens": 47957513.0,
      "step": 21485
    },
    {
      "entropy": 5.476819801330566,
      "epoch": 2.413545235019936,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044248311629079657,
      "loss": 5.1942,
      "mean_token_accuracy": 0.19590148329734802,
      "num_tokens": 47968817.0,
      "step": 21490
    },
    {
      "entropy": 5.492171716690064,
      "epoch": 2.4141068119278937,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004424563035921882,
      "loss": 5.1331,
      "mean_token_accuracy": 0.2000505194067955,
      "num_tokens": 47979034.0,
      "step": 21495
    },
    {
      "entropy": 5.542651653289795,
      "epoch": 2.414668388835851,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044242948556170634,
      "loss": 5.2197,
      "mean_token_accuracy": 0.2020846948027611,
      "num_tokens": 47990459.0,
      "step": 21500
    },
    {
      "entropy": 5.4815925598144535,
      "epoch": 2.4152299657438085,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00044240266220020455,
      "loss": 5.2593,
      "mean_token_accuracy": 0.19531419426202773,
      "num_tokens": 48000007.0,
      "step": 21505
    },
    {
      "entropy": 5.507301902770996,
      "epoch": 2.415791542651766,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004423758335085371,
      "loss": 5.2188,
      "mean_token_accuracy": 0.19626193791627883,
      "num_tokens": 48012962.0,
      "step": 21510
    },
    {
      "entropy": 5.465673303604126,
      "epoch": 2.416353119559724,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044234899948755827,
      "loss": 5.2093,
      "mean_token_accuracy": 0.1957627147436142,
      "num_tokens": 48025916.0,
      "step": 21515
    },
    {
      "entropy": 5.455223655700683,
      "epoch": 2.416914696467681,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044232216013812226,
      "loss": 5.2095,
      "mean_token_accuracy": 0.20134610384702684,
      "num_tokens": 48036255.0,
      "step": 21520
    },
    {
      "entropy": 5.489645862579346,
      "epoch": 2.4174762733756388,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004422953154610838,
      "loss": 5.2089,
      "mean_token_accuracy": 0.20038269311189652,
      "num_tokens": 48045943.0,
      "step": 21525
    },
    {
      "entropy": 5.406678152084351,
      "epoch": 2.4180378502835964,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004422684654572977,
      "loss": 5.1716,
      "mean_token_accuracy": 0.20007821619510652,
      "num_tokens": 48056562.0,
      "step": 21530
    },
    {
      "entropy": 5.414686012268066,
      "epoch": 2.418599427191554,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044224161012761874,
      "loss": 5.1455,
      "mean_token_accuracy": 0.19966147243976592,
      "num_tokens": 48066997.0,
      "step": 21535
    },
    {
      "entropy": 5.52472152709961,
      "epoch": 2.4191610040995113,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044221474947290205,
      "loss": 5.2624,
      "mean_token_accuracy": 0.19443010538816452,
      "num_tokens": 48077810.0,
      "step": 21540
    },
    {
      "entropy": 5.542490816116333,
      "epoch": 2.419722581007469,
      "grad_norm": 1.125,
      "learning_rate": 0.0004421878834940029,
      "loss": 5.3266,
      "mean_token_accuracy": 0.18872090429067612,
      "num_tokens": 48089422.0,
      "step": 21545
    },
    {
      "entropy": 5.4717436790466305,
      "epoch": 2.4202841579154266,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044216101219177667,
      "loss": 5.18,
      "mean_token_accuracy": 0.19570807367563248,
      "num_tokens": 48101114.0,
      "step": 21550
    },
    {
      "entropy": 5.48274073600769,
      "epoch": 2.4208457348233843,
      "grad_norm": 1.125,
      "learning_rate": 0.00044213413556707905,
      "loss": 5.1681,
      "mean_token_accuracy": 0.2029666617512703,
      "num_tokens": 48110991.0,
      "step": 21555
    },
    {
      "entropy": 5.508170175552368,
      "epoch": 2.4214073117313415,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004421072536207656,
      "loss": 5.2141,
      "mean_token_accuracy": 0.2002914935350418,
      "num_tokens": 48122889.0,
      "step": 21560
    },
    {
      "entropy": 5.533147811889648,
      "epoch": 2.421968888639299,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044208036635369247,
      "loss": 5.2218,
      "mean_token_accuracy": 0.20043204873800277,
      "num_tokens": 48133586.0,
      "step": 21565
    },
    {
      "entropy": 5.443800926208496,
      "epoch": 2.422530465547257,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004420534737667157,
      "loss": 5.1527,
      "mean_token_accuracy": 0.20333350151777269,
      "num_tokens": 48144060.0,
      "step": 21570
    },
    {
      "entropy": 5.54448938369751,
      "epoch": 2.423092042455214,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044202657586069153,
      "loss": 5.2599,
      "mean_token_accuracy": 0.2001522049307823,
      "num_tokens": 48155197.0,
      "step": 21575
    },
    {
      "entropy": 5.5493550300598145,
      "epoch": 2.4236536193631717,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044199967263647635,
      "loss": 5.3169,
      "mean_token_accuracy": 0.19386566579341888,
      "num_tokens": 48166990.0,
      "step": 21580
    },
    {
      "entropy": 5.567315483093262,
      "epoch": 2.4242151962711294,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004419727640949268,
      "loss": 5.3224,
      "mean_token_accuracy": 0.18650861829519272,
      "num_tokens": 48179454.0,
      "step": 21585
    },
    {
      "entropy": 5.418863344192505,
      "epoch": 2.424776773179087,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004419458502368998,
      "loss": 5.1721,
      "mean_token_accuracy": 0.20034961998462678,
      "num_tokens": 48192009.0,
      "step": 21590
    },
    {
      "entropy": 5.5574883937835695,
      "epoch": 2.4253383500870442,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044191893106325207,
      "loss": 5.2552,
      "mean_token_accuracy": 0.197950841486454,
      "num_tokens": 48203111.0,
      "step": 21595
    },
    {
      "entropy": 5.595216894149781,
      "epoch": 2.425899926995002,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004418920065748409,
      "loss": 5.4157,
      "mean_token_accuracy": 0.1881189227104187,
      "num_tokens": 48214766.0,
      "step": 21600
    },
    {
      "entropy": 5.544545269012451,
      "epoch": 2.4264615039029596,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044186507677252345,
      "loss": 5.3159,
      "mean_token_accuracy": 0.2030021384358406,
      "num_tokens": 48225061.0,
      "step": 21605
    },
    {
      "entropy": 5.469171810150146,
      "epoch": 2.427023080810917,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004418381416571573,
      "loss": 5.2038,
      "mean_token_accuracy": 0.19727595150470734,
      "num_tokens": 48236101.0,
      "step": 21610
    },
    {
      "entropy": 5.4103844165802,
      "epoch": 2.4275846577188744,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004418112012296,
      "loss": 5.1742,
      "mean_token_accuracy": 0.1981567308306694,
      "num_tokens": 48248080.0,
      "step": 21615
    },
    {
      "entropy": 5.42638144493103,
      "epoch": 2.428146234626832,
      "grad_norm": 1.484375,
      "learning_rate": 0.00044178425549070934,
      "loss": 5.1343,
      "mean_token_accuracy": 0.20125969499349594,
      "num_tokens": 48259177.0,
      "step": 21620
    },
    {
      "entropy": 5.496696472167969,
      "epoch": 2.4287078115347898,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044175730444134326,
      "loss": 5.1755,
      "mean_token_accuracy": 0.19469186514616013,
      "num_tokens": 48271345.0,
      "step": 21625
    },
    {
      "entropy": 5.517582368850708,
      "epoch": 2.4292693884427474,
      "grad_norm": 1.125,
      "learning_rate": 0.00044173034808236,
      "loss": 5.2608,
      "mean_token_accuracy": 0.19164094030857087,
      "num_tokens": 48284219.0,
      "step": 21630
    },
    {
      "entropy": 5.498813247680664,
      "epoch": 2.4298309653507046,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004417033864146177,
      "loss": 5.1755,
      "mean_token_accuracy": 0.19849569201469422,
      "num_tokens": 48294626.0,
      "step": 21635
    },
    {
      "entropy": 5.380316543579101,
      "epoch": 2.4303925422586623,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004416764194389751,
      "loss": 5.1676,
      "mean_token_accuracy": 0.1984646886587143,
      "num_tokens": 48305459.0,
      "step": 21640
    },
    {
      "entropy": 5.464716053009033,
      "epoch": 2.43095411916662,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004416494471562904,
      "loss": 5.1715,
      "mean_token_accuracy": 0.2041002705693245,
      "num_tokens": 48316232.0,
      "step": 21645
    },
    {
      "entropy": 5.470191144943238,
      "epoch": 2.4315156960745776,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004416224695674229,
      "loss": 5.145,
      "mean_token_accuracy": 0.2063722714781761,
      "num_tokens": 48326711.0,
      "step": 21650
    },
    {
      "entropy": 5.318240833282471,
      "epoch": 2.432077272982535,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00044159548667323126,
      "loss": 4.9612,
      "mean_token_accuracy": 0.22063706517219545,
      "num_tokens": 48336860.0,
      "step": 21655
    },
    {
      "entropy": 5.347515010833741,
      "epoch": 2.4326388498904925,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004415684984745747,
      "loss": 5.2694,
      "mean_token_accuracy": 0.19604750275611876,
      "num_tokens": 48348503.0,
      "step": 21660
    },
    {
      "entropy": 5.524180173873901,
      "epoch": 2.43320042679845,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004415415049723126,
      "loss": 5.2929,
      "mean_token_accuracy": 0.19336607158184052,
      "num_tokens": 48359855.0,
      "step": 21665
    },
    {
      "entropy": 5.56565842628479,
      "epoch": 2.4337620037064074,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004415145061673043,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1919916331768036,
      "num_tokens": 48370188.0,
      "step": 21670
    },
    {
      "entropy": 5.478146171569824,
      "epoch": 2.434323580614365,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004414875020604096,
      "loss": 5.1624,
      "mean_token_accuracy": 0.20773671567440033,
      "num_tokens": 48381152.0,
      "step": 21675
    },
    {
      "entropy": 5.454023694992065,
      "epoch": 2.4348851575223227,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004414604926524882,
      "loss": 5.2145,
      "mean_token_accuracy": 0.1982116147875786,
      "num_tokens": 48392721.0,
      "step": 21680
    },
    {
      "entropy": 5.455103158950806,
      "epoch": 2.4354467344302804,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004414334779444002,
      "loss": 5.1895,
      "mean_token_accuracy": 0.1979516863822937,
      "num_tokens": 48404071.0,
      "step": 21685
    },
    {
      "entropy": 5.473447799682617,
      "epoch": 2.4360083113382376,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044140645793700573,
      "loss": 5.2013,
      "mean_token_accuracy": 0.19501965045928954,
      "num_tokens": 48416001.0,
      "step": 21690
    },
    {
      "entropy": 5.4764386177062985,
      "epoch": 2.4365698882461952,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044137943263116515,
      "loss": 5.204,
      "mean_token_accuracy": 0.19583554565906525,
      "num_tokens": 48426880.0,
      "step": 21695
    },
    {
      "entropy": 5.583257532119751,
      "epoch": 2.437131465154153,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004413524020277388,
      "loss": 5.3075,
      "mean_token_accuracy": 0.19080916196107864,
      "num_tokens": 48437147.0,
      "step": 21700
    },
    {
      "entropy": 5.4255836486816404,
      "epoch": 2.4376930420621106,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004413253661275875,
      "loss": 5.0882,
      "mean_token_accuracy": 0.2037639334797859,
      "num_tokens": 48448706.0,
      "step": 21705
    },
    {
      "entropy": 5.435296058654785,
      "epoch": 2.438254618970068,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044129832493157195,
      "loss": 5.1096,
      "mean_token_accuracy": 0.20613910853862763,
      "num_tokens": 48458873.0,
      "step": 21710
    },
    {
      "entropy": 5.489786100387573,
      "epoch": 2.4388161958780255,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044127127844055335,
      "loss": 5.2789,
      "mean_token_accuracy": 0.19208460450172424,
      "num_tokens": 48470726.0,
      "step": 21715
    },
    {
      "entropy": 5.503612422943116,
      "epoch": 2.439377772785983,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044124422665539273,
      "loss": 5.103,
      "mean_token_accuracy": 0.20683069825172423,
      "num_tokens": 48481912.0,
      "step": 21720
    },
    {
      "entropy": 5.476216745376587,
      "epoch": 2.4399393496939408,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004412171695769515,
      "loss": 5.228,
      "mean_token_accuracy": 0.20174122005701065,
      "num_tokens": 48492048.0,
      "step": 21725
    },
    {
      "entropy": 5.468652677536011,
      "epoch": 2.440500926601898,
      "grad_norm": 1.125,
      "learning_rate": 0.00044119010720609106,
      "loss": 5.2593,
      "mean_token_accuracy": 0.19893279373645784,
      "num_tokens": 48503190.0,
      "step": 21730
    },
    {
      "entropy": 5.560579586029053,
      "epoch": 2.4410625035098557,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044116303954367314,
      "loss": 5.1611,
      "mean_token_accuracy": 0.19728001654148103,
      "num_tokens": 48514956.0,
      "step": 21735
    },
    {
      "entropy": 5.579486322402954,
      "epoch": 2.4416240804178133,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004411359665905596,
      "loss": 5.3054,
      "mean_token_accuracy": 0.19709682762622832,
      "num_tokens": 48526372.0,
      "step": 21740
    },
    {
      "entropy": 5.377999258041382,
      "epoch": 2.442185657325771,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004411088883476125,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20003522783517838,
      "num_tokens": 48536985.0,
      "step": 21745
    },
    {
      "entropy": 5.459693336486817,
      "epoch": 2.442747234233728,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044108180481569387,
      "loss": 5.2078,
      "mean_token_accuracy": 0.20046344995498658,
      "num_tokens": 48547505.0,
      "step": 21750
    },
    {
      "entropy": 5.511417961120605,
      "epoch": 2.443308811141686,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044105471599566626,
      "loss": 5.2378,
      "mean_token_accuracy": 0.19674644917249678,
      "num_tokens": 48560104.0,
      "step": 21755
    },
    {
      "entropy": 5.521594619750976,
      "epoch": 2.4438703880496435,
      "grad_norm": 1.109375,
      "learning_rate": 0.000441027621888392,
      "loss": 5.3197,
      "mean_token_accuracy": 0.19993076771497725,
      "num_tokens": 48571203.0,
      "step": 21760
    },
    {
      "entropy": 5.4625563621521,
      "epoch": 2.4444319649576007,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044100052249473393,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19899989515542985,
      "num_tokens": 48581480.0,
      "step": 21765
    },
    {
      "entropy": 5.509932231903076,
      "epoch": 2.4449935418655584,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004409734178155548,
      "loss": 5.1919,
      "mean_token_accuracy": 0.20069716721773148,
      "num_tokens": 48592811.0,
      "step": 21770
    },
    {
      "entropy": 5.420577096939087,
      "epoch": 2.445555118773516,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004409463078517178,
      "loss": 5.2034,
      "mean_token_accuracy": 0.203411503136158,
      "num_tokens": 48604245.0,
      "step": 21775
    },
    {
      "entropy": 5.5184472560882565,
      "epoch": 2.4461166956814737,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044091919260408585,
      "loss": 5.1847,
      "mean_token_accuracy": 0.20057608783245087,
      "num_tokens": 48615138.0,
      "step": 21780
    },
    {
      "entropy": 5.48538908958435,
      "epoch": 2.446678272589431,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004408920720735225,
      "loss": 5.2473,
      "mean_token_accuracy": 0.19808796793222427,
      "num_tokens": 48626084.0,
      "step": 21785
    },
    {
      "entropy": 5.453407573699951,
      "epoch": 2.4472398494973886,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044086494626089127,
      "loss": 5.1067,
      "mean_token_accuracy": 0.1999150887131691,
      "num_tokens": 48636425.0,
      "step": 21790
    },
    {
      "entropy": 5.417506361007691,
      "epoch": 2.4478014264053463,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004408378151670558,
      "loss": 5.2456,
      "mean_token_accuracy": 0.1862782880663872,
      "num_tokens": 48646941.0,
      "step": 21795
    },
    {
      "entropy": 5.443913698196411,
      "epoch": 2.448363003313304,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044081067879288,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20425938665866852,
      "num_tokens": 48657481.0,
      "step": 21800
    },
    {
      "entropy": 5.540360164642334,
      "epoch": 2.448924580221261,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004407835371392279,
      "loss": 5.2293,
      "mean_token_accuracy": 0.19924686402082442,
      "num_tokens": 48669306.0,
      "step": 21805
    },
    {
      "entropy": 5.438464879989624,
      "epoch": 2.449486157129219,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004407563902069636,
      "loss": 5.1822,
      "mean_token_accuracy": 0.20206608772277831,
      "num_tokens": 48680789.0,
      "step": 21810
    },
    {
      "entropy": 5.499397230148316,
      "epoch": 2.4500477340371765,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004407292379969515,
      "loss": 5.2781,
      "mean_token_accuracy": 0.19427161812782287,
      "num_tokens": 48691811.0,
      "step": 21815
    },
    {
      "entropy": 5.505946969985962,
      "epoch": 2.450609310945134,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004407020805100563,
      "loss": 5.268,
      "mean_token_accuracy": 0.19791758358478545,
      "num_tokens": 48704248.0,
      "step": 21820
    },
    {
      "entropy": 5.492702436447144,
      "epoch": 2.4511708878530913,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004406749177471425,
      "loss": 5.1747,
      "mean_token_accuracy": 0.20821383446455002,
      "num_tokens": 48714452.0,
      "step": 21825
    },
    {
      "entropy": 5.477393770217896,
      "epoch": 2.451732464761049,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044064774970907516,
      "loss": 5.1556,
      "mean_token_accuracy": 0.20354018807411195,
      "num_tokens": 48725481.0,
      "step": 21830
    },
    {
      "entropy": 5.532561540603638,
      "epoch": 2.4522940416690067,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004406205763967193,
      "loss": 5.2707,
      "mean_token_accuracy": 0.1989121988415718,
      "num_tokens": 48737704.0,
      "step": 21835
    },
    {
      "entropy": 5.521845436096191,
      "epoch": 2.4528556185769643,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004405933978109398,
      "loss": 5.2581,
      "mean_token_accuracy": 0.1946515366435051,
      "num_tokens": 48749085.0,
      "step": 21840
    },
    {
      "entropy": 5.5241645812988285,
      "epoch": 2.4534171954849215,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044056621395260247,
      "loss": 5.2389,
      "mean_token_accuracy": 0.19579836428165437,
      "num_tokens": 48759432.0,
      "step": 21845
    },
    {
      "entropy": 5.38377537727356,
      "epoch": 2.453978772392879,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044053902482257265,
      "loss": 5.1048,
      "mean_token_accuracy": 0.1968403398990631,
      "num_tokens": 48769549.0,
      "step": 21850
    },
    {
      "entropy": 5.428978109359742,
      "epoch": 2.454540349300837,
      "grad_norm": 1.15625,
      "learning_rate": 0.000440511830421716,
      "loss": 5.1542,
      "mean_token_accuracy": 0.20203926116228105,
      "num_tokens": 48779630.0,
      "step": 21855
    },
    {
      "entropy": 5.5175073623657225,
      "epoch": 2.455101926208794,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004404846307508985,
      "loss": 5.2374,
      "mean_token_accuracy": 0.19475793093442917,
      "num_tokens": 48790489.0,
      "step": 21860
    },
    {
      "entropy": 5.581151103973388,
      "epoch": 2.4556635031167517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044045742581098624,
      "loss": 5.2892,
      "mean_token_accuracy": 0.19683291167020797,
      "num_tokens": 48802310.0,
      "step": 21865
    },
    {
      "entropy": 5.43082184791565,
      "epoch": 2.4562250800247094,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004404302156028453,
      "loss": 5.1457,
      "mean_token_accuracy": 0.20765147358179092,
      "num_tokens": 48812066.0,
      "step": 21870
    },
    {
      "entropy": 5.533325910568237,
      "epoch": 2.456786656932667,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044040300012734217,
      "loss": 5.3751,
      "mean_token_accuracy": 0.18804677575826645,
      "num_tokens": 48824477.0,
      "step": 21875
    },
    {
      "entropy": 5.545376491546631,
      "epoch": 2.4573482338406243,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004403757793853434,
      "loss": 5.2161,
      "mean_token_accuracy": 0.1918596476316452,
      "num_tokens": 48835147.0,
      "step": 21880
    },
    {
      "entropy": 5.436745023727417,
      "epoch": 2.457909810748582,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004403485533777157,
      "loss": 5.2452,
      "mean_token_accuracy": 0.19451164007186889,
      "num_tokens": 48846147.0,
      "step": 21885
    },
    {
      "entropy": 5.581682920455933,
      "epoch": 2.4584713876565396,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004403213221053258,
      "loss": 5.3094,
      "mean_token_accuracy": 0.19171257615089415,
      "num_tokens": 48857340.0,
      "step": 21890
    },
    {
      "entropy": 5.432861423492431,
      "epoch": 2.4590329645644973,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044029408556904093,
      "loss": 5.1574,
      "mean_token_accuracy": 0.2004673019051552,
      "num_tokens": 48868256.0,
      "step": 21895
    },
    {
      "entropy": 5.587519788742066,
      "epoch": 2.4595945414724545,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004402668437697283,
      "loss": 5.2699,
      "mean_token_accuracy": 0.19470858573913574,
      "num_tokens": 48879346.0,
      "step": 21900
    },
    {
      "entropy": 5.5088897228240965,
      "epoch": 2.460156118380412,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00044023959670825533,
      "loss": 5.3029,
      "mean_token_accuracy": 0.18948156833648683,
      "num_tokens": 48890939.0,
      "step": 21905
    },
    {
      "entropy": 5.495760679244995,
      "epoch": 2.46071769528837,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004402123443854894,
      "loss": 5.2963,
      "mean_token_accuracy": 0.19572442770004272,
      "num_tokens": 48902239.0,
      "step": 21910
    },
    {
      "entropy": 5.535546922683716,
      "epoch": 2.4612792721963275,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004401850868022984,
      "loss": 5.2571,
      "mean_token_accuracy": 0.19931843429803847,
      "num_tokens": 48913810.0,
      "step": 21915
    },
    {
      "entropy": 5.428094720840454,
      "epoch": 2.4618408491042847,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004401578239595501,
      "loss": 5.1343,
      "mean_token_accuracy": 0.2021462619304657,
      "num_tokens": 48925180.0,
      "step": 21920
    },
    {
      "entropy": 5.452257299423218,
      "epoch": 2.4624024260122424,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044013055585811274,
      "loss": 5.2308,
      "mean_token_accuracy": 0.1900040999054909,
      "num_tokens": 48937067.0,
      "step": 21925
    },
    {
      "entropy": 5.523362016677856,
      "epoch": 2.4629640029202,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044010328249885434,
      "loss": 5.2799,
      "mean_token_accuracy": 0.19059958010911943,
      "num_tokens": 48949220.0,
      "step": 21930
    },
    {
      "entropy": 5.5175337314605715,
      "epoch": 2.4635255798281577,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004400760038826434,
      "loss": 5.222,
      "mean_token_accuracy": 0.1970979243516922,
      "num_tokens": 48960604.0,
      "step": 21935
    },
    {
      "entropy": 5.439581394195557,
      "epoch": 2.464087156736115,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044004872001034856,
      "loss": 5.1474,
      "mean_token_accuracy": 0.20640051662921904,
      "num_tokens": 48970561.0,
      "step": 21940
    },
    {
      "entropy": 5.477488899230957,
      "epoch": 2.4646487336440726,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004400214308828384,
      "loss": 5.2323,
      "mean_token_accuracy": 0.19332142323255538,
      "num_tokens": 48981639.0,
      "step": 21945
    },
    {
      "entropy": 5.587163972854614,
      "epoch": 2.46521031055203,
      "grad_norm": 2.328125,
      "learning_rate": 0.00043999413650098193,
      "loss": 5.2597,
      "mean_token_accuracy": 0.19523866027593612,
      "num_tokens": 48992872.0,
      "step": 21950
    },
    {
      "entropy": 5.494712686538696,
      "epoch": 2.4657718874599874,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004399668368656481,
      "loss": 5.2396,
      "mean_token_accuracy": 0.19345155358314514,
      "num_tokens": 49004942.0,
      "step": 21955
    },
    {
      "entropy": 5.554857301712036,
      "epoch": 2.466333464367945,
      "grad_norm": 1.125,
      "learning_rate": 0.0004399395319777062,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19897417575120926,
      "num_tokens": 49016493.0,
      "step": 21960
    },
    {
      "entropy": 5.5130510330200195,
      "epoch": 2.4668950412759028,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043991222183802564,
      "loss": 5.2279,
      "mean_token_accuracy": 0.19936139434576033,
      "num_tokens": 49027780.0,
      "step": 21965
    },
    {
      "entropy": 5.462654781341553,
      "epoch": 2.4674566181838604,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000439884906447476,
      "loss": 5.2214,
      "mean_token_accuracy": 0.2021893158555031,
      "num_tokens": 49039441.0,
      "step": 21970
    },
    {
      "entropy": 5.486788558959961,
      "epoch": 2.4680181950918176,
      "grad_norm": 1.21875,
      "learning_rate": 0.000439857585806927,
      "loss": 5.2275,
      "mean_token_accuracy": 0.20156155675649642,
      "num_tokens": 49050740.0,
      "step": 21975
    },
    {
      "entropy": 5.44096884727478,
      "epoch": 2.4685797719997753,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043983025991724847,
      "loss": 5.1937,
      "mean_token_accuracy": 0.20292574316263198,
      "num_tokens": 49062216.0,
      "step": 21980
    },
    {
      "entropy": 5.4507551193237305,
      "epoch": 2.469141348907733,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004398029287793106,
      "loss": 5.1409,
      "mean_token_accuracy": 0.2044414147734642,
      "num_tokens": 49072848.0,
      "step": 21985
    },
    {
      "entropy": 5.417329263687134,
      "epoch": 2.4697029258156906,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043977559239398353,
      "loss": 5.113,
      "mean_token_accuracy": 0.20254876464605331,
      "num_tokens": 49083382.0,
      "step": 21990
    },
    {
      "entropy": 5.438059234619141,
      "epoch": 2.470264502723648,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004397482507621376,
      "loss": 5.157,
      "mean_token_accuracy": 0.20100490599870682,
      "num_tokens": 49095102.0,
      "step": 21995
    },
    {
      "entropy": 5.436920309066773,
      "epoch": 2.4708260796316055,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004397209038846436,
      "loss": 5.18,
      "mean_token_accuracy": 0.1987630844116211,
      "num_tokens": 49105964.0,
      "step": 22000
    },
    {
      "entropy": 5.499797201156616,
      "epoch": 2.471387656539563,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043969355176237207,
      "loss": 5.2966,
      "mean_token_accuracy": 0.18655095845460892,
      "num_tokens": 49117748.0,
      "step": 22005
    },
    {
      "entropy": 5.416675806045532,
      "epoch": 2.471949233447521,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004396661943961941,
      "loss": 5.038,
      "mean_token_accuracy": 0.20946522504091264,
      "num_tokens": 49128217.0,
      "step": 22010
    },
    {
      "entropy": 5.483416557312012,
      "epoch": 2.472510810355478,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004396388317869805,
      "loss": 5.3003,
      "mean_token_accuracy": 0.19545274823904038,
      "num_tokens": 49139272.0,
      "step": 22015
    },
    {
      "entropy": 5.467224025726319,
      "epoch": 2.4730723872634357,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004396114639356027,
      "loss": 5.1912,
      "mean_token_accuracy": 0.19524904191493989,
      "num_tokens": 49149570.0,
      "step": 22020
    },
    {
      "entropy": 5.424163293838501,
      "epoch": 2.4736339641713934,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000439584090842932,
      "loss": 5.1432,
      "mean_token_accuracy": 0.20249621123075484,
      "num_tokens": 49160892.0,
      "step": 22025
    },
    {
      "entropy": 5.456919622421265,
      "epoch": 2.474195541079351,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043955671250984005,
      "loss": 5.1554,
      "mean_token_accuracy": 0.19948485046625136,
      "num_tokens": 49172476.0,
      "step": 22030
    },
    {
      "entropy": 5.538638353347778,
      "epoch": 2.4747571179873082,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043952932893719847,
      "loss": 5.2515,
      "mean_token_accuracy": 0.20217922627925872,
      "num_tokens": 49184052.0,
      "step": 22035
    },
    {
      "entropy": 5.451108980178833,
      "epoch": 2.475318694895266,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004395019401258794,
      "loss": 5.2382,
      "mean_token_accuracy": 0.191593736410141,
      "num_tokens": 49195441.0,
      "step": 22040
    },
    {
      "entropy": 5.394437646865844,
      "epoch": 2.4758802718032236,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043947454607675464,
      "loss": 5.1473,
      "mean_token_accuracy": 0.20122386366128922,
      "num_tokens": 49206395.0,
      "step": 22045
    },
    {
      "entropy": 5.50679383277893,
      "epoch": 2.476441848711181,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004394471467906966,
      "loss": 5.2255,
      "mean_token_accuracy": 0.19376638978719712,
      "num_tokens": 49217705.0,
      "step": 22050
    },
    {
      "entropy": 5.476305437088013,
      "epoch": 2.4770034256191384,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043941974226857763,
      "loss": 5.0971,
      "mean_token_accuracy": 0.20508005619049072,
      "num_tokens": 49228348.0,
      "step": 22055
    },
    {
      "entropy": 5.401055812835693,
      "epoch": 2.477565002527096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004393923325112702,
      "loss": 5.1686,
      "mean_token_accuracy": 0.19700507819652557,
      "num_tokens": 49239031.0,
      "step": 22060
    },
    {
      "entropy": 5.494280433654785,
      "epoch": 2.4781265794350538,
      "grad_norm": 1.375,
      "learning_rate": 0.0004393649175196472,
      "loss": 5.2852,
      "mean_token_accuracy": 0.19698485136032104,
      "num_tokens": 49250918.0,
      "step": 22065
    },
    {
      "entropy": 5.4980357646942135,
      "epoch": 2.478688156343011,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004393374972945815,
      "loss": 5.2338,
      "mean_token_accuracy": 0.19626788049936295,
      "num_tokens": 49262305.0,
      "step": 22070
    },
    {
      "entropy": 5.432592868804932,
      "epoch": 2.4792497332509686,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043931007183694616,
      "loss": 5.1535,
      "mean_token_accuracy": 0.19390892535448073,
      "num_tokens": 49273106.0,
      "step": 22075
    },
    {
      "entropy": 5.387322902679443,
      "epoch": 2.4798113101589263,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043928264114761437,
      "loss": 5.1157,
      "mean_token_accuracy": 0.208780637383461,
      "num_tokens": 49284519.0,
      "step": 22080
    },
    {
      "entropy": 5.422154474258423,
      "epoch": 2.480372887066884,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043925520522745953,
      "loss": 5.1449,
      "mean_token_accuracy": 0.2027816042304039,
      "num_tokens": 49295753.0,
      "step": 22085
    },
    {
      "entropy": 5.475007152557373,
      "epoch": 2.480934463974841,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004392277640773553,
      "loss": 5.239,
      "mean_token_accuracy": 0.20320538878440858,
      "num_tokens": 49307682.0,
      "step": 22090
    },
    {
      "entropy": 5.445680761337281,
      "epoch": 2.481496040882799,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043920031769817543,
      "loss": 5.1785,
      "mean_token_accuracy": 0.19945676475763321,
      "num_tokens": 49318445.0,
      "step": 22095
    },
    {
      "entropy": 5.479609203338623,
      "epoch": 2.4820576177907565,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043917286609079364,
      "loss": 5.3012,
      "mean_token_accuracy": 0.19467789083719253,
      "num_tokens": 49328559.0,
      "step": 22100
    },
    {
      "entropy": 5.451708126068115,
      "epoch": 2.482619194698714,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043914540925608426,
      "loss": 5.1993,
      "mean_token_accuracy": 0.19196963608264922,
      "num_tokens": 49340177.0,
      "step": 22105
    },
    {
      "entropy": 5.479378271102905,
      "epoch": 2.4831807716066714,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004391179471949213,
      "loss": 5.1648,
      "mean_token_accuracy": 0.20082276165485383,
      "num_tokens": 49351214.0,
      "step": 22110
    },
    {
      "entropy": 5.456046485900879,
      "epoch": 2.483742348514629,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043909047990817917,
      "loss": 5.2031,
      "mean_token_accuracy": 0.20369723588228225,
      "num_tokens": 49362697.0,
      "step": 22115
    },
    {
      "entropy": 5.489879846572876,
      "epoch": 2.4843039254225867,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043906300739673265,
      "loss": 5.1976,
      "mean_token_accuracy": 0.19850983321666718,
      "num_tokens": 49374439.0,
      "step": 22120
    },
    {
      "entropy": 5.426640176773072,
      "epoch": 2.4848655023305444,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043903552966145624,
      "loss": 5.1042,
      "mean_token_accuracy": 0.19865205734968186,
      "num_tokens": 49385834.0,
      "step": 22125
    },
    {
      "entropy": 5.398649120330811,
      "epoch": 2.4854270792385016,
      "grad_norm": 1.140625,
      "learning_rate": 0.000439008046703225,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20927066951990128,
      "num_tokens": 49396562.0,
      "step": 22130
    },
    {
      "entropy": 5.3768110275268555,
      "epoch": 2.4859886561464593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000438980558522914,
      "loss": 5.1646,
      "mean_token_accuracy": 0.19566589295864106,
      "num_tokens": 49407339.0,
      "step": 22135
    },
    {
      "entropy": 5.489607906341552,
      "epoch": 2.486550233054417,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004389530651213983,
      "loss": 5.1722,
      "mean_token_accuracy": 0.20103365033864976,
      "num_tokens": 49419190.0,
      "step": 22140
    },
    {
      "entropy": 5.600993347167969,
      "epoch": 2.487111809962374,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043892556649955354,
      "loss": 5.3011,
      "mean_token_accuracy": 0.1875125840306282,
      "num_tokens": 49430559.0,
      "step": 22145
    },
    {
      "entropy": 5.476229190826416,
      "epoch": 2.487673386870332,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004388980626582551,
      "loss": 5.242,
      "mean_token_accuracy": 0.2010987728834152,
      "num_tokens": 49441862.0,
      "step": 22150
    },
    {
      "entropy": 5.444450426101684,
      "epoch": 2.4882349637782895,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004388705535983788,
      "loss": 5.175,
      "mean_token_accuracy": 0.19661868512630462,
      "num_tokens": 49452839.0,
      "step": 22155
    },
    {
      "entropy": 5.452764892578125,
      "epoch": 2.488796540686247,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004388430393208005,
      "loss": 5.1788,
      "mean_token_accuracy": 0.20446619242429734,
      "num_tokens": 49463683.0,
      "step": 22160
    },
    {
      "entropy": 5.446353578567505,
      "epoch": 2.4893581175942043,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004388155198263963,
      "loss": 5.1476,
      "mean_token_accuracy": 0.20625588297843933,
      "num_tokens": 49474338.0,
      "step": 22165
    },
    {
      "entropy": 5.427867126464844,
      "epoch": 2.489919694502162,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004387879951160424,
      "loss": 5.2375,
      "mean_token_accuracy": 0.19345259219408034,
      "num_tokens": 49485888.0,
      "step": 22170
    },
    {
      "entropy": 5.536109256744385,
      "epoch": 2.4904812714101197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004387604651906152,
      "loss": 5.2665,
      "mean_token_accuracy": 0.19629083275794984,
      "num_tokens": 49497270.0,
      "step": 22175
    },
    {
      "entropy": 5.575623607635498,
      "epoch": 2.4910428483180773,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004387329300509913,
      "loss": 5.2388,
      "mean_token_accuracy": 0.19408587366342545,
      "num_tokens": 49509099.0,
      "step": 22180
    },
    {
      "entropy": 5.409361982345581,
      "epoch": 2.4916044252260345,
      "grad_norm": 1.25,
      "learning_rate": 0.00043870538969804733,
      "loss": 5.1599,
      "mean_token_accuracy": 0.20205197781324385,
      "num_tokens": 49520110.0,
      "step": 22185
    },
    {
      "entropy": 5.396758556365967,
      "epoch": 2.492166002133992,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004386778441326603,
      "loss": 5.137,
      "mean_token_accuracy": 0.20351825505495072,
      "num_tokens": 49531866.0,
      "step": 22190
    },
    {
      "entropy": 5.498166799545288,
      "epoch": 2.49272757904195,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043865029335570725,
      "loss": 5.2137,
      "mean_token_accuracy": 0.19389698952436446,
      "num_tokens": 49541022.0,
      "step": 22195
    },
    {
      "entropy": 5.445477771759033,
      "epoch": 2.4932891559499075,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004386227373680655,
      "loss": 5.212,
      "mean_token_accuracy": 0.1959021046757698,
      "num_tokens": 49551908.0,
      "step": 22200
    },
    {
      "entropy": 5.492675018310547,
      "epoch": 2.4938507328578647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004385951761706122,
      "loss": 5.1593,
      "mean_token_accuracy": 0.1995590701699257,
      "num_tokens": 49562475.0,
      "step": 22205
    },
    {
      "entropy": 5.437918663024902,
      "epoch": 2.4944123097658224,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043856760976422506,
      "loss": 5.1178,
      "mean_token_accuracy": 0.203217613697052,
      "num_tokens": 49572916.0,
      "step": 22210
    },
    {
      "entropy": 5.459088373184204,
      "epoch": 2.49497388667378,
      "grad_norm": 1.25,
      "learning_rate": 0.00043854003814978174,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19507265239953994,
      "num_tokens": 49584584.0,
      "step": 22215
    },
    {
      "entropy": 5.373010683059692,
      "epoch": 2.4955354635817377,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043851246132816025,
      "loss": 5.1991,
      "mean_token_accuracy": 0.19947243630886077,
      "num_tokens": 49595792.0,
      "step": 22220
    },
    {
      "entropy": 5.495765686035156,
      "epoch": 2.496097040489695,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004384848793002385,
      "loss": 5.2434,
      "mean_token_accuracy": 0.1998839020729065,
      "num_tokens": 49607596.0,
      "step": 22225
    },
    {
      "entropy": 5.537977695465088,
      "epoch": 2.4966586173976526,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004384572920668949,
      "loss": 5.1741,
      "mean_token_accuracy": 0.19015215188264847,
      "num_tokens": 49617990.0,
      "step": 22230
    },
    {
      "entropy": 5.540008544921875,
      "epoch": 2.4972201943056103,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004384296996290076,
      "loss": 5.2626,
      "mean_token_accuracy": 0.19752211570739747,
      "num_tokens": 49629458.0,
      "step": 22235
    },
    {
      "entropy": 5.544511508941651,
      "epoch": 2.4977817712135675,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043840210198745533,
      "loss": 5.2221,
      "mean_token_accuracy": 0.20264557451009751,
      "num_tokens": 49641227.0,
      "step": 22240
    },
    {
      "entropy": 5.440906286239624,
      "epoch": 2.498343348121525,
      "grad_norm": 1.125,
      "learning_rate": 0.00043837449914311676,
      "loss": 5.2815,
      "mean_token_accuracy": 0.19710609763860704,
      "num_tokens": 49652655.0,
      "step": 22245
    },
    {
      "entropy": 5.441756629943848,
      "epoch": 2.498904925029483,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043834689109687077,
      "loss": 5.2406,
      "mean_token_accuracy": 0.2054051637649536,
      "num_tokens": 49664308.0,
      "step": 22250
    },
    {
      "entropy": 5.509242963790894,
      "epoch": 2.4994665019374405,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004383192778495964,
      "loss": 5.1792,
      "mean_token_accuracy": 0.1972736269235611,
      "num_tokens": 49676656.0,
      "step": 22255
    },
    {
      "entropy": 5.466150522232056,
      "epoch": 2.500028078845398,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000438291659402173,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19730150699615479,
      "num_tokens": 49687706.0,
      "step": 22260
    },
    {
      "entropy": 5.483819627761841,
      "epoch": 2.5005896557533553,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043826403575547974,
      "loss": 5.1724,
      "mean_token_accuracy": 0.2053607791662216,
      "num_tokens": 49699139.0,
      "step": 22265
    },
    {
      "entropy": 5.318334102630615,
      "epoch": 2.501151232661313,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004382364069103962,
      "loss": 5.1748,
      "mean_token_accuracy": 0.20242644846439362,
      "num_tokens": 49710690.0,
      "step": 22270
    },
    {
      "entropy": 5.437481832504273,
      "epoch": 2.5017128095692707,
      "grad_norm": 1.125,
      "learning_rate": 0.00043820877286780224,
      "loss": 5.2127,
      "mean_token_accuracy": 0.20191237777471543,
      "num_tokens": 49721920.0,
      "step": 22275
    },
    {
      "entropy": 5.555009126663208,
      "epoch": 2.502274386477228,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004381811336285776,
      "loss": 5.2056,
      "mean_token_accuracy": 0.19992334693670272,
      "num_tokens": 49732923.0,
      "step": 22280
    },
    {
      "entropy": 5.366896629333496,
      "epoch": 2.5028359633851855,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004381534891936024,
      "loss": 5.089,
      "mean_token_accuracy": 0.20133124887943268,
      "num_tokens": 49744755.0,
      "step": 22285
    },
    {
      "entropy": 5.423555660247803,
      "epoch": 2.503397540293143,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004381258395637568,
      "loss": 5.1612,
      "mean_token_accuracy": 0.20267312824726105,
      "num_tokens": 49756126.0,
      "step": 22290
    },
    {
      "entropy": 5.458829307556153,
      "epoch": 2.5039591172011004,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004380981847399212,
      "loss": 5.1599,
      "mean_token_accuracy": 0.20250371396541594,
      "num_tokens": 49767137.0,
      "step": 22295
    },
    {
      "entropy": 5.454941463470459,
      "epoch": 2.504520694109058,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043807052472297626,
      "loss": 5.2492,
      "mean_token_accuracy": 0.1909842610359192,
      "num_tokens": 49778533.0,
      "step": 22300
    },
    {
      "entropy": 5.4665198802948,
      "epoch": 2.5050822710170157,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004380428595138024,
      "loss": 5.1992,
      "mean_token_accuracy": 0.20529264956712723,
      "num_tokens": 49789494.0,
      "step": 22305
    },
    {
      "entropy": 5.496833467483521,
      "epoch": 2.5056438479249734,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043801518911328074,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19738632589578628,
      "num_tokens": 49799546.0,
      "step": 22310
    },
    {
      "entropy": 5.388626050949097,
      "epoch": 2.506205424832931,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043798751352229227,
      "loss": 5.099,
      "mean_token_accuracy": 0.20380640029907227,
      "num_tokens": 49810643.0,
      "step": 22315
    },
    {
      "entropy": 5.487656784057617,
      "epoch": 2.5067670017408883,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043795983274171806,
      "loss": 5.2221,
      "mean_token_accuracy": 0.1948302373290062,
      "num_tokens": 49821410.0,
      "step": 22320
    },
    {
      "entropy": 5.460285186767578,
      "epoch": 2.507328578648846,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043793214677243967,
      "loss": 5.2431,
      "mean_token_accuracy": 0.19352083653211594,
      "num_tokens": 49832911.0,
      "step": 22325
    },
    {
      "entropy": 5.508481025695801,
      "epoch": 2.5078901555568036,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004379044556153384,
      "loss": 5.2308,
      "mean_token_accuracy": 0.19389263391494752,
      "num_tokens": 49845458.0,
      "step": 22330
    },
    {
      "entropy": 5.5135880470275875,
      "epoch": 2.508451732464761,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043787675927129625,
      "loss": 5.3314,
      "mean_token_accuracy": 0.18994753211736679,
      "num_tokens": 49857054.0,
      "step": 22335
    },
    {
      "entropy": 5.456147384643555,
      "epoch": 2.5090133093727185,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043784905774119483,
      "loss": 5.1243,
      "mean_token_accuracy": 0.2006603807210922,
      "num_tokens": 49867223.0,
      "step": 22340
    },
    {
      "entropy": 5.401931667327881,
      "epoch": 2.509574886280676,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043782135102591626,
      "loss": 5.1865,
      "mean_token_accuracy": 0.1990416169166565,
      "num_tokens": 49877801.0,
      "step": 22345
    },
    {
      "entropy": 5.417321872711182,
      "epoch": 2.510136463188634,
      "grad_norm": 1.125,
      "learning_rate": 0.0004377936391263427,
      "loss": 5.1859,
      "mean_token_accuracy": 0.19792087376117706,
      "num_tokens": 49888853.0,
      "step": 22350
    },
    {
      "entropy": 5.473415946960449,
      "epoch": 2.5106980400965915,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043776592204335655,
      "loss": 5.1834,
      "mean_token_accuracy": 0.20324978977441788,
      "num_tokens": 49899989.0,
      "step": 22355
    },
    {
      "entropy": 5.481160116195679,
      "epoch": 2.5112596170045487,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004377381997778403,
      "loss": 5.1571,
      "mean_token_accuracy": 0.20386865437030793,
      "num_tokens": 49910993.0,
      "step": 22360
    },
    {
      "entropy": 5.478058338165283,
      "epoch": 2.5118211939125064,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004377104723306767,
      "loss": 5.2194,
      "mean_token_accuracy": 0.19816150814294814,
      "num_tokens": 49922169.0,
      "step": 22365
    },
    {
      "entropy": 5.466850137710571,
      "epoch": 2.512382770820464,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004376827397027485,
      "loss": 5.2746,
      "mean_token_accuracy": 0.1932577133178711,
      "num_tokens": 49933700.0,
      "step": 22370
    },
    {
      "entropy": 5.500583648681641,
      "epoch": 2.5129443477284212,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004376550018949388,
      "loss": 5.197,
      "mean_token_accuracy": 0.19715379923582077,
      "num_tokens": 49944689.0,
      "step": 22375
    },
    {
      "entropy": 5.479571199417114,
      "epoch": 2.513505924636379,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043762725890813083,
      "loss": 5.2258,
      "mean_token_accuracy": 0.19828970432281495,
      "num_tokens": 49955818.0,
      "step": 22380
    },
    {
      "entropy": 5.396988105773926,
      "epoch": 2.5140675015443366,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004375995107432077,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20480271130800248,
      "num_tokens": 49965799.0,
      "step": 22385
    },
    {
      "entropy": 5.441124486923218,
      "epoch": 2.5146290784522938,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043757175740105316,
      "loss": 5.1998,
      "mean_token_accuracy": 0.19720589220523835,
      "num_tokens": 49977742.0,
      "step": 22390
    },
    {
      "entropy": 5.502358627319336,
      "epoch": 2.5151906553602514,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004375439988825508,
      "loss": 5.1811,
      "mean_token_accuracy": 0.1966991201043129,
      "num_tokens": 49988382.0,
      "step": 22395
    },
    {
      "entropy": 5.50432333946228,
      "epoch": 2.515752232268209,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004375162351885845,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19249147176742554,
      "num_tokens": 50000190.0,
      "step": 22400
    },
    {
      "entropy": 5.528553104400634,
      "epoch": 2.5163138091761668,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043748846632003823,
      "loss": 5.2138,
      "mean_token_accuracy": 0.20334574729204177,
      "num_tokens": 50011297.0,
      "step": 22405
    },
    {
      "entropy": 5.4859010696411135,
      "epoch": 2.5168753860841244,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043746069227779615,
      "loss": 5.2268,
      "mean_token_accuracy": 0.1949320390820503,
      "num_tokens": 50023165.0,
      "step": 22410
    },
    {
      "entropy": 5.510264921188354,
      "epoch": 2.5174369629920816,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004374329130627427,
      "loss": 5.234,
      "mean_token_accuracy": 0.19958595037460328,
      "num_tokens": 50035803.0,
      "step": 22415
    },
    {
      "entropy": 5.445142030715942,
      "epoch": 2.5179985399000393,
      "grad_norm": 1.125,
      "learning_rate": 0.00043740512867576224,
      "loss": 5.1186,
      "mean_token_accuracy": 0.20306495279073716,
      "num_tokens": 50046599.0,
      "step": 22420
    },
    {
      "entropy": 5.471498155593872,
      "epoch": 2.518560116807997,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004373773391177395,
      "loss": 5.3363,
      "mean_token_accuracy": 0.18993393331766129,
      "num_tokens": 50058170.0,
      "step": 22425
    },
    {
      "entropy": 5.440329074859619,
      "epoch": 2.519121693715954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004373495443895593,
      "loss": 5.1424,
      "mean_token_accuracy": 0.20907102674245834,
      "num_tokens": 50068818.0,
      "step": 22430
    },
    {
      "entropy": 5.506471872329712,
      "epoch": 2.519683270623912,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004373217444921066,
      "loss": 5.2896,
      "mean_token_accuracy": 0.19338531494140626,
      "num_tokens": 50080533.0,
      "step": 22435
    },
    {
      "entropy": 5.434464168548584,
      "epoch": 2.5202448475318695,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043729393942626665,
      "loss": 5.1146,
      "mean_token_accuracy": 0.20435652881860733,
      "num_tokens": 50090897.0,
      "step": 22440
    },
    {
      "entropy": 5.371607303619385,
      "epoch": 2.520806424439827,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043726612919292476,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20662710517644883,
      "num_tokens": 50101594.0,
      "step": 22445
    },
    {
      "entropy": 5.419979047775269,
      "epoch": 2.521368001347785,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004372383137929663,
      "loss": 5.1598,
      "mean_token_accuracy": 0.20202449411153794,
      "num_tokens": 50111923.0,
      "step": 22450
    },
    {
      "entropy": 5.500240087509155,
      "epoch": 2.521929578255742,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043721049322727705,
      "loss": 5.2375,
      "mean_token_accuracy": 0.2016470327973366,
      "num_tokens": 50123295.0,
      "step": 22455
    },
    {
      "entropy": 5.538443660736084,
      "epoch": 2.5224911551636997,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004371826674967428,
      "loss": 5.2995,
      "mean_token_accuracy": 0.19865900725126268,
      "num_tokens": 50133245.0,
      "step": 22460
    },
    {
      "entropy": 5.416790199279785,
      "epoch": 2.5230527320716574,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043715483660224943,
      "loss": 5.1531,
      "mean_token_accuracy": 0.2004664197564125,
      "num_tokens": 50144827.0,
      "step": 22465
    },
    {
      "entropy": 5.56309289932251,
      "epoch": 2.5236143089796146,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004371270005446833,
      "loss": 5.2798,
      "mean_token_accuracy": 0.19326076358556749,
      "num_tokens": 50156244.0,
      "step": 22470
    },
    {
      "entropy": 5.544798040390015,
      "epoch": 2.5241758858875722,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004370991593249305,
      "loss": 5.2527,
      "mean_token_accuracy": 0.19526829123497008,
      "num_tokens": 50168363.0,
      "step": 22475
    },
    {
      "entropy": 5.50516448020935,
      "epoch": 2.52473746279553,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043707131294387766,
      "loss": 5.1814,
      "mean_token_accuracy": 0.1983770489692688,
      "num_tokens": 50179370.0,
      "step": 22480
    },
    {
      "entropy": 5.5335962772369385,
      "epoch": 2.525299039703487,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004370434614024113,
      "loss": 5.2165,
      "mean_token_accuracy": 0.2053406596183777,
      "num_tokens": 50191583.0,
      "step": 22485
    },
    {
      "entropy": 5.489752197265625,
      "epoch": 2.525860616611445,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043701560470141824,
      "loss": 5.1718,
      "mean_token_accuracy": 0.1955375775694847,
      "num_tokens": 50202008.0,
      "step": 22490
    },
    {
      "entropy": 5.420732212066651,
      "epoch": 2.5264221935194024,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004369877428417856,
      "loss": 5.226,
      "mean_token_accuracy": 0.19432284235954284,
      "num_tokens": 50212632.0,
      "step": 22495
    },
    {
      "entropy": 5.474782037734985,
      "epoch": 2.52698377042736,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004369598758244003,
      "loss": 5.2266,
      "mean_token_accuracy": 0.20370943248271942,
      "num_tokens": 50223485.0,
      "step": 22500
    },
    {
      "entropy": 5.567892932891846,
      "epoch": 2.5275453473353178,
      "grad_norm": 1.125,
      "learning_rate": 0.00043693200365014985,
      "loss": 5.2945,
      "mean_token_accuracy": 0.19657267183065413,
      "num_tokens": 50235516.0,
      "step": 22505
    },
    {
      "entropy": 5.475609064102173,
      "epoch": 2.528106924243275,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043690412631992154,
      "loss": 5.1424,
      "mean_token_accuracy": 0.20403032153844833,
      "num_tokens": 50246846.0,
      "step": 22510
    },
    {
      "entropy": 5.420966577529907,
      "epoch": 2.5286685011512327,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043687624383460306,
      "loss": 5.219,
      "mean_token_accuracy": 0.1896538943052292,
      "num_tokens": 50258433.0,
      "step": 22515
    },
    {
      "entropy": 5.553134536743164,
      "epoch": 2.5292300780591903,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043684835619508215,
      "loss": 5.3207,
      "mean_token_accuracy": 0.1976487845182419,
      "num_tokens": 50269111.0,
      "step": 22520
    },
    {
      "entropy": 5.492173671722412,
      "epoch": 2.5297916549671475,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004368204634022469,
      "loss": 5.3164,
      "mean_token_accuracy": 0.19301936626434327,
      "num_tokens": 50280137.0,
      "step": 22525
    },
    {
      "entropy": 5.49051604270935,
      "epoch": 2.530353231875105,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004367925654569852,
      "loss": 5.1619,
      "mean_token_accuracy": 0.20840794444084168,
      "num_tokens": 50291717.0,
      "step": 22530
    },
    {
      "entropy": 5.510760068893433,
      "epoch": 2.530914808783063,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004367646623601856,
      "loss": 5.246,
      "mean_token_accuracy": 0.19451362788677215,
      "num_tokens": 50303107.0,
      "step": 22535
    },
    {
      "entropy": 5.374897003173828,
      "epoch": 2.5314763856910205,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004367367541127364,
      "loss": 5.16,
      "mean_token_accuracy": 0.20004209876060486,
      "num_tokens": 50313997.0,
      "step": 22540
    },
    {
      "entropy": 5.437381267547607,
      "epoch": 2.532037962598978,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004367088407155261,
      "loss": 5.1711,
      "mean_token_accuracy": 0.20653210282325746,
      "num_tokens": 50324327.0,
      "step": 22545
    },
    {
      "entropy": 5.412775897979737,
      "epoch": 2.5325995395069354,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004366809221694437,
      "loss": 5.1087,
      "mean_token_accuracy": 0.20845651477575303,
      "num_tokens": 50334787.0,
      "step": 22550
    },
    {
      "entropy": 5.393438243865967,
      "epoch": 2.533161116414893,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043665299847537805,
      "loss": 5.0917,
      "mean_token_accuracy": 0.20833459943532945,
      "num_tokens": 50345418.0,
      "step": 22555
    },
    {
      "entropy": 5.4462038516998295,
      "epoch": 2.5337226933228507,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043662506963421815,
      "loss": 5.21,
      "mean_token_accuracy": 0.2001356914639473,
      "num_tokens": 50357047.0,
      "step": 22560
    },
    {
      "entropy": 5.437898397445679,
      "epoch": 2.534284270230808,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004365971356468534,
      "loss": 5.1666,
      "mean_token_accuracy": 0.2075835272669792,
      "num_tokens": 50368385.0,
      "step": 22565
    },
    {
      "entropy": 5.465651798248291,
      "epoch": 2.5348458471387656,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043656919651417314,
      "loss": 5.2319,
      "mean_token_accuracy": 0.2005552276968956,
      "num_tokens": 50379389.0,
      "step": 22570
    },
    {
      "entropy": 5.500934362411499,
      "epoch": 2.5354074240467233,
      "grad_norm": 1.15625,
      "learning_rate": 0.000436541252237067,
      "loss": 5.1982,
      "mean_token_accuracy": 0.20028745979070664,
      "num_tokens": 50390124.0,
      "step": 22575
    },
    {
      "entropy": 5.495947074890137,
      "epoch": 2.5359690009546805,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004365133028164248,
      "loss": 5.212,
      "mean_token_accuracy": 0.20381075590848924,
      "num_tokens": 50401385.0,
      "step": 22580
    },
    {
      "entropy": 5.4653639793396,
      "epoch": 2.536530577862638,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004364853482531363,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19280862361192702,
      "num_tokens": 50412890.0,
      "step": 22585
    },
    {
      "entropy": 5.487146472930908,
      "epoch": 2.537092154770596,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043645738854809177,
      "loss": 5.2635,
      "mean_token_accuracy": 0.20081137865781784,
      "num_tokens": 50424403.0,
      "step": 22590
    },
    {
      "entropy": 5.578891134262085,
      "epoch": 2.5376537316785535,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004364294237021813,
      "loss": 5.2179,
      "mean_token_accuracy": 0.19921068251132965,
      "num_tokens": 50434598.0,
      "step": 22595
    },
    {
      "entropy": 5.575385189056396,
      "epoch": 2.538215308586511,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043640145371629544,
      "loss": 5.2863,
      "mean_token_accuracy": 0.19035972505807877,
      "num_tokens": 50446661.0,
      "step": 22600
    },
    {
      "entropy": 5.566793584823609,
      "epoch": 2.5387768854944683,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004363734785913246,
      "loss": 5.2841,
      "mean_token_accuracy": 0.19049828946590425,
      "num_tokens": 50458188.0,
      "step": 22605
    },
    {
      "entropy": 5.502074956893921,
      "epoch": 2.539338462402426,
      "grad_norm": 1.125,
      "learning_rate": 0.0004363454983281597,
      "loss": 5.2551,
      "mean_token_accuracy": 0.20060526132583617,
      "num_tokens": 50469305.0,
      "step": 22610
    },
    {
      "entropy": 5.516038227081299,
      "epoch": 2.5399000393103837,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004363175129276915,
      "loss": 5.2024,
      "mean_token_accuracy": 0.20006683319807053,
      "num_tokens": 50480450.0,
      "step": 22615
    },
    {
      "entropy": 5.439752578735352,
      "epoch": 2.540461616218341,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043628952239081113,
      "loss": 5.1684,
      "mean_token_accuracy": 0.19881519228219985,
      "num_tokens": 50490397.0,
      "step": 22620
    },
    {
      "entropy": 5.377489328384399,
      "epoch": 2.5410231931262985,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043626152671840984,
      "loss": 5.1397,
      "mean_token_accuracy": 0.20063460767269134,
      "num_tokens": 50501068.0,
      "step": 22625
    },
    {
      "entropy": 5.532137823104859,
      "epoch": 2.541584770034256,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000436233525911379,
      "loss": 5.4097,
      "mean_token_accuracy": 0.1885458707809448,
      "num_tokens": 50513268.0,
      "step": 22630
    },
    {
      "entropy": 5.467395544052124,
      "epoch": 2.542146346942214,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004362055199706101,
      "loss": 5.1054,
      "mean_token_accuracy": 0.20370217710733413,
      "num_tokens": 50523806.0,
      "step": 22635
    },
    {
      "entropy": 5.480954504013061,
      "epoch": 2.5427079238501715,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043617750889699496,
      "loss": 5.1612,
      "mean_token_accuracy": 0.1982889235019684,
      "num_tokens": 50535000.0,
      "step": 22640
    },
    {
      "entropy": 5.490712070465088,
      "epoch": 2.5432695007581287,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043614949269142546,
      "loss": 5.2655,
      "mean_token_accuracy": 0.19457219839096068,
      "num_tokens": 50546042.0,
      "step": 22645
    },
    {
      "entropy": 5.574697399139405,
      "epoch": 2.5438310776660864,
      "grad_norm": 1.125,
      "learning_rate": 0.00043612147135479356,
      "loss": 5.2525,
      "mean_token_accuracy": 0.19598281979560853,
      "num_tokens": 50556645.0,
      "step": 22650
    },
    {
      "entropy": 5.558626699447632,
      "epoch": 2.544392654574044,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004360934448879916,
      "loss": 5.2675,
      "mean_token_accuracy": 0.1959284171462059,
      "num_tokens": 50567904.0,
      "step": 22655
    },
    {
      "entropy": 5.389188003540039,
      "epoch": 2.5449542314820013,
      "grad_norm": 1.125,
      "learning_rate": 0.0004360654132919118,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20017922520637513,
      "num_tokens": 50578280.0,
      "step": 22660
    },
    {
      "entropy": 5.408729982376099,
      "epoch": 2.545515808389959,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004360373765674467,
      "loss": 5.1929,
      "mean_token_accuracy": 0.20137256681919097,
      "num_tokens": 50590232.0,
      "step": 22665
    },
    {
      "entropy": 5.453761339187622,
      "epoch": 2.5460773852979166,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043600933471548927,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20683782547712326,
      "num_tokens": 50601469.0,
      "step": 22670
    },
    {
      "entropy": 5.401768207550049,
      "epoch": 2.546638962205874,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004359812877369321,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20404326915740967,
      "num_tokens": 50612090.0,
      "step": 22675
    },
    {
      "entropy": 5.395300674438476,
      "epoch": 2.5472005391138315,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043595323563266815,
      "loss": 5.1831,
      "mean_token_accuracy": 0.19551308006048201,
      "num_tokens": 50623834.0,
      "step": 22680
    },
    {
      "entropy": 5.485017538070679,
      "epoch": 2.547762116021789,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043592517840359093,
      "loss": 5.2384,
      "mean_token_accuracy": 0.1948726862668991,
      "num_tokens": 50633966.0,
      "step": 22685
    },
    {
      "entropy": 5.438422155380249,
      "epoch": 2.548323692929747,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004358971160505936,
      "loss": 5.0889,
      "mean_token_accuracy": 0.20339812934398652,
      "num_tokens": 50644354.0,
      "step": 22690
    },
    {
      "entropy": 5.429444885253906,
      "epoch": 2.5488852698377045,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043586904857456963,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20113642066717147,
      "num_tokens": 50655525.0,
      "step": 22695
    },
    {
      "entropy": 5.491053676605224,
      "epoch": 2.5494468467456617,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043584097597641275,
      "loss": 5.3234,
      "mean_token_accuracy": 0.19047256857156752,
      "num_tokens": 50667879.0,
      "step": 22700
    },
    {
      "entropy": 5.393489217758178,
      "epoch": 2.5500084236536193,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043581289825701683,
      "loss": 5.0668,
      "mean_token_accuracy": 0.20350636243820192,
      "num_tokens": 50679568.0,
      "step": 22705
    },
    {
      "entropy": 5.448760032653809,
      "epoch": 2.550570000561577,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004357848154172759,
      "loss": 5.2297,
      "mean_token_accuracy": 0.20112274587154388,
      "num_tokens": 50690172.0,
      "step": 22710
    },
    {
      "entropy": 5.433311319351196,
      "epoch": 2.5511315774695342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043575672745808404,
      "loss": 5.1244,
      "mean_token_accuracy": 0.20036081075668336,
      "num_tokens": 50701150.0,
      "step": 22715
    },
    {
      "entropy": 5.471509742736816,
      "epoch": 2.551693154377492,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043572863438033556,
      "loss": 5.2306,
      "mean_token_accuracy": 0.20599816590547562,
      "num_tokens": 50712407.0,
      "step": 22720
    },
    {
      "entropy": 5.449364757537841,
      "epoch": 2.5522547312854496,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004357005361849251,
      "loss": 5.1526,
      "mean_token_accuracy": 0.19876994341611862,
      "num_tokens": 50723122.0,
      "step": 22725
    },
    {
      "entropy": 5.477801656723022,
      "epoch": 2.552816308193407,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043567243287274726,
      "loss": 5.2063,
      "mean_token_accuracy": 0.19595978260040284,
      "num_tokens": 50734479.0,
      "step": 22730
    },
    {
      "entropy": 5.380903196334839,
      "epoch": 2.553377885101365,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043564432444469686,
      "loss": 5.168,
      "mean_token_accuracy": 0.20650181621313096,
      "num_tokens": 50745098.0,
      "step": 22735
    },
    {
      "entropy": 5.489142274856567,
      "epoch": 2.553939462009322,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043561621090166885,
      "loss": 5.2122,
      "mean_token_accuracy": 0.1987846463918686,
      "num_tokens": 50755793.0,
      "step": 22740
    },
    {
      "entropy": 5.4806722640991214,
      "epoch": 2.5545010389172798,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004355880922445584,
      "loss": 5.2219,
      "mean_token_accuracy": 0.19182201474905014,
      "num_tokens": 50767685.0,
      "step": 22745
    },
    {
      "entropy": 5.517179250717163,
      "epoch": 2.5550626158252374,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00043555996847426077,
      "loss": 5.2962,
      "mean_token_accuracy": 0.19072840809822084,
      "num_tokens": 50778583.0,
      "step": 22750
    },
    {
      "entropy": 5.492299604415893,
      "epoch": 2.5556241927331946,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043553183959167154,
      "loss": 5.1876,
      "mean_token_accuracy": 0.2016289323568344,
      "num_tokens": 50789651.0,
      "step": 22755
    },
    {
      "entropy": 5.546571874618531,
      "epoch": 2.5561857696411523,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043550370559768614,
      "loss": 5.3283,
      "mean_token_accuracy": 0.19424901008605958,
      "num_tokens": 50800846.0,
      "step": 22760
    },
    {
      "entropy": 5.488478755950927,
      "epoch": 2.55674734654911,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004354755664932006,
      "loss": 5.0754,
      "mean_token_accuracy": 0.2037645012140274,
      "num_tokens": 50812532.0,
      "step": 22765
    },
    {
      "entropy": 5.517052888870239,
      "epoch": 2.557308923457067,
      "grad_norm": 1.125,
      "learning_rate": 0.00043544742227911076,
      "loss": 5.235,
      "mean_token_accuracy": 0.19830751568078994,
      "num_tokens": 50823592.0,
      "step": 22770
    },
    {
      "entropy": 5.446296691894531,
      "epoch": 2.557870500365025,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043541927295631276,
      "loss": 5.1188,
      "mean_token_accuracy": 0.2011640876531601,
      "num_tokens": 50834392.0,
      "step": 22775
    },
    {
      "entropy": 5.541642093658448,
      "epoch": 2.5584320772729825,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043539111852570296,
      "loss": 5.2677,
      "mean_token_accuracy": 0.19484104365110397,
      "num_tokens": 50845371.0,
      "step": 22780
    },
    {
      "entropy": 5.506129312515259,
      "epoch": 2.55899365418094,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043536295898817767,
      "loss": 5.2706,
      "mean_token_accuracy": 0.1952488377690315,
      "num_tokens": 50857748.0,
      "step": 22785
    },
    {
      "entropy": 5.424519681930542,
      "epoch": 2.559555231088898,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004353347943446336,
      "loss": 5.0747,
      "mean_token_accuracy": 0.2082340955734253,
      "num_tokens": 50868646.0,
      "step": 22790
    },
    {
      "entropy": 5.4368438720703125,
      "epoch": 2.560116807996855,
      "grad_norm": 1.0,
      "learning_rate": 0.00043530662459596744,
      "loss": 5.1143,
      "mean_token_accuracy": 0.204285891354084,
      "num_tokens": 50880566.0,
      "step": 22795
    },
    {
      "entropy": 5.462652015686035,
      "epoch": 2.5606783849048127,
      "grad_norm": 1.125,
      "learning_rate": 0.0004352784497430762,
      "loss": 5.2017,
      "mean_token_accuracy": 0.2011650323867798,
      "num_tokens": 50891487.0,
      "step": 22800
    },
    {
      "entropy": 5.471800041198731,
      "epoch": 2.5612399618127704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043525026978685706,
      "loss": 5.245,
      "mean_token_accuracy": 0.19139301478862764,
      "num_tokens": 50902522.0,
      "step": 22805
    },
    {
      "entropy": 5.590824174880981,
      "epoch": 2.5618015387207276,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043522208472820706,
      "loss": 5.2266,
      "mean_token_accuracy": 0.19290973246097565,
      "num_tokens": 50913363.0,
      "step": 22810
    },
    {
      "entropy": 5.479837369918823,
      "epoch": 2.5623631156286852,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004351938945680238,
      "loss": 5.2048,
      "mean_token_accuracy": 0.20074327141046525,
      "num_tokens": 50923724.0,
      "step": 22815
    },
    {
      "entropy": 5.470147228240966,
      "epoch": 2.562924692536643,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004351656993072048,
      "loss": 5.1929,
      "mean_token_accuracy": 0.20039606094360352,
      "num_tokens": 50933575.0,
      "step": 22820
    },
    {
      "entropy": 5.43289155960083,
      "epoch": 2.5634862694446006,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004351374989466477,
      "loss": 5.1707,
      "mean_token_accuracy": 0.2038708046078682,
      "num_tokens": 50944619.0,
      "step": 22825
    },
    {
      "entropy": 5.409995031356812,
      "epoch": 2.5640478463525582,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043510929348725065,
      "loss": 5.179,
      "mean_token_accuracy": 0.20432940125465393,
      "num_tokens": 50955701.0,
      "step": 22830
    },
    {
      "entropy": 5.415213918685913,
      "epoch": 2.5646094232605154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043508108292991156,
      "loss": 5.043,
      "mean_token_accuracy": 0.20145202428102493,
      "num_tokens": 50966544.0,
      "step": 22835
    },
    {
      "entropy": 5.467957401275635,
      "epoch": 2.565171000168473,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004350528672755287,
      "loss": 5.2507,
      "mean_token_accuracy": 0.20001019388437272,
      "num_tokens": 50976648.0,
      "step": 22840
    },
    {
      "entropy": 5.550426721572876,
      "epoch": 2.5657325770764308,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004350246465250005,
      "loss": 5.2135,
      "mean_token_accuracy": 0.19686073064804077,
      "num_tokens": 50988173.0,
      "step": 22845
    },
    {
      "entropy": 5.446621656417847,
      "epoch": 2.566294153984388,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043499642067922544,
      "loss": 5.079,
      "mean_token_accuracy": 0.20425282418727875,
      "num_tokens": 50999274.0,
      "step": 22850
    },
    {
      "entropy": 5.344490814208984,
      "epoch": 2.5668557308923456,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043496818973910235,
      "loss": 5.1445,
      "mean_token_accuracy": 0.20603708922863007,
      "num_tokens": 51010213.0,
      "step": 22855
    },
    {
      "entropy": 5.440465784072876,
      "epoch": 2.5674173078003033,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043493995370553,
      "loss": 5.2107,
      "mean_token_accuracy": 0.20147900581359862,
      "num_tokens": 51021460.0,
      "step": 22860
    },
    {
      "entropy": 5.586348199844361,
      "epoch": 2.5679788847082605,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043491171257940755,
      "loss": 5.2025,
      "mean_token_accuracy": 0.19831876158714296,
      "num_tokens": 51031713.0,
      "step": 22865
    },
    {
      "entropy": 5.510029125213623,
      "epoch": 2.568540461616218,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043488346636163405,
      "loss": 5.2917,
      "mean_token_accuracy": 0.19384196400642395,
      "num_tokens": 51043026.0,
      "step": 22870
    },
    {
      "entropy": 5.571712589263916,
      "epoch": 2.569102038524176,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004348552150531091,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19675661921501159,
      "num_tokens": 51054838.0,
      "step": 22875
    },
    {
      "entropy": 5.467933416366577,
      "epoch": 2.5696636154321335,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043482695865473207,
      "loss": 5.0715,
      "mean_token_accuracy": 0.20232197344303132,
      "num_tokens": 51065574.0,
      "step": 22880
    },
    {
      "entropy": 5.409094667434692,
      "epoch": 2.570225192340091,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004347986971674026,
      "loss": 5.1494,
      "mean_token_accuracy": 0.20092492699623107,
      "num_tokens": 51076814.0,
      "step": 22885
    },
    {
      "entropy": 5.415322494506836,
      "epoch": 2.5707867692480484,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043477043059202077,
      "loss": 5.2407,
      "mean_token_accuracy": 0.19303952008485795,
      "num_tokens": 51089075.0,
      "step": 22890
    },
    {
      "entropy": 5.505451917648315,
      "epoch": 2.571348346156006,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004347421589294864,
      "loss": 5.2667,
      "mean_token_accuracy": 0.1963049903512001,
      "num_tokens": 51100148.0,
      "step": 22895
    },
    {
      "entropy": 5.574727296829224,
      "epoch": 2.5719099230639637,
      "grad_norm": 1.25,
      "learning_rate": 0.0004347138821806997,
      "loss": 5.3076,
      "mean_token_accuracy": 0.1986745685338974,
      "num_tokens": 51111148.0,
      "step": 22900
    },
    {
      "entropy": 5.625747489929199,
      "epoch": 2.572471499971921,
      "grad_norm": 1.140625,
      "learning_rate": 0.000434685600346561,
      "loss": 5.2892,
      "mean_token_accuracy": 0.19100103825330733,
      "num_tokens": 51122563.0,
      "step": 22905
    },
    {
      "entropy": 5.407369470596313,
      "epoch": 2.5730330768798786,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000434657313427971,
      "loss": 5.1609,
      "mean_token_accuracy": 0.19958069324493408,
      "num_tokens": 51132676.0,
      "step": 22910
    },
    {
      "entropy": 5.435240173339844,
      "epoch": 2.5735946537878363,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004346290214258301,
      "loss": 5.1567,
      "mean_token_accuracy": 0.19375762194395066,
      "num_tokens": 51144317.0,
      "step": 22915
    },
    {
      "entropy": 5.6386651515960695,
      "epoch": 2.574156230695794,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043460072434103934,
      "loss": 5.3721,
      "mean_token_accuracy": 0.18261118978261948,
      "num_tokens": 51156181.0,
      "step": 22920
    },
    {
      "entropy": 5.426962900161743,
      "epoch": 2.5747178076037516,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004345724221744996,
      "loss": 5.1135,
      "mean_token_accuracy": 0.20535062551498412,
      "num_tokens": 51166825.0,
      "step": 22925
    },
    {
      "entropy": 5.504230594635009,
      "epoch": 2.575279384511709,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043454411492711197,
      "loss": 5.3378,
      "mean_token_accuracy": 0.1932621866464615,
      "num_tokens": 51178839.0,
      "step": 22930
    },
    {
      "entropy": 5.446401357650757,
      "epoch": 2.5758409614196665,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043451580259977793,
      "loss": 5.2076,
      "mean_token_accuracy": 0.19552009999752046,
      "num_tokens": 51190811.0,
      "step": 22935
    },
    {
      "entropy": 5.473272323608398,
      "epoch": 2.576402538327624,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043448748519339875,
      "loss": 5.1934,
      "mean_token_accuracy": 0.19591692537069322,
      "num_tokens": 51201991.0,
      "step": 22940
    },
    {
      "entropy": 5.526747941970825,
      "epoch": 2.5769641152355813,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043445916270887626,
      "loss": 5.2256,
      "mean_token_accuracy": 0.20157719552516937,
      "num_tokens": 51213017.0,
      "step": 22945
    },
    {
      "entropy": 5.49044189453125,
      "epoch": 2.577525692143539,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004344308351471123,
      "loss": 5.1731,
      "mean_token_accuracy": 0.1987592473626137,
      "num_tokens": 51224412.0,
      "step": 22950
    },
    {
      "entropy": 5.507965898513794,
      "epoch": 2.5780872690514967,
      "grad_norm": 1.125,
      "learning_rate": 0.00043440250250900854,
      "loss": 5.24,
      "mean_token_accuracy": 0.19685512632131577,
      "num_tokens": 51235043.0,
      "step": 22955
    },
    {
      "entropy": 5.484808254241943,
      "epoch": 2.578648845959454,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004343741647954673,
      "loss": 5.2642,
      "mean_token_accuracy": 0.1931155279278755,
      "num_tokens": 51246182.0,
      "step": 22960
    },
    {
      "entropy": 5.567646455764771,
      "epoch": 2.5792104228674115,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004343458220073909,
      "loss": 5.327,
      "mean_token_accuracy": 0.19570422619581224,
      "num_tokens": 51259791.0,
      "step": 22965
    },
    {
      "entropy": 5.588500118255615,
      "epoch": 2.579771999775369,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004343174741456817,
      "loss": 5.327,
      "mean_token_accuracy": 0.1901649221777916,
      "num_tokens": 51270530.0,
      "step": 22970
    },
    {
      "entropy": 5.408082675933838,
      "epoch": 2.580333576683327,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004342891212112423,
      "loss": 5.1613,
      "mean_token_accuracy": 0.19526751935482026,
      "num_tokens": 51281969.0,
      "step": 22975
    },
    {
      "entropy": 5.436216831207275,
      "epoch": 2.5808951535912845,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043426076320497557,
      "loss": 5.086,
      "mean_token_accuracy": 0.2031923860311508,
      "num_tokens": 51292790.0,
      "step": 22980
    },
    {
      "entropy": 5.453107118606567,
      "epoch": 2.5814567304992417,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043423240012778437,
      "loss": 5.1859,
      "mean_token_accuracy": 0.19469663947820665,
      "num_tokens": 51303949.0,
      "step": 22985
    },
    {
      "entropy": 5.448466014862061,
      "epoch": 2.5820183074071994,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004342040319805717,
      "loss": 5.1273,
      "mean_token_accuracy": 0.20429383367300033,
      "num_tokens": 51314570.0,
      "step": 22990
    },
    {
      "entropy": 5.350348806381225,
      "epoch": 2.582579884315157,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000434175658764241,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20541241019964218,
      "num_tokens": 51325210.0,
      "step": 22995
    },
    {
      "entropy": 5.420697450637817,
      "epoch": 2.5831414612231143,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043414728047969553,
      "loss": 5.0501,
      "mean_token_accuracy": 0.2020192414522171,
      "num_tokens": 51335871.0,
      "step": 23000
    },
    {
      "entropy": 5.459543085098266,
      "epoch": 2.583703038131072,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043411889712783897,
      "loss": 5.2261,
      "mean_token_accuracy": 0.20133862346410752,
      "num_tokens": 51346821.0,
      "step": 23005
    },
    {
      "entropy": 5.494036054611206,
      "epoch": 2.5842646150390296,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043409050870957505,
      "loss": 5.2234,
      "mean_token_accuracy": 0.1899457350373268,
      "num_tokens": 51358746.0,
      "step": 23010
    },
    {
      "entropy": 5.5528693199157715,
      "epoch": 2.5848261919469873,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004340621152258075,
      "loss": 5.2553,
      "mean_token_accuracy": 0.19268431812524794,
      "num_tokens": 51369628.0,
      "step": 23015
    },
    {
      "entropy": 5.575812244415284,
      "epoch": 2.585387768854945,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043403371667744055,
      "loss": 5.2254,
      "mean_token_accuracy": 0.20492194443941117,
      "num_tokens": 51382160.0,
      "step": 23020
    },
    {
      "entropy": 5.488270616531372,
      "epoch": 2.585949345762902,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004340053130653785,
      "loss": 5.2482,
      "mean_token_accuracy": 0.1930004522204399,
      "num_tokens": 51394843.0,
      "step": 23025
    },
    {
      "entropy": 5.458188247680664,
      "epoch": 2.58651092267086,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004339769043905255,
      "loss": 5.1908,
      "mean_token_accuracy": 0.2022606372833252,
      "num_tokens": 51405714.0,
      "step": 23030
    },
    {
      "entropy": 5.505884313583374,
      "epoch": 2.5870724995788175,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043394849065378624,
      "loss": 5.1488,
      "mean_token_accuracy": 0.1990939497947693,
      "num_tokens": 51416292.0,
      "step": 23035
    },
    {
      "entropy": 5.4181712627410885,
      "epoch": 2.5876340764867747,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043392007185606533,
      "loss": 5.1202,
      "mean_token_accuracy": 0.2047677993774414,
      "num_tokens": 51426993.0,
      "step": 23040
    },
    {
      "entropy": 5.430881786346435,
      "epoch": 2.5881956533947323,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004338916479982678,
      "loss": 5.261,
      "mean_token_accuracy": 0.19708542823791503,
      "num_tokens": 51439493.0,
      "step": 23045
    },
    {
      "entropy": 5.531647872924805,
      "epoch": 2.58875723030269,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043386321908129856,
      "loss": 5.1841,
      "mean_token_accuracy": 0.20050522834062576,
      "num_tokens": 51451705.0,
      "step": 23050
    },
    {
      "entropy": 5.560615825653076,
      "epoch": 2.589318807210647,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00043383478510606276,
      "loss": 5.155,
      "mean_token_accuracy": 0.20101606696844102,
      "num_tokens": 51463044.0,
      "step": 23055
    },
    {
      "entropy": 5.390435457229614,
      "epoch": 2.589880384118605,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004338063460734658,
      "loss": 5.1888,
      "mean_token_accuracy": 0.2007523164153099,
      "num_tokens": 51475009.0,
      "step": 23060
    },
    {
      "entropy": 5.416891384124756,
      "epoch": 2.5904419610265625,
      "grad_norm": 1.125,
      "learning_rate": 0.0004337779019844132,
      "loss": 5.1968,
      "mean_token_accuracy": 0.1994771331548691,
      "num_tokens": 51486068.0,
      "step": 23065
    },
    {
      "entropy": 5.532764530181884,
      "epoch": 2.59100353793452,
      "grad_norm": 1.125,
      "learning_rate": 0.00043374945283981066,
      "loss": 5.2478,
      "mean_token_accuracy": 0.20258497595787048,
      "num_tokens": 51497642.0,
      "step": 23070
    },
    {
      "entropy": 5.575276184082031,
      "epoch": 2.591565114842478,
      "grad_norm": 1.0,
      "learning_rate": 0.0004337209986405639,
      "loss": 5.248,
      "mean_token_accuracy": 0.19231952279806136,
      "num_tokens": 51511013.0,
      "step": 23075
    },
    {
      "entropy": 5.547844696044922,
      "epoch": 2.592126691750435,
      "grad_norm": 1.09375,
      "learning_rate": 0.000433692539387579,
      "loss": 5.3667,
      "mean_token_accuracy": 0.1921800434589386,
      "num_tokens": 51521687.0,
      "step": 23080
    },
    {
      "entropy": 5.525119400024414,
      "epoch": 2.5926882686583927,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004336640750817621,
      "loss": 5.2325,
      "mean_token_accuracy": 0.19389444291591645,
      "num_tokens": 51532288.0,
      "step": 23085
    },
    {
      "entropy": 5.534699010848999,
      "epoch": 2.5932498455663504,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043363560572401953,
      "loss": 5.2335,
      "mean_token_accuracy": 0.19883880764245987,
      "num_tokens": 51543101.0,
      "step": 23090
    },
    {
      "entropy": 5.462389039993286,
      "epoch": 2.5938114224743076,
      "grad_norm": 1.125,
      "learning_rate": 0.00043360713131525774,
      "loss": 5.2866,
      "mean_token_accuracy": 0.19957807213068007,
      "num_tokens": 51554729.0,
      "step": 23095
    },
    {
      "entropy": 5.492091512680053,
      "epoch": 2.5943729993822653,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004335786518563834,
      "loss": 5.2021,
      "mean_token_accuracy": 0.19596800506114959,
      "num_tokens": 51567232.0,
      "step": 23100
    },
    {
      "entropy": 5.553070068359375,
      "epoch": 2.594934576290223,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004335501673483033,
      "loss": 5.1923,
      "mean_token_accuracy": 0.1995108276605606,
      "num_tokens": 51577679.0,
      "step": 23105
    },
    {
      "entropy": 5.453506088256836,
      "epoch": 2.5954961531981806,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043352167779192423,
      "loss": 5.1237,
      "mean_token_accuracy": 0.196798375248909,
      "num_tokens": 51588656.0,
      "step": 23110
    },
    {
      "entropy": 5.468847608566284,
      "epoch": 2.5960577301061383,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043349318318815363,
      "loss": 5.2261,
      "mean_token_accuracy": 0.1949308842420578,
      "num_tokens": 51599016.0,
      "step": 23115
    },
    {
      "entropy": 5.317247819900513,
      "epoch": 2.5966193070140955,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043346468353789853,
      "loss": 4.9954,
      "mean_token_accuracy": 0.21364231258630753,
      "num_tokens": 51610288.0,
      "step": 23120
    },
    {
      "entropy": 5.483580780029297,
      "epoch": 2.597180883922053,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004334361788420664,
      "loss": 5.2992,
      "mean_token_accuracy": 0.19442235231399535,
      "num_tokens": 51622450.0,
      "step": 23125
    },
    {
      "entropy": 5.531235551834106,
      "epoch": 2.597742460830011,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043340766910156497,
      "loss": 5.1118,
      "mean_token_accuracy": 0.20397164821624755,
      "num_tokens": 51635111.0,
      "step": 23130
    },
    {
      "entropy": 5.473828744888306,
      "epoch": 2.598304037737968,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043337915431730195,
      "loss": 5.2457,
      "mean_token_accuracy": 0.1937500387430191,
      "num_tokens": 51646088.0,
      "step": 23135
    },
    {
      "entropy": 5.447720718383789,
      "epoch": 2.5988656146459257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043335063449018515,
      "loss": 5.1647,
      "mean_token_accuracy": 0.19829271733760834,
      "num_tokens": 51659337.0,
      "step": 23140
    },
    {
      "entropy": 5.547460508346558,
      "epoch": 2.5994271915538834,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004333221096211227,
      "loss": 5.3022,
      "mean_token_accuracy": 0.1992781475186348,
      "num_tokens": 51670954.0,
      "step": 23145
    },
    {
      "entropy": 5.447976112365723,
      "epoch": 2.599988768461841,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043329357971102296,
      "loss": 5.1381,
      "mean_token_accuracy": 0.1973046228289604,
      "num_tokens": 51681104.0,
      "step": 23150
    },
    {
      "entropy": 5.321099185943604,
      "epoch": 2.6005503453697982,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043326504476079425,
      "loss": 5.0318,
      "mean_token_accuracy": 0.2073124334216118,
      "num_tokens": 51691337.0,
      "step": 23155
    },
    {
      "entropy": 5.390917158126831,
      "epoch": 2.601111922277756,
      "grad_norm": 1.125,
      "learning_rate": 0.0004332365047713451,
      "loss": 5.1477,
      "mean_token_accuracy": 0.2008199766278267,
      "num_tokens": 51702488.0,
      "step": 23160
    },
    {
      "entropy": 5.53355302810669,
      "epoch": 2.6016734991857136,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004332079597435842,
      "loss": 5.2438,
      "mean_token_accuracy": 0.19955824315547943,
      "num_tokens": 51714439.0,
      "step": 23165
    },
    {
      "entropy": 5.630756664276123,
      "epoch": 2.602235076093671,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004331794096784206,
      "loss": 5.288,
      "mean_token_accuracy": 0.19511518627405167,
      "num_tokens": 51726423.0,
      "step": 23170
    },
    {
      "entropy": 5.529940795898438,
      "epoch": 2.6027966530016284,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004331508545767633,
      "loss": 5.2563,
      "mean_token_accuracy": 0.20738064795732497,
      "num_tokens": 51736198.0,
      "step": 23175
    },
    {
      "entropy": 5.395144462585449,
      "epoch": 2.603358229909586,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004331222944395213,
      "loss": 5.1376,
      "mean_token_accuracy": 0.20435541719198227,
      "num_tokens": 51746993.0,
      "step": 23180
    },
    {
      "entropy": 5.437501335144043,
      "epoch": 2.6039198068175438,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004330937292676042,
      "loss": 5.1534,
      "mean_token_accuracy": 0.20632504522800446,
      "num_tokens": 51756981.0,
      "step": 23185
    },
    {
      "entropy": 5.400945997238159,
      "epoch": 2.604481383725501,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004330651590619214,
      "loss": 5.161,
      "mean_token_accuracy": 0.20466522127389908,
      "num_tokens": 51768192.0,
      "step": 23190
    },
    {
      "entropy": 5.4939188957214355,
      "epoch": 2.6050429606334586,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004330365838233827,
      "loss": 5.2165,
      "mean_token_accuracy": 0.19679141342639922,
      "num_tokens": 51778996.0,
      "step": 23195
    },
    {
      "entropy": 5.5277821063995365,
      "epoch": 2.6056045375414163,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043300800355289786,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1924500972032547,
      "num_tokens": 51790179.0,
      "step": 23200
    },
    {
      "entropy": 5.470696496963501,
      "epoch": 2.606166114449374,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043297941825137684,
      "loss": 5.2489,
      "mean_token_accuracy": 0.20165283530950545,
      "num_tokens": 51801312.0,
      "step": 23205
    },
    {
      "entropy": 5.46109676361084,
      "epoch": 2.6067276913573316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043295082791972993,
      "loss": 5.1642,
      "mean_token_accuracy": 0.2015443429350853,
      "num_tokens": 51811960.0,
      "step": 23210
    },
    {
      "entropy": 5.454913663864136,
      "epoch": 2.607289268265289,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043292223255886737,
      "loss": 5.2618,
      "mean_token_accuracy": 0.19244524836540222,
      "num_tokens": 51823166.0,
      "step": 23215
    },
    {
      "entropy": 5.479498434066772,
      "epoch": 2.6078508451732465,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043289363216969973,
      "loss": 5.1326,
      "mean_token_accuracy": 0.19873957484960555,
      "num_tokens": 51835517.0,
      "step": 23220
    },
    {
      "entropy": 5.452497148513794,
      "epoch": 2.608412422081204,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043286502675313757,
      "loss": 5.1666,
      "mean_token_accuracy": 0.2023353934288025,
      "num_tokens": 51846690.0,
      "step": 23225
    },
    {
      "entropy": 5.389071798324585,
      "epoch": 2.6089739989891614,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004328364163100917,
      "loss": 5.159,
      "mean_token_accuracy": 0.20219334661960603,
      "num_tokens": 51857234.0,
      "step": 23230
    },
    {
      "entropy": 5.48324146270752,
      "epoch": 2.609535575897119,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004328078008414731,
      "loss": 5.243,
      "mean_token_accuracy": 0.20180547386407852,
      "num_tokens": 51867913.0,
      "step": 23235
    },
    {
      "entropy": 5.437986230850219,
      "epoch": 2.6100971528050767,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000432779180348193,
      "loss": 5.1851,
      "mean_token_accuracy": 0.2051671788096428,
      "num_tokens": 51878252.0,
      "step": 23240
    },
    {
      "entropy": 5.383799028396607,
      "epoch": 2.6106587297130344,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043275055483116254,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20199795216321945,
      "num_tokens": 51888407.0,
      "step": 23245
    },
    {
      "entropy": 5.508200931549072,
      "epoch": 2.6112203066209916,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004327219242912933,
      "loss": 5.1982,
      "mean_token_accuracy": 0.19967449456453323,
      "num_tokens": 51900042.0,
      "step": 23250
    },
    {
      "entropy": 5.510458517074585,
      "epoch": 2.6117818835289492,
      "grad_norm": 1.0,
      "learning_rate": 0.0004326932887294967,
      "loss": 5.21,
      "mean_token_accuracy": 0.19857455641031266,
      "num_tokens": 51909691.0,
      "step": 23255
    },
    {
      "entropy": 5.560432863235474,
      "epoch": 2.612343460436907,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004326646481466848,
      "loss": 5.2666,
      "mean_token_accuracy": 0.19549815654754638,
      "num_tokens": 51921691.0,
      "step": 23260
    },
    {
      "entropy": 5.457535886764527,
      "epoch": 2.6129050373448646,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004326360025437692,
      "loss": 5.1382,
      "mean_token_accuracy": 0.20069834738969802,
      "num_tokens": 51934050.0,
      "step": 23265
    },
    {
      "entropy": 5.486303091049194,
      "epoch": 2.613466614252822,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043260735192166217,
      "loss": 5.1686,
      "mean_token_accuracy": 0.2039589539170265,
      "num_tokens": 51944566.0,
      "step": 23270
    },
    {
      "entropy": 5.459819507598877,
      "epoch": 2.6140281911607794,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000432578696281276,
      "loss": 5.0944,
      "mean_token_accuracy": 0.20982656031847,
      "num_tokens": 51955485.0,
      "step": 23275
    },
    {
      "entropy": 5.425382089614868,
      "epoch": 2.614589768068737,
      "grad_norm": 1.234375,
      "learning_rate": 0.000432550035623523,
      "loss": 5.1682,
      "mean_token_accuracy": 0.20477787256240845,
      "num_tokens": 51965124.0,
      "step": 23280
    },
    {
      "entropy": 5.551755666732788,
      "epoch": 2.6151513449766943,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004325213699493157,
      "loss": 5.2096,
      "mean_token_accuracy": 0.2066628560423851,
      "num_tokens": 51976362.0,
      "step": 23285
    },
    {
      "entropy": 5.376646900177002,
      "epoch": 2.615712921884652,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004324926992595669,
      "loss": 5.0861,
      "mean_token_accuracy": 0.19803429394960403,
      "num_tokens": 51988068.0,
      "step": 23290
    },
    {
      "entropy": 5.44796347618103,
      "epoch": 2.6162744987926096,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004324640235551895,
      "loss": 5.1431,
      "mean_token_accuracy": 0.20669957399368286,
      "num_tokens": 51998392.0,
      "step": 23295
    },
    {
      "entropy": 5.409528064727783,
      "epoch": 2.6168360757005673,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004324353428370965,
      "loss": 5.1153,
      "mean_token_accuracy": 0.19969575405120848,
      "num_tokens": 52009416.0,
      "step": 23300
    },
    {
      "entropy": 5.540070486068726,
      "epoch": 2.617397652608525,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004324066571062012,
      "loss": 5.1928,
      "mean_token_accuracy": 0.20413899719715117,
      "num_tokens": 52020554.0,
      "step": 23305
    },
    {
      "entropy": 5.405008935928345,
      "epoch": 2.617959229516482,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004323779663634168,
      "loss": 5.1964,
      "mean_token_accuracy": 0.19754453748464584,
      "num_tokens": 52031301.0,
      "step": 23310
    },
    {
      "entropy": 5.419484329223633,
      "epoch": 2.61852080642444,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000432349270609657,
      "loss": 5.1435,
      "mean_token_accuracy": 0.2014838293194771,
      "num_tokens": 52041720.0,
      "step": 23315
    },
    {
      "entropy": 5.39289870262146,
      "epoch": 2.6190823833323975,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043232056984583527,
      "loss": 5.0732,
      "mean_token_accuracy": 0.2068752959370613,
      "num_tokens": 52051911.0,
      "step": 23320
    },
    {
      "entropy": 5.439510345458984,
      "epoch": 2.6196439602403547,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004322918640728656,
      "loss": 5.1656,
      "mean_token_accuracy": 0.20078354924917222,
      "num_tokens": 52062842.0,
      "step": 23325
    },
    {
      "entropy": 5.469933795928955,
      "epoch": 2.6202055371483124,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043226315329166203,
      "loss": 5.305,
      "mean_token_accuracy": 0.19540444612503052,
      "num_tokens": 52075034.0,
      "step": 23330
    },
    {
      "entropy": 5.499102926254272,
      "epoch": 2.62076711405627,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043223443750313873,
      "loss": 5.0946,
      "mean_token_accuracy": 0.21070630103349686,
      "num_tokens": 52084782.0,
      "step": 23335
    },
    {
      "entropy": 5.3721761226654055,
      "epoch": 2.6213286909642277,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004322057167082099,
      "loss": 5.1122,
      "mean_token_accuracy": 0.19998955577611924,
      "num_tokens": 52095623.0,
      "step": 23340
    },
    {
      "entropy": 5.439248418807983,
      "epoch": 2.621890267872185,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043217699090779006,
      "loss": 5.1704,
      "mean_token_accuracy": 0.20737988203763963,
      "num_tokens": 52107240.0,
      "step": 23345
    },
    {
      "entropy": 5.4496763229370115,
      "epoch": 2.6224518447801426,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043214826010279383,
      "loss": 5.2667,
      "mean_token_accuracy": 0.19826025366783143,
      "num_tokens": 52118157.0,
      "step": 23350
    },
    {
      "entropy": 5.565532588958741,
      "epoch": 2.6230134216881003,
      "grad_norm": 1.375,
      "learning_rate": 0.0004321195242941361,
      "loss": 5.2864,
      "mean_token_accuracy": 0.18660368919372558,
      "num_tokens": 52130578.0,
      "step": 23355
    },
    {
      "entropy": 5.540079402923584,
      "epoch": 2.623574998596058,
      "grad_norm": 1.25,
      "learning_rate": 0.00043209078348273175,
      "loss": 5.1551,
      "mean_token_accuracy": 0.2040082111954689,
      "num_tokens": 52142057.0,
      "step": 23360
    },
    {
      "entropy": 5.5023033618927,
      "epoch": 2.624136575504015,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004320620376694959,
      "loss": 5.2609,
      "mean_token_accuracy": 0.19550875574350357,
      "num_tokens": 52152370.0,
      "step": 23365
    },
    {
      "entropy": 5.404144763946533,
      "epoch": 2.624698152411973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043203328685534396,
      "loss": 5.1575,
      "mean_token_accuracy": 0.20197988003492356,
      "num_tokens": 52162139.0,
      "step": 23370
    },
    {
      "entropy": 5.493215322494507,
      "epoch": 2.6252597293199305,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004320045310411912,
      "loss": 5.2252,
      "mean_token_accuracy": 0.20067870914936065,
      "num_tokens": 52172820.0,
      "step": 23375
    },
    {
      "entropy": 5.429520177841186,
      "epoch": 2.6258213062278877,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043197577022795327,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20700942426919938,
      "num_tokens": 52183315.0,
      "step": 23380
    },
    {
      "entropy": 5.355142116546631,
      "epoch": 2.6263828831358453,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004319470044165459,
      "loss": 5.1411,
      "mean_token_accuracy": 0.20844318121671676,
      "num_tokens": 52194776.0,
      "step": 23385
    },
    {
      "entropy": 5.56971287727356,
      "epoch": 2.626944460043803,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004319182336078852,
      "loss": 5.3254,
      "mean_token_accuracy": 0.19389017671346664,
      "num_tokens": 52207370.0,
      "step": 23390
    },
    {
      "entropy": 5.484973049163818,
      "epoch": 2.6275060369517607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000431889457802887,
      "loss": 5.1299,
      "mean_token_accuracy": 0.20800730139017104,
      "num_tokens": 52218287.0,
      "step": 23395
    },
    {
      "entropy": 5.44383020401001,
      "epoch": 2.6280676138597183,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004318606770024675,
      "loss": 5.255,
      "mean_token_accuracy": 0.19475822299718856,
      "num_tokens": 52229418.0,
      "step": 23400
    },
    {
      "entropy": 5.563400936126709,
      "epoch": 2.6286291907676755,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004318318912075433,
      "loss": 5.2024,
      "mean_token_accuracy": 0.19832005500793456,
      "num_tokens": 52240382.0,
      "step": 23405
    },
    {
      "entropy": 5.518240451812744,
      "epoch": 2.629190767675633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004318031004190308,
      "loss": 5.2272,
      "mean_token_accuracy": 0.19684348702430726,
      "num_tokens": 52252608.0,
      "step": 23410
    },
    {
      "entropy": 5.431765222549439,
      "epoch": 2.629752344583591,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043177430463784685,
      "loss": 5.133,
      "mean_token_accuracy": 0.2142768010497093,
      "num_tokens": 52262940.0,
      "step": 23415
    },
    {
      "entropy": 5.367644214630127,
      "epoch": 2.630313921491548,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043174550386490816,
      "loss": 5.0846,
      "mean_token_accuracy": 0.20578556060791015,
      "num_tokens": 52273270.0,
      "step": 23420
    },
    {
      "entropy": 5.427347135543823,
      "epoch": 2.6308754983995057,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004317166981011319,
      "loss": 5.1611,
      "mean_token_accuracy": 0.200395405292511,
      "num_tokens": 52284433.0,
      "step": 23425
    },
    {
      "entropy": 5.548338174819946,
      "epoch": 2.6314370753074634,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043168788734743516,
      "loss": 5.1775,
      "mean_token_accuracy": 0.20473337322473525,
      "num_tokens": 52296088.0,
      "step": 23430
    },
    {
      "entropy": 5.497882080078125,
      "epoch": 2.631998652215421,
      "grad_norm": 1.125,
      "learning_rate": 0.00043165907160473525,
      "loss": 5.2498,
      "mean_token_accuracy": 0.19243747889995574,
      "num_tokens": 52307683.0,
      "step": 23435
    },
    {
      "entropy": 5.447935438156128,
      "epoch": 2.6325602291233783,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004316302508739499,
      "loss": 5.2304,
      "mean_token_accuracy": 0.19726531058549882,
      "num_tokens": 52319845.0,
      "step": 23440
    },
    {
      "entropy": 5.488337135314941,
      "epoch": 2.633121806031336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004316014251559965,
      "loss": 5.1653,
      "mean_token_accuracy": 0.2008945792913437,
      "num_tokens": 52330196.0,
      "step": 23445
    },
    {
      "entropy": 5.593804883956909,
      "epoch": 2.6336833829392936,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043157259445179304,
      "loss": 5.3748,
      "mean_token_accuracy": 0.19053859859704972,
      "num_tokens": 52341510.0,
      "step": 23450
    },
    {
      "entropy": 5.471933364868164,
      "epoch": 2.6342449598472513,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004315437587622574,
      "loss": 5.1341,
      "mean_token_accuracy": 0.2022364392876625,
      "num_tokens": 52351848.0,
      "step": 23455
    },
    {
      "entropy": 5.483906698226929,
      "epoch": 2.6348065367552085,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043151491808830777,
      "loss": 5.1955,
      "mean_token_accuracy": 0.1995692566037178,
      "num_tokens": 52363698.0,
      "step": 23460
    },
    {
      "entropy": 5.477593946456909,
      "epoch": 2.635368113663166,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043148607243086255,
      "loss": 5.1555,
      "mean_token_accuracy": 0.1993621587753296,
      "num_tokens": 52374433.0,
      "step": 23465
    },
    {
      "entropy": 5.4144477367401125,
      "epoch": 2.635929690571124,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004314572217908399,
      "loss": 5.2165,
      "mean_token_accuracy": 0.20255142897367479,
      "num_tokens": 52385762.0,
      "step": 23470
    },
    {
      "entropy": 5.481233692169189,
      "epoch": 2.636491267479081,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004314283661691587,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21361292004585267,
      "num_tokens": 52397058.0,
      "step": 23475
    },
    {
      "entropy": 5.512315082550049,
      "epoch": 2.6370528443870387,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004313995055667377,
      "loss": 5.2547,
      "mean_token_accuracy": 0.1977449432015419,
      "num_tokens": 52407671.0,
      "step": 23480
    },
    {
      "entropy": 5.477112340927124,
      "epoch": 2.6376144212949963,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004313706399844958,
      "loss": 5.2396,
      "mean_token_accuracy": 0.1980699732899666,
      "num_tokens": 52419511.0,
      "step": 23485
    },
    {
      "entropy": 5.394292974472046,
      "epoch": 2.638175998202954,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000431341769423352,
      "loss": 5.1114,
      "mean_token_accuracy": 0.1963311478495598,
      "num_tokens": 52431311.0,
      "step": 23490
    },
    {
      "entropy": 5.493609428405762,
      "epoch": 2.6387375751109117,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004313128938842257,
      "loss": 5.152,
      "mean_token_accuracy": 0.19902120381593705,
      "num_tokens": 52442184.0,
      "step": 23495
    },
    {
      "entropy": 5.515880727767945,
      "epoch": 2.639299152018869,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043128401336803616,
      "loss": 5.223,
      "mean_token_accuracy": 0.19565608203411103,
      "num_tokens": 52454837.0,
      "step": 23500
    },
    {
      "entropy": 5.5168108463287355,
      "epoch": 2.6398607289268265,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043125512787570305,
      "loss": 5.2155,
      "mean_token_accuracy": 0.199556864798069,
      "num_tokens": 52465378.0,
      "step": 23505
    },
    {
      "entropy": 5.483809185028076,
      "epoch": 2.640422305834784,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004312262374081461,
      "loss": 5.2363,
      "mean_token_accuracy": 0.19988692104816436,
      "num_tokens": 52476025.0,
      "step": 23510
    },
    {
      "entropy": 5.536465692520141,
      "epoch": 2.6409838827427414,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043119734196628507,
      "loss": 5.1809,
      "mean_token_accuracy": 0.1984606832265854,
      "num_tokens": 52487408.0,
      "step": 23515
    },
    {
      "entropy": 5.496368551254273,
      "epoch": 2.641545459650699,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043116844155104025,
      "loss": 5.1245,
      "mean_token_accuracy": 0.20686034560203553,
      "num_tokens": 52497801.0,
      "step": 23520
    },
    {
      "entropy": 5.411055946350098,
      "epoch": 2.6421070365586568,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004311395361633316,
      "loss": 5.1853,
      "mean_token_accuracy": 0.19698130935430527,
      "num_tokens": 52508918.0,
      "step": 23525
    },
    {
      "entropy": 5.412768316268921,
      "epoch": 2.6426686134666144,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004311106258040795,
      "loss": 5.1194,
      "mean_token_accuracy": 0.20621820837259291,
      "num_tokens": 52519951.0,
      "step": 23530
    },
    {
      "entropy": 5.340115308761597,
      "epoch": 2.643230190374572,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043108171047420463,
      "loss": 5.122,
      "mean_token_accuracy": 0.1990943282842636,
      "num_tokens": 52530710.0,
      "step": 23535
    },
    {
      "entropy": 5.361363506317138,
      "epoch": 2.6437917672825293,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004310527901746275,
      "loss": 5.086,
      "mean_token_accuracy": 0.20682694166898727,
      "num_tokens": 52542188.0,
      "step": 23540
    },
    {
      "entropy": 5.488569259643555,
      "epoch": 2.644353344190487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000431023864906269,
      "loss": 5.2185,
      "mean_token_accuracy": 0.20059878677129744,
      "num_tokens": 52553848.0,
      "step": 23545
    },
    {
      "entropy": 5.427157735824585,
      "epoch": 2.6449149210984446,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043099493467005015,
      "loss": 5.1484,
      "mean_token_accuracy": 0.2035217434167862,
      "num_tokens": 52563338.0,
      "step": 23550
    },
    {
      "entropy": 5.470952558517456,
      "epoch": 2.645476498006402,
      "grad_norm": 1.203125,
      "learning_rate": 0.000430965999466892,
      "loss": 5.2901,
      "mean_token_accuracy": 0.19292771965265273,
      "num_tokens": 52576369.0,
      "step": 23555
    },
    {
      "entropy": 5.481601095199585,
      "epoch": 2.6460380749143595,
      "grad_norm": 1.125,
      "learning_rate": 0.00043093705929771604,
      "loss": 5.199,
      "mean_token_accuracy": 0.20293420553207397,
      "num_tokens": 52587409.0,
      "step": 23560
    },
    {
      "entropy": 5.458005094528199,
      "epoch": 2.646599651822317,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004309081141634436,
      "loss": 5.138,
      "mean_token_accuracy": 0.19908208400011063,
      "num_tokens": 52599398.0,
      "step": 23565
    },
    {
      "entropy": 5.462128734588623,
      "epoch": 2.6471612287302744,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043087916406499635,
      "loss": 5.1289,
      "mean_token_accuracy": 0.2062810480594635,
      "num_tokens": 52609565.0,
      "step": 23570
    },
    {
      "entropy": 5.42417426109314,
      "epoch": 2.647722805638232,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043085020900329604,
      "loss": 5.1929,
      "mean_token_accuracy": 0.203407683968544,
      "num_tokens": 52621150.0,
      "step": 23575
    },
    {
      "entropy": 5.440845346450805,
      "epoch": 2.6482843825461897,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043082124897926463,
      "loss": 5.1862,
      "mean_token_accuracy": 0.20108286887407303,
      "num_tokens": 52632030.0,
      "step": 23580
    },
    {
      "entropy": 5.509593296051025,
      "epoch": 2.6488459594541474,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004307922839938242,
      "loss": 5.2972,
      "mean_token_accuracy": 0.195563405752182,
      "num_tokens": 52643447.0,
      "step": 23585
    },
    {
      "entropy": 5.509720754623413,
      "epoch": 2.649407536362105,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043076331404789703,
      "loss": 5.1771,
      "mean_token_accuracy": 0.2076278507709503,
      "num_tokens": 52655017.0,
      "step": 23590
    },
    {
      "entropy": 5.5232213020324705,
      "epoch": 2.6499691132700622,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004307343391424055,
      "loss": 5.2104,
      "mean_token_accuracy": 0.19614923596382142,
      "num_tokens": 52665674.0,
      "step": 23595
    },
    {
      "entropy": 5.39630389213562,
      "epoch": 2.65053069017802,
      "grad_norm": 1.125,
      "learning_rate": 0.0004307053592782723,
      "loss": 5.1602,
      "mean_token_accuracy": 0.20166541785001754,
      "num_tokens": 52678177.0,
      "step": 23600
    },
    {
      "entropy": 5.471887302398682,
      "epoch": 2.6510922670859776,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043067637445642004,
      "loss": 5.2367,
      "mean_token_accuracy": 0.199528306722641,
      "num_tokens": 52689326.0,
      "step": 23605
    },
    {
      "entropy": 5.503168153762817,
      "epoch": 2.651653843993935,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043064738467777166,
      "loss": 5.2686,
      "mean_token_accuracy": 0.1897120252251625,
      "num_tokens": 52701686.0,
      "step": 23610
    },
    {
      "entropy": 5.455009794235229,
      "epoch": 2.6522154209018924,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043061838994325007,
      "loss": 5.1746,
      "mean_token_accuracy": 0.20472166687250137,
      "num_tokens": 52713422.0,
      "step": 23615
    },
    {
      "entropy": 5.4553224563598635,
      "epoch": 2.65277699780985,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043058939025377867,
      "loss": 5.2259,
      "mean_token_accuracy": 0.19934384673833846,
      "num_tokens": 52724438.0,
      "step": 23620
    },
    {
      "entropy": 5.455756664276123,
      "epoch": 2.6533385747178078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004305603856102808,
      "loss": 5.1782,
      "mean_token_accuracy": 0.20492662936449052,
      "num_tokens": 52734657.0,
      "step": 23625
    },
    {
      "entropy": 5.455070829391479,
      "epoch": 2.6539001516257654,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004305313760136799,
      "loss": 5.1379,
      "mean_token_accuracy": 0.20828091055154802,
      "num_tokens": 52745832.0,
      "step": 23630
    },
    {
      "entropy": 5.506795835494995,
      "epoch": 2.6544617285337226,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043050236146489955,
      "loss": 5.3156,
      "mean_token_accuracy": 0.1952225759625435,
      "num_tokens": 52758795.0,
      "step": 23635
    },
    {
      "entropy": 5.488795757293701,
      "epoch": 2.6550233054416803,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004304733419648638,
      "loss": 5.2211,
      "mean_token_accuracy": 0.2016402930021286,
      "num_tokens": 52769154.0,
      "step": 23640
    },
    {
      "entropy": 5.472024202346802,
      "epoch": 2.655584882349638,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043044431751449656,
      "loss": 5.1591,
      "mean_token_accuracy": 0.20421334505081176,
      "num_tokens": 52779189.0,
      "step": 23645
    },
    {
      "entropy": 5.516177177429199,
      "epoch": 2.656146459257595,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004304152881147219,
      "loss": 5.2928,
      "mean_token_accuracy": 0.20140977799892426,
      "num_tokens": 52789545.0,
      "step": 23650
    },
    {
      "entropy": 5.587547445297242,
      "epoch": 2.656708036165553,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004303862537664642,
      "loss": 5.4089,
      "mean_token_accuracy": 0.18406397551298143,
      "num_tokens": 52802410.0,
      "step": 23655
    },
    {
      "entropy": 5.433873414993286,
      "epoch": 2.6572696130735105,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000430357214470648,
      "loss": 5.1877,
      "mean_token_accuracy": 0.20008882135152817,
      "num_tokens": 52814223.0,
      "step": 23660
    },
    {
      "entropy": 5.400897598266601,
      "epoch": 2.6578311899814677,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004303281702281977,
      "loss": 5.1332,
      "mean_token_accuracy": 0.2064850002527237,
      "num_tokens": 52824554.0,
      "step": 23665
    },
    {
      "entropy": 5.553000831604004,
      "epoch": 2.6583927668894254,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043029912104003827,
      "loss": 5.3241,
      "mean_token_accuracy": 0.1995888665318489,
      "num_tokens": 52836058.0,
      "step": 23670
    },
    {
      "entropy": 5.467930746078491,
      "epoch": 2.658954343797383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043027006690709457,
      "loss": 5.265,
      "mean_token_accuracy": 0.19593652337789536,
      "num_tokens": 52847220.0,
      "step": 23675
    },
    {
      "entropy": 5.468914222717285,
      "epoch": 2.6595159207053407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004302410078302918,
      "loss": 5.1019,
      "mean_token_accuracy": 0.20464342683553696,
      "num_tokens": 52858242.0,
      "step": 23680
    },
    {
      "entropy": 5.384401178359985,
      "epoch": 2.6600774976132984,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043021194381055505,
      "loss": 5.0472,
      "mean_token_accuracy": 0.20356269925832748,
      "num_tokens": 52869439.0,
      "step": 23685
    },
    {
      "entropy": 5.583893728256226,
      "epoch": 2.6606390745212556,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043018287484880986,
      "loss": 5.3961,
      "mean_token_accuracy": 0.19009190201759338,
      "num_tokens": 52879893.0,
      "step": 23690
    },
    {
      "entropy": 5.540567922592163,
      "epoch": 2.6612006514292132,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004301538009459817,
      "loss": 5.2299,
      "mean_token_accuracy": 0.1975465565919876,
      "num_tokens": 52891352.0,
      "step": 23695
    },
    {
      "entropy": 5.431209564208984,
      "epoch": 2.661762228337171,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004301247221029964,
      "loss": 5.2186,
      "mean_token_accuracy": 0.19581498354673385,
      "num_tokens": 52904202.0,
      "step": 23700
    },
    {
      "entropy": 5.466895246505738,
      "epoch": 2.662323805245128,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043009563832077977,
      "loss": 5.1392,
      "mean_token_accuracy": 0.2047729477286339,
      "num_tokens": 52914495.0,
      "step": 23705
    },
    {
      "entropy": 5.477147245407105,
      "epoch": 2.662885382153086,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004300665496002578,
      "loss": 5.213,
      "mean_token_accuracy": 0.19413714557886125,
      "num_tokens": 52925930.0,
      "step": 23710
    },
    {
      "entropy": 5.504684257507324,
      "epoch": 2.6634469590610435,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043003745594235687,
      "loss": 5.1831,
      "mean_token_accuracy": 0.19708070307970046,
      "num_tokens": 52936310.0,
      "step": 23715
    },
    {
      "entropy": 5.488216876983643,
      "epoch": 2.664008535969001,
      "grad_norm": 1.375,
      "learning_rate": 0.0004300083573480031,
      "loss": 5.2992,
      "mean_token_accuracy": 0.19424211531877517,
      "num_tokens": 52948801.0,
      "step": 23720
    },
    {
      "entropy": 5.553876447677612,
      "epoch": 2.6645701128769588,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042997925381812313,
      "loss": 5.3393,
      "mean_token_accuracy": 0.19497064799070357,
      "num_tokens": 52960541.0,
      "step": 23725
    },
    {
      "entropy": 5.481603574752808,
      "epoch": 2.665131689784916,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042995014535364363,
      "loss": 5.2075,
      "mean_token_accuracy": 0.1960158884525299,
      "num_tokens": 52971973.0,
      "step": 23730
    },
    {
      "entropy": 5.4324750900268555,
      "epoch": 2.6656932666928737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004299210319554914,
      "loss": 5.1737,
      "mean_token_accuracy": 0.21714765429496766,
      "num_tokens": 52983124.0,
      "step": 23735
    },
    {
      "entropy": 5.479479217529297,
      "epoch": 2.6662548436008313,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042989191362459347,
      "loss": 5.1177,
      "mean_token_accuracy": 0.2090711385011673,
      "num_tokens": 52994402.0,
      "step": 23740
    },
    {
      "entropy": 5.466092920303344,
      "epoch": 2.6668164205087885,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004298627903618769,
      "loss": 5.2374,
      "mean_token_accuracy": 0.19495458900928497,
      "num_tokens": 53005253.0,
      "step": 23745
    },
    {
      "entropy": 5.487114906311035,
      "epoch": 2.667377997416746,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042983366216826907,
      "loss": 5.2558,
      "mean_token_accuracy": 0.20063638240098952,
      "num_tokens": 53016599.0,
      "step": 23750
    },
    {
      "entropy": 5.452792739868164,
      "epoch": 2.667939574324704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004298045290446973,
      "loss": 5.1789,
      "mean_token_accuracy": 0.20471510142087937,
      "num_tokens": 53027412.0,
      "step": 23755
    },
    {
      "entropy": 5.491108417510986,
      "epoch": 2.668501151232661,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042977539099208937,
      "loss": 5.2177,
      "mean_token_accuracy": 0.19474119395017625,
      "num_tokens": 53039490.0,
      "step": 23760
    },
    {
      "entropy": 5.478243064880371,
      "epoch": 2.6690627281406187,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042974624801137283,
      "loss": 5.1946,
      "mean_token_accuracy": 0.2009686827659607,
      "num_tokens": 53049837.0,
      "step": 23765
    },
    {
      "entropy": 5.433179664611816,
      "epoch": 2.6696243050485764,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042971710010347594,
      "loss": 5.1118,
      "mean_token_accuracy": 0.211330346763134,
      "num_tokens": 53060187.0,
      "step": 23770
    },
    {
      "entropy": 5.42611174583435,
      "epoch": 2.670185881956534,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042968794726932634,
      "loss": 5.1688,
      "mean_token_accuracy": 0.2024862825870514,
      "num_tokens": 53072050.0,
      "step": 23775
    },
    {
      "entropy": 5.451820421218872,
      "epoch": 2.6707474588644917,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004296587895098527,
      "loss": 5.2358,
      "mean_token_accuracy": 0.19662144035100937,
      "num_tokens": 53083235.0,
      "step": 23780
    },
    {
      "entropy": 5.5170825004577635,
      "epoch": 2.671309035772449,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042962962682598316,
      "loss": 5.2409,
      "mean_token_accuracy": 0.19715406000614166,
      "num_tokens": 53093615.0,
      "step": 23785
    },
    {
      "entropy": 5.5463752269744875,
      "epoch": 2.6718706126804066,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042960045921864627,
      "loss": 5.2889,
      "mean_token_accuracy": 0.19199395626783372,
      "num_tokens": 53105611.0,
      "step": 23790
    },
    {
      "entropy": 5.471039342880249,
      "epoch": 2.6724321895883643,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004295712866887709,
      "loss": 5.0522,
      "mean_token_accuracy": 0.2054540604352951,
      "num_tokens": 53116343.0,
      "step": 23795
    },
    {
      "entropy": 5.454292106628418,
      "epoch": 2.6729937664963215,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042954210923728565,
      "loss": 5.2195,
      "mean_token_accuracy": 0.2033547967672348,
      "num_tokens": 53128754.0,
      "step": 23800
    },
    {
      "entropy": 5.388036823272705,
      "epoch": 2.673555343404279,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004295129268651198,
      "loss": 5.091,
      "mean_token_accuracy": 0.210646091401577,
      "num_tokens": 53139508.0,
      "step": 23805
    },
    {
      "entropy": 5.411625480651855,
      "epoch": 2.674116920312237,
      "grad_norm": 1.25,
      "learning_rate": 0.0004294837395732025,
      "loss": 5.0998,
      "mean_token_accuracy": 0.20751608312129974,
      "num_tokens": 53149925.0,
      "step": 23810
    },
    {
      "entropy": 5.435418653488159,
      "epoch": 2.6746784972201945,
      "grad_norm": 1.125,
      "learning_rate": 0.00042945454736246293,
      "loss": 5.1634,
      "mean_token_accuracy": 0.20198342949151993,
      "num_tokens": 53160053.0,
      "step": 23815
    },
    {
      "entropy": 5.495969581604004,
      "epoch": 2.675240074128152,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042942535023383075,
      "loss": 5.1703,
      "mean_token_accuracy": 0.20624391883611679,
      "num_tokens": 53170723.0,
      "step": 23820
    },
    {
      "entropy": 5.39514856338501,
      "epoch": 2.6758016510361093,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042939614818823547,
      "loss": 5.2044,
      "mean_token_accuracy": 0.19773222655057907,
      "num_tokens": 53182120.0,
      "step": 23825
    },
    {
      "entropy": 5.435149335861206,
      "epoch": 2.676363227944067,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004293669412266069,
      "loss": 5.1888,
      "mean_token_accuracy": 0.19621680080890655,
      "num_tokens": 53192481.0,
      "step": 23830
    },
    {
      "entropy": 5.4549168109893795,
      "epoch": 2.6769248048520247,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004293377293498751,
      "loss": 5.1761,
      "mean_token_accuracy": 0.1969471037387848,
      "num_tokens": 53203710.0,
      "step": 23835
    },
    {
      "entropy": 5.467357540130616,
      "epoch": 2.677486381759982,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042930851255897014,
      "loss": 5.0953,
      "mean_token_accuracy": 0.20897049307823182,
      "num_tokens": 53214051.0,
      "step": 23840
    },
    {
      "entropy": 5.39466700553894,
      "epoch": 2.6780479586679395,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004292792908548223,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20404789447784424,
      "num_tokens": 53224981.0,
      "step": 23845
    },
    {
      "entropy": 5.334466934204102,
      "epoch": 2.678609535575897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042925006423836197,
      "loss": 4.9957,
      "mean_token_accuracy": 0.2121786430478096,
      "num_tokens": 53236336.0,
      "step": 23850
    },
    {
      "entropy": 5.428548002243042,
      "epoch": 2.6791711124838544,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042922083271051975,
      "loss": 5.2174,
      "mean_token_accuracy": 0.20053329914808274,
      "num_tokens": 53248212.0,
      "step": 23855
    },
    {
      "entropy": 5.479628276824951,
      "epoch": 2.679732689391812,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004291915962722265,
      "loss": 5.1958,
      "mean_token_accuracy": 0.20091208964586257,
      "num_tokens": 53258910.0,
      "step": 23860
    },
    {
      "entropy": 5.550708818435669,
      "epoch": 2.6802942662997697,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004291623549244129,
      "loss": 5.2285,
      "mean_token_accuracy": 0.1984703153371811,
      "num_tokens": 53269445.0,
      "step": 23865
    },
    {
      "entropy": 5.455594396591186,
      "epoch": 2.6808558432077274,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042913310866801024,
      "loss": 5.1566,
      "mean_token_accuracy": 0.20768319070339203,
      "num_tokens": 53280119.0,
      "step": 23870
    },
    {
      "entropy": 5.500387048721313,
      "epoch": 2.681417420115685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042910385750394964,
      "loss": 5.1934,
      "mean_token_accuracy": 0.20669564604759216,
      "num_tokens": 53291496.0,
      "step": 23875
    },
    {
      "entropy": 5.46678991317749,
      "epoch": 2.6819789970236423,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004290746014331623,
      "loss": 5.1631,
      "mean_token_accuracy": 0.19575610905885696,
      "num_tokens": 53303470.0,
      "step": 23880
    },
    {
      "entropy": 5.520257139205933,
      "epoch": 2.6825405739316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042904534045658,
      "loss": 5.2588,
      "mean_token_accuracy": 0.20200514942407607,
      "num_tokens": 53313552.0,
      "step": 23885
    },
    {
      "entropy": 5.437272071838379,
      "epoch": 2.6831021508395576,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004290160745751343,
      "loss": 5.2804,
      "mean_token_accuracy": 0.1971744939684868,
      "num_tokens": 53325280.0,
      "step": 23890
    },
    {
      "entropy": 5.3667339324951175,
      "epoch": 2.683663727747515,
      "grad_norm": 1.125,
      "learning_rate": 0.0004289868037897571,
      "loss": 5.044,
      "mean_token_accuracy": 0.2140959158539772,
      "num_tokens": 53334982.0,
      "step": 23895
    },
    {
      "entropy": 5.50390248298645,
      "epoch": 2.6842253046554725,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004289575281013802,
      "loss": 5.2249,
      "mean_token_accuracy": 0.19505434781312941,
      "num_tokens": 53346367.0,
      "step": 23900
    },
    {
      "entropy": 5.463916778564453,
      "epoch": 2.68478688156343,
      "grad_norm": 1.0625,
      "learning_rate": 0.000428928247510936,
      "loss": 5.1597,
      "mean_token_accuracy": 0.20073503702878953,
      "num_tokens": 53356907.0,
      "step": 23905
    },
    {
      "entropy": 5.451604270935059,
      "epoch": 2.685348458471388,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004288989620193567,
      "loss": 5.1413,
      "mean_token_accuracy": 0.2048221156001091,
      "num_tokens": 53366619.0,
      "step": 23910
    },
    {
      "entropy": 5.489072227478028,
      "epoch": 2.6859100353793455,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042886967162757475,
      "loss": 5.2027,
      "mean_token_accuracy": 0.20132599025964737,
      "num_tokens": 53377578.0,
      "step": 23915
    },
    {
      "entropy": 5.532238483428955,
      "epoch": 2.6864716122873027,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042884037633652276,
      "loss": 5.1694,
      "mean_token_accuracy": 0.19484029412269593,
      "num_tokens": 53389416.0,
      "step": 23920
    },
    {
      "entropy": 5.399318790435791,
      "epoch": 2.6870331891952604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042881107614713355,
      "loss": 5.0257,
      "mean_token_accuracy": 0.2138148456811905,
      "num_tokens": 53399453.0,
      "step": 23925
    },
    {
      "entropy": 5.337710618972778,
      "epoch": 2.687594766103218,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042878177106034,
      "loss": 5.1795,
      "mean_token_accuracy": 0.19844985008239746,
      "num_tokens": 53409933.0,
      "step": 23930
    },
    {
      "entropy": 5.467342567443848,
      "epoch": 2.6881563430111752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004287524610770752,
      "loss": 5.2255,
      "mean_token_accuracy": 0.20246859788894653,
      "num_tokens": 53420943.0,
      "step": 23935
    },
    {
      "entropy": 5.390620326995849,
      "epoch": 2.688717919919133,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042872314619827234,
      "loss": 5.0613,
      "mean_token_accuracy": 0.2050351619720459,
      "num_tokens": 53431450.0,
      "step": 23940
    },
    {
      "entropy": 5.334903001785278,
      "epoch": 2.6892794968270906,
      "grad_norm": 1.0625,
      "learning_rate": 0.000428693826424865,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2082084074616432,
      "num_tokens": 53442198.0,
      "step": 23945
    },
    {
      "entropy": 5.386764860153198,
      "epoch": 2.6898410737350478,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004286645017577866,
      "loss": 5.0642,
      "mean_token_accuracy": 0.20550597161054612,
      "num_tokens": 53452411.0,
      "step": 23950
    },
    {
      "entropy": 5.39258222579956,
      "epoch": 2.6904026506430054,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042863517219797085,
      "loss": 5.1279,
      "mean_token_accuracy": 0.20574878752231598,
      "num_tokens": 53462644.0,
      "step": 23955
    },
    {
      "entropy": 5.3428794860839846,
      "epoch": 2.690964227550963,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042860583774635165,
      "loss": 5.1363,
      "mean_token_accuracy": 0.2001335307955742,
      "num_tokens": 53473258.0,
      "step": 23960
    },
    {
      "entropy": 5.435537242889405,
      "epoch": 2.6915258044589208,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000428576498403863,
      "loss": 5.128,
      "mean_token_accuracy": 0.20437007397413254,
      "num_tokens": 53484011.0,
      "step": 23965
    },
    {
      "entropy": 5.470778512954712,
      "epoch": 2.6920873813668784,
      "grad_norm": 1.1875,
      "learning_rate": 0.000428547154171439,
      "loss": 5.168,
      "mean_token_accuracy": 0.20754805952310562,
      "num_tokens": 53495549.0,
      "step": 23970
    },
    {
      "entropy": 5.4761251449584964,
      "epoch": 2.6926489582748356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004285178050500142,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1998326763510704,
      "num_tokens": 53506721.0,
      "step": 23975
    },
    {
      "entropy": 5.546339750289917,
      "epoch": 2.6932105351827933,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004284884510405228,
      "loss": 5.3391,
      "mean_token_accuracy": 0.19034197330474853,
      "num_tokens": 53518363.0,
      "step": 23980
    },
    {
      "entropy": 5.354467725753784,
      "epoch": 2.693772112090751,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004284590921438997,
      "loss": 5.077,
      "mean_token_accuracy": 0.20418502986431122,
      "num_tokens": 53529140.0,
      "step": 23985
    },
    {
      "entropy": 5.362538194656372,
      "epoch": 2.694333688998708,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042842972836107954,
      "loss": 5.1498,
      "mean_token_accuracy": 0.19923700243234635,
      "num_tokens": 53540707.0,
      "step": 23990
    },
    {
      "entropy": 5.478306627273559,
      "epoch": 2.694895265906666,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004284003596929973,
      "loss": 5.1625,
      "mean_token_accuracy": 0.2027212679386139,
      "num_tokens": 53551297.0,
      "step": 23995
    },
    {
      "entropy": 5.385969114303589,
      "epoch": 2.6954568428146235,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042837098614058815,
      "loss": 5.0659,
      "mean_token_accuracy": 0.19970440119504929,
      "num_tokens": 53562207.0,
      "step": 24000
    },
    {
      "epoch": 2.6954568428146235,
      "eval_entropy": 5.194862051356967,
      "eval_loss": 5.287596225738525,
      "eval_mean_token_accuracy": 0.2040473656488166,
      "eval_num_tokens": 53562207.0,
      "eval_runtime": 23.7018,
      "eval_samples_per_second": 1308.465,
      "eval_steps_per_second": 163.574,
      "step": 24000
    },
    {
      "entropy": 5.4051800727844235,
      "epoch": 2.696018419722581,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042834160770478726,
      "loss": 5.2464,
      "mean_token_accuracy": 0.1972460463643074,
      "num_tokens": 53573823.0,
      "step": 24005
    },
    {
      "entropy": 5.408265018463135,
      "epoch": 2.696579996630539,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042831222438653016,
      "loss": 5.0789,
      "mean_token_accuracy": 0.214223475754261,
      "num_tokens": 53584724.0,
      "step": 24010
    },
    {
      "entropy": 5.518650007247925,
      "epoch": 2.697141573538496,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004282828361867524,
      "loss": 5.225,
      "mean_token_accuracy": 0.19809203147888182,
      "num_tokens": 53596212.0,
      "step": 24015
    },
    {
      "entropy": 5.502169561386109,
      "epoch": 2.6977031504464537,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042825344310638966,
      "loss": 5.196,
      "mean_token_accuracy": 0.2016608238220215,
      "num_tokens": 53608341.0,
      "step": 24020
    },
    {
      "entropy": 5.422328186035156,
      "epoch": 2.6982647273544114,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042822404514637785,
      "loss": 5.1809,
      "mean_token_accuracy": 0.20706566274166108,
      "num_tokens": 53619598.0,
      "step": 24025
    },
    {
      "entropy": 5.4138418674469,
      "epoch": 2.6988263042623686,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000428194642307653,
      "loss": 5.1209,
      "mean_token_accuracy": 0.20084422677755356,
      "num_tokens": 53630174.0,
      "step": 24030
    },
    {
      "entropy": 5.47109203338623,
      "epoch": 2.6993878811703262,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004281652345911513,
      "loss": 5.2088,
      "mean_token_accuracy": 0.19838862270116805,
      "num_tokens": 53641929.0,
      "step": 24035
    },
    {
      "entropy": 5.412718820571899,
      "epoch": 2.699949458078284,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042813582199780913,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20229452252388,
      "num_tokens": 53652262.0,
      "step": 24040
    },
    {
      "entropy": 5.531603908538818,
      "epoch": 2.700511034986241,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004281064045285631,
      "loss": 5.2366,
      "mean_token_accuracy": 0.2037740558385849,
      "num_tokens": 53663351.0,
      "step": 24045
    },
    {
      "entropy": 5.323830604553223,
      "epoch": 2.701072611894199,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042807698218434964,
      "loss": 5.1197,
      "mean_token_accuracy": 0.20090827345848083,
      "num_tokens": 53674632.0,
      "step": 24050
    },
    {
      "entropy": 5.365088939666748,
      "epoch": 2.7016341888021564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042804755496610576,
      "loss": 5.2035,
      "mean_token_accuracy": 0.20120566040277482,
      "num_tokens": 53686346.0,
      "step": 24055
    },
    {
      "entropy": 5.538683080673218,
      "epoch": 2.702195765710114,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042801812287476836,
      "loss": 5.2256,
      "mean_token_accuracy": 0.19962630569934844,
      "num_tokens": 53697519.0,
      "step": 24060
    },
    {
      "entropy": 5.462648534774781,
      "epoch": 2.7027573426180718,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004279886859112746,
      "loss": 5.1441,
      "mean_token_accuracy": 0.2016030043363571,
      "num_tokens": 53707858.0,
      "step": 24065
    },
    {
      "entropy": 5.376752090454102,
      "epoch": 2.703318919526029,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042795924407656175,
      "loss": 5.1269,
      "mean_token_accuracy": 0.2048349052667618,
      "num_tokens": 53719168.0,
      "step": 24070
    },
    {
      "entropy": 5.458829498291015,
      "epoch": 2.7038804964339866,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042792979737156717,
      "loss": 5.1322,
      "mean_token_accuracy": 0.20487259775400163,
      "num_tokens": 53730601.0,
      "step": 24075
    },
    {
      "entropy": 5.42026104927063,
      "epoch": 2.7044420733419443,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004279003457972286,
      "loss": 5.1294,
      "mean_token_accuracy": 0.20773958414793015,
      "num_tokens": 53741410.0,
      "step": 24080
    },
    {
      "entropy": 5.347302055358886,
      "epoch": 2.7050036502499015,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004278708893544837,
      "loss": 5.1303,
      "mean_token_accuracy": 0.20953479111194612,
      "num_tokens": 53752585.0,
      "step": 24085
    },
    {
      "entropy": 5.3899242877960205,
      "epoch": 2.705565227157859,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004278414280442704,
      "loss": 5.1339,
      "mean_token_accuracy": 0.20496701449155807,
      "num_tokens": 53762604.0,
      "step": 24090
    },
    {
      "entropy": 5.398458766937256,
      "epoch": 2.706126804065817,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042781196186752676,
      "loss": 5.1623,
      "mean_token_accuracy": 0.19578926563262938,
      "num_tokens": 53773938.0,
      "step": 24095
    },
    {
      "entropy": 5.542138624191284,
      "epoch": 2.7066883809737745,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004277824908251909,
      "loss": 5.2552,
      "mean_token_accuracy": 0.2030252441763878,
      "num_tokens": 53785142.0,
      "step": 24100
    },
    {
      "entropy": 5.551196956634522,
      "epoch": 2.707249957881732,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004277530149182014,
      "loss": 5.2576,
      "mean_token_accuracy": 0.1965601846575737,
      "num_tokens": 53797352.0,
      "step": 24105
    },
    {
      "entropy": 5.47598819732666,
      "epoch": 2.7078115347896894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042772353414749655,
      "loss": 5.1862,
      "mean_token_accuracy": 0.20036886334419252,
      "num_tokens": 53808027.0,
      "step": 24110
    },
    {
      "entropy": 5.284035110473633,
      "epoch": 2.708373111697647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042769404851401524,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2085837334394455,
      "num_tokens": 53818520.0,
      "step": 24115
    },
    {
      "entropy": 5.417389583587647,
      "epoch": 2.7089346886056047,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004276645580186961,
      "loss": 5.1264,
      "mean_token_accuracy": 0.20566834211349488,
      "num_tokens": 53829234.0,
      "step": 24120
    },
    {
      "entropy": 5.4402289390563965,
      "epoch": 2.709496265513562,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042763506266247834,
      "loss": 5.0737,
      "mean_token_accuracy": 0.20758898854255675,
      "num_tokens": 53839097.0,
      "step": 24125
    },
    {
      "entropy": 5.3463664054870605,
      "epoch": 2.7100578424215196,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004276055624463009,
      "loss": 5.133,
      "mean_token_accuracy": 0.2015272855758667,
      "num_tokens": 53850375.0,
      "step": 24130
    },
    {
      "entropy": 5.338164186477661,
      "epoch": 2.7106194193294773,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042757605737110324,
      "loss": 5.1088,
      "mean_token_accuracy": 0.21365642845630645,
      "num_tokens": 53861711.0,
      "step": 24135
    },
    {
      "entropy": 5.428396940231323,
      "epoch": 2.7111809962374345,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004275465474378246,
      "loss": 5.1527,
      "mean_token_accuracy": 0.20046297758817672,
      "num_tokens": 53872081.0,
      "step": 24140
    },
    {
      "entropy": 5.400820350646972,
      "epoch": 2.711742573145392,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004275170326474049,
      "loss": 5.1368,
      "mean_token_accuracy": 0.2025968015193939,
      "num_tokens": 53882591.0,
      "step": 24145
    },
    {
      "entropy": 5.466067361831665,
      "epoch": 2.71230415005335,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042748751300078367,
      "loss": 5.1226,
      "mean_token_accuracy": 0.21349797546863555,
      "num_tokens": 53893246.0,
      "step": 24150
    },
    {
      "entropy": 5.357647228240967,
      "epoch": 2.7128657269613075,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004274579884989008,
      "loss": 5.0641,
      "mean_token_accuracy": 0.2175005629658699,
      "num_tokens": 53904579.0,
      "step": 24155
    },
    {
      "entropy": 5.311406230926513,
      "epoch": 2.713427303869265,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042742845914269645,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20274033397436142,
      "num_tokens": 53916342.0,
      "step": 24160
    },
    {
      "entropy": 5.436368322372436,
      "epoch": 2.7139888807772223,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042739892493311095,
      "loss": 5.1318,
      "mean_token_accuracy": 0.20680449157953262,
      "num_tokens": 53927414.0,
      "step": 24165
    },
    {
      "entropy": 5.429965305328369,
      "epoch": 2.71455045768518,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042736938587108447,
      "loss": 5.1647,
      "mean_token_accuracy": 0.2029428333044052,
      "num_tokens": 53938679.0,
      "step": 24170
    },
    {
      "entropy": 5.447264575958252,
      "epoch": 2.7151120345931377,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004273398419575577,
      "loss": 5.2298,
      "mean_token_accuracy": 0.19512159377336502,
      "num_tokens": 53951364.0,
      "step": 24175
    },
    {
      "entropy": 5.542708539962769,
      "epoch": 2.715673611501095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004273102931934713,
      "loss": 5.2864,
      "mean_token_accuracy": 0.19286893159151078,
      "num_tokens": 53963590.0,
      "step": 24180
    },
    {
      "entropy": 5.554969263076782,
      "epoch": 2.7162351884090525,
      "grad_norm": 1.0625,
      "learning_rate": 0.000427280739579766,
      "loss": 5.2353,
      "mean_token_accuracy": 0.20057183802127837,
      "num_tokens": 53974315.0,
      "step": 24185
    },
    {
      "entropy": 5.442768144607544,
      "epoch": 2.71679676531701,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00042725118111738305,
      "loss": 5.2325,
      "mean_token_accuracy": 0.19446465373039246,
      "num_tokens": 53985058.0,
      "step": 24190
    },
    {
      "entropy": 5.405373668670654,
      "epoch": 2.717358342224968,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004272216178072633,
      "loss": 5.1143,
      "mean_token_accuracy": 0.2096636101603508,
      "num_tokens": 53994876.0,
      "step": 24195
    },
    {
      "entropy": 5.447998809814453,
      "epoch": 2.7179199191329255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004271920496503483,
      "loss": 5.2055,
      "mean_token_accuracy": 0.19569581300020217,
      "num_tokens": 54005705.0,
      "step": 24200
    },
    {
      "entropy": 5.489930057525635,
      "epoch": 2.7184814960408827,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042716247664757934,
      "loss": 5.1404,
      "mean_token_accuracy": 0.2035662278532982,
      "num_tokens": 54015787.0,
      "step": 24205
    },
    {
      "entropy": 5.493834733963013,
      "epoch": 2.7190430729488404,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042713289879989816,
      "loss": 5.198,
      "mean_token_accuracy": 0.19779807776212693,
      "num_tokens": 54027545.0,
      "step": 24210
    },
    {
      "entropy": 5.485737562179565,
      "epoch": 2.719604649856798,
      "grad_norm": 1.125,
      "learning_rate": 0.0004271033161082464,
      "loss": 5.2127,
      "mean_token_accuracy": 0.20279552191495895,
      "num_tokens": 54038560.0,
      "step": 24215
    },
    {
      "entropy": 5.522479152679443,
      "epoch": 2.7201662267647553,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042707372857356615,
      "loss": 5.2497,
      "mean_token_accuracy": 0.19392069578170776,
      "num_tokens": 54049286.0,
      "step": 24220
    },
    {
      "entropy": 5.415632724761963,
      "epoch": 2.720727803672713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004270441361967994,
      "loss": 5.0861,
      "mean_token_accuracy": 0.20544077455997467,
      "num_tokens": 54059894.0,
      "step": 24225
    },
    {
      "entropy": 5.477279424667358,
      "epoch": 2.7212893805806706,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004270145389788883,
      "loss": 5.2319,
      "mean_token_accuracy": 0.2013646736741066,
      "num_tokens": 54070084.0,
      "step": 24230
    },
    {
      "entropy": 5.3923492431640625,
      "epoch": 2.721850957488628,
      "grad_norm": 1.25,
      "learning_rate": 0.00042698493692077533,
      "loss": 5.1583,
      "mean_token_accuracy": 0.20531713366508483,
      "num_tokens": 54081607.0,
      "step": 24235
    },
    {
      "entropy": 5.47043924331665,
      "epoch": 2.7224125343965855,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004269553300234031,
      "loss": 5.22,
      "mean_token_accuracy": 0.20225788801908492,
      "num_tokens": 54093364.0,
      "step": 24240
    },
    {
      "entropy": 5.4724555015563965,
      "epoch": 2.722974111304543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042692571828771423,
      "loss": 5.0977,
      "mean_token_accuracy": 0.20471826791763306,
      "num_tokens": 54105199.0,
      "step": 24245
    },
    {
      "entropy": 5.394800424575806,
      "epoch": 2.723535688212501,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042689610171465147,
      "loss": 5.0583,
      "mean_token_accuracy": 0.2145303651690483,
      "num_tokens": 54116013.0,
      "step": 24250
    },
    {
      "entropy": 5.4568017482757565,
      "epoch": 2.7240972651204585,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000426866480305158,
      "loss": 5.2213,
      "mean_token_accuracy": 0.20276910215616226,
      "num_tokens": 54128457.0,
      "step": 24255
    },
    {
      "entropy": 5.385050821304321,
      "epoch": 2.7246588420284157,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004268368540601769,
      "loss": 5.1453,
      "mean_token_accuracy": 0.20343485623598098,
      "num_tokens": 54139865.0,
      "step": 24260
    },
    {
      "entropy": 5.53568058013916,
      "epoch": 2.7252204189363733,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004268072229806514,
      "loss": 5.3528,
      "mean_token_accuracy": 0.19820138067007065,
      "num_tokens": 54152264.0,
      "step": 24265
    },
    {
      "entropy": 5.527226305007934,
      "epoch": 2.725781995844331,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004267775870675251,
      "loss": 5.1967,
      "mean_token_accuracy": 0.20990683883428574,
      "num_tokens": 54164339.0,
      "step": 24270
    },
    {
      "entropy": 5.315124130249023,
      "epoch": 2.7263435727522882,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004267479463217415,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21235856413841248,
      "num_tokens": 54174041.0,
      "step": 24275
    },
    {
      "entropy": 5.432281589508056,
      "epoch": 2.726905149660246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042671830074424453,
      "loss": 5.1573,
      "mean_token_accuracy": 0.20573486536741256,
      "num_tokens": 54186457.0,
      "step": 24280
    },
    {
      "entropy": 5.431314182281494,
      "epoch": 2.7274667265682035,
      "grad_norm": 1.109375,
      "learning_rate": 0.000426688650335978,
      "loss": 5.1767,
      "mean_token_accuracy": 0.20273495614528655,
      "num_tokens": 54197217.0,
      "step": 24285
    },
    {
      "entropy": 5.458189344406128,
      "epoch": 2.728028303476161,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042665899509788593,
      "loss": 5.2006,
      "mean_token_accuracy": 0.19937868416309357,
      "num_tokens": 54209773.0,
      "step": 24290
    },
    {
      "entropy": 5.406728315353393,
      "epoch": 2.728589880384119,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004266293350309128,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20669412463903428,
      "num_tokens": 54219772.0,
      "step": 24295
    },
    {
      "entropy": 5.371670961380005,
      "epoch": 2.729151457292076,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004265996701360026,
      "loss": 5.1423,
      "mean_token_accuracy": 0.20119674801826476,
      "num_tokens": 54231222.0,
      "step": 24300
    },
    {
      "entropy": 5.515133714675903,
      "epoch": 2.7297130342000337,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042657000041410026,
      "loss": 5.1561,
      "mean_token_accuracy": 0.20618900507688523,
      "num_tokens": 54242113.0,
      "step": 24305
    },
    {
      "entropy": 5.466952514648438,
      "epoch": 2.7302746111079914,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004265403258661503,
      "loss": 5.2014,
      "mean_token_accuracy": 0.2055875301361084,
      "num_tokens": 54252073.0,
      "step": 24310
    },
    {
      "entropy": 5.438057947158813,
      "epoch": 2.7308361880159486,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004265106464930976,
      "loss": 5.2322,
      "mean_token_accuracy": 0.20073012113571168,
      "num_tokens": 54262920.0,
      "step": 24315
    },
    {
      "entropy": 5.430406808853149,
      "epoch": 2.7313977649239063,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042648096229588715,
      "loss": 5.2074,
      "mean_token_accuracy": 0.20107340812683105,
      "num_tokens": 54274525.0,
      "step": 24320
    },
    {
      "entropy": 5.485376787185669,
      "epoch": 2.731959341831864,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042645127327546405,
      "loss": 5.0958,
      "mean_token_accuracy": 0.20889943838119507,
      "num_tokens": 54284312.0,
      "step": 24325
    },
    {
      "entropy": 5.398770761489868,
      "epoch": 2.732520918739821,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042642157943277375,
      "loss": 5.1264,
      "mean_token_accuracy": 0.20519059002399445,
      "num_tokens": 54295735.0,
      "step": 24330
    },
    {
      "entropy": 5.35915174484253,
      "epoch": 2.733082495647779,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004263918807687616,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20631640553474426,
      "num_tokens": 54306435.0,
      "step": 24335
    },
    {
      "entropy": 5.379607629776001,
      "epoch": 2.7336440725557365,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004263621772843733,
      "loss": 5.0613,
      "mean_token_accuracy": 0.211759290099144,
      "num_tokens": 54316454.0,
      "step": 24340
    },
    {
      "entropy": 5.453609561920166,
      "epoch": 2.734205649463694,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004263324689805545,
      "loss": 5.2128,
      "mean_token_accuracy": 0.20510256588459014,
      "num_tokens": 54327270.0,
      "step": 24345
    },
    {
      "entropy": 5.476011133193969,
      "epoch": 2.734767226371652,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004263027558582512,
      "loss": 5.1809,
      "mean_token_accuracy": 0.2007264494895935,
      "num_tokens": 54338788.0,
      "step": 24350
    },
    {
      "entropy": 5.560351276397705,
      "epoch": 2.735328803279609,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004262730379184095,
      "loss": 5.2729,
      "mean_token_accuracy": 0.196891950070858,
      "num_tokens": 54350568.0,
      "step": 24355
    },
    {
      "entropy": 5.3842747688293455,
      "epoch": 2.7358903801875667,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042624331516197554,
      "loss": 5.1169,
      "mean_token_accuracy": 0.20631497949361802,
      "num_tokens": 54361318.0,
      "step": 24360
    },
    {
      "entropy": 5.3753941535949705,
      "epoch": 2.7364519570955244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042621358758989584,
      "loss": 5.0842,
      "mean_token_accuracy": 0.20430964827537537,
      "num_tokens": 54371548.0,
      "step": 24365
    },
    {
      "entropy": 5.460762023925781,
      "epoch": 2.7370135340034816,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004261838552031169,
      "loss": 5.3079,
      "mean_token_accuracy": 0.19245846420526505,
      "num_tokens": 54384485.0,
      "step": 24370
    },
    {
      "entropy": 5.541934776306152,
      "epoch": 2.7375751109114392,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042615411800258523,
      "loss": 5.2357,
      "mean_token_accuracy": 0.1944488063454628,
      "num_tokens": 54396155.0,
      "step": 24375
    },
    {
      "entropy": 5.3857834815979,
      "epoch": 2.738136687819397,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042612437598924793,
      "loss": 5.1721,
      "mean_token_accuracy": 0.20549881607294082,
      "num_tokens": 54406222.0,
      "step": 24380
    },
    {
      "entropy": 5.564047622680664,
      "epoch": 2.7386982647273546,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004260946291640519,
      "loss": 5.3023,
      "mean_token_accuracy": 0.19118553400039673,
      "num_tokens": 54419271.0,
      "step": 24385
    },
    {
      "entropy": 5.452545261383056,
      "epoch": 2.739259841635312,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042606487752794417,
      "loss": 5.2499,
      "mean_token_accuracy": 0.19906802326440812,
      "num_tokens": 54430517.0,
      "step": 24390
    },
    {
      "entropy": 5.474771785736084,
      "epoch": 2.7398214185432694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042603512108187224,
      "loss": 5.2031,
      "mean_token_accuracy": 0.19740011990070344,
      "num_tokens": 54442273.0,
      "step": 24395
    },
    {
      "entropy": 5.38907675743103,
      "epoch": 2.740382995451227,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042600535982678347,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20888710916042327,
      "num_tokens": 54452727.0,
      "step": 24400
    },
    {
      "entropy": 5.48522424697876,
      "epoch": 2.7409445723591848,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042597559376362544,
      "loss": 5.2001,
      "mean_token_accuracy": 0.20540001839399338,
      "num_tokens": 54463133.0,
      "step": 24405
    },
    {
      "entropy": 5.480113124847412,
      "epoch": 2.741506149267142,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004259458228933459,
      "loss": 5.2474,
      "mean_token_accuracy": 0.18812239170074463,
      "num_tokens": 54475525.0,
      "step": 24410
    },
    {
      "entropy": 5.477261924743653,
      "epoch": 2.7420677261750996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042591604721689297,
      "loss": 5.1739,
      "mean_token_accuracy": 0.20273682177066804,
      "num_tokens": 54488012.0,
      "step": 24415
    },
    {
      "entropy": 5.409173631668091,
      "epoch": 2.7426293030830573,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004258862667352145,
      "loss": 5.0831,
      "mean_token_accuracy": 0.21084702461957933,
      "num_tokens": 54499485.0,
      "step": 24420
    },
    {
      "entropy": 5.479358720779419,
      "epoch": 2.7431908799910145,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004258564814492587,
      "loss": 5.2455,
      "mean_token_accuracy": 0.20549583435058594,
      "num_tokens": 54509883.0,
      "step": 24425
    },
    {
      "entropy": 5.462230205535889,
      "epoch": 2.743752456898972,
      "grad_norm": 1.171875,
      "learning_rate": 0.000425826691359974,
      "loss": 5.2614,
      "mean_token_accuracy": 0.19527233242988587,
      "num_tokens": 54521320.0,
      "step": 24430
    },
    {
      "entropy": 5.530293273925781,
      "epoch": 2.74431403380693,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004257968964683091,
      "loss": 5.2261,
      "mean_token_accuracy": 0.19720340967178346,
      "num_tokens": 54532537.0,
      "step": 24435
    },
    {
      "entropy": 5.398940896987915,
      "epoch": 2.7448756107148875,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004257670967752125,
      "loss": 5.1112,
      "mean_token_accuracy": 0.19985959827899932,
      "num_tokens": 54542494.0,
      "step": 24440
    },
    {
      "entropy": 5.4648839950561525,
      "epoch": 2.745437187622845,
      "grad_norm": 1.015625,
      "learning_rate": 0.000425737292281633,
      "loss": 5.2677,
      "mean_token_accuracy": 0.19331849366426468,
      "num_tokens": 54553583.0,
      "step": 24445
    },
    {
      "entropy": 5.479203128814698,
      "epoch": 2.7459987645308024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004257074829885196,
      "loss": 5.1579,
      "mean_token_accuracy": 0.20669061690568924,
      "num_tokens": 54565655.0,
      "step": 24450
    },
    {
      "entropy": 5.394396018981934,
      "epoch": 2.74656034143876,
      "grad_norm": 1.0,
      "learning_rate": 0.0004256776688968216,
      "loss": 5.1509,
      "mean_token_accuracy": 0.19818404763936998,
      "num_tokens": 54578615.0,
      "step": 24455
    },
    {
      "entropy": 5.513056707382202,
      "epoch": 2.7471219183467177,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042564785000748814,
      "loss": 5.2153,
      "mean_token_accuracy": 0.2025950238108635,
      "num_tokens": 54590645.0,
      "step": 24460
    },
    {
      "entropy": 5.542813491821289,
      "epoch": 2.747683495254675,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004256180263214687,
      "loss": 5.3139,
      "mean_token_accuracy": 0.1915415719151497,
      "num_tokens": 54602326.0,
      "step": 24465
    },
    {
      "entropy": 5.5383281230926515,
      "epoch": 2.7482450721626326,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042558819783971295,
      "loss": 5.3005,
      "mean_token_accuracy": 0.18920648097991943,
      "num_tokens": 54614008.0,
      "step": 24470
    },
    {
      "entropy": 5.455666208267212,
      "epoch": 2.7488066490705902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004255583645631705,
      "loss": 5.1417,
      "mean_token_accuracy": 0.2042185351252556,
      "num_tokens": 54625482.0,
      "step": 24475
    },
    {
      "entropy": 5.589465379714966,
      "epoch": 2.749368225978548,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004255285264927913,
      "loss": 5.2311,
      "mean_token_accuracy": 0.19621275812387468,
      "num_tokens": 54636423.0,
      "step": 24480
    },
    {
      "entropy": 5.4568106174469,
      "epoch": 2.7499298028865056,
      "grad_norm": 1.0,
      "learning_rate": 0.00042549868362952556,
      "loss": 5.2039,
      "mean_token_accuracy": 0.1975671634078026,
      "num_tokens": 54648152.0,
      "step": 24485
    },
    {
      "entropy": 5.430043792724609,
      "epoch": 2.750491379794463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042546883597432327,
      "loss": 5.2158,
      "mean_token_accuracy": 0.20439807176589966,
      "num_tokens": 54661141.0,
      "step": 24490
    },
    {
      "entropy": 5.511838674545288,
      "epoch": 2.7510529567024204,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042543898352813485,
      "loss": 5.1964,
      "mean_token_accuracy": 0.19709533900022508,
      "num_tokens": 54671180.0,
      "step": 24495
    },
    {
      "entropy": 5.349125671386719,
      "epoch": 2.751614533610378,
      "grad_norm": 1.15625,
      "learning_rate": 0.000425409126291911,
      "loss": 5.0621,
      "mean_token_accuracy": 0.21101489961147307,
      "num_tokens": 54681451.0,
      "step": 24500
    },
    {
      "entropy": 5.3607542514801025,
      "epoch": 2.7521761105183353,
      "grad_norm": 1.125,
      "learning_rate": 0.0004253792642666021,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20588713139295578,
      "num_tokens": 54691760.0,
      "step": 24505
    },
    {
      "entropy": 5.473189449310302,
      "epoch": 2.752737687426293,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042534939745315914,
      "loss": 5.1624,
      "mean_token_accuracy": 0.1964871421456337,
      "num_tokens": 54703761.0,
      "step": 24510
    },
    {
      "entropy": 5.482636880874634,
      "epoch": 2.7532992643342507,
      "grad_norm": 1.046875,
      "learning_rate": 0.000425319525852533,
      "loss": 5.22,
      "mean_token_accuracy": 0.2023416668176651,
      "num_tokens": 54715495.0,
      "step": 24515
    },
    {
      "entropy": 5.391024398803711,
      "epoch": 2.753860841242208,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004252896494656749,
      "loss": 5.102,
      "mean_token_accuracy": 0.2031466171145439,
      "num_tokens": 54727061.0,
      "step": 24520
    },
    {
      "entropy": 5.445971393585205,
      "epoch": 2.7544224181501655,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004252597682935362,
      "loss": 5.0887,
      "mean_token_accuracy": 0.20333671867847442,
      "num_tokens": 54738368.0,
      "step": 24525
    },
    {
      "entropy": 5.421975088119507,
      "epoch": 2.754983995058123,
      "grad_norm": 1.03125,
      "learning_rate": 0.000425229882337068,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19622305482625962,
      "num_tokens": 54749372.0,
      "step": 24530
    },
    {
      "entropy": 5.494048547744751,
      "epoch": 2.755545571966081,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042519999159722216,
      "loss": 5.2869,
      "mean_token_accuracy": 0.19349543750286102,
      "num_tokens": 54761309.0,
      "step": 24535
    },
    {
      "entropy": 5.461278772354126,
      "epoch": 2.7561071488740385,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042517009607495035,
      "loss": 5.2037,
      "mean_token_accuracy": 0.19828810840845107,
      "num_tokens": 54773072.0,
      "step": 24540
    },
    {
      "entropy": 5.431031656265259,
      "epoch": 2.7566687257819957,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042514019577120444,
      "loss": 5.1438,
      "mean_token_accuracy": 0.2016252487897873,
      "num_tokens": 54783603.0,
      "step": 24545
    },
    {
      "entropy": 5.4024372577667235,
      "epoch": 2.7572303026899534,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004251102906869364,
      "loss": 5.1767,
      "mean_token_accuracy": 0.2045640081167221,
      "num_tokens": 54794706.0,
      "step": 24550
    },
    {
      "entropy": 5.5191004276275635,
      "epoch": 2.757791879597911,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004250803808230986,
      "loss": 5.2611,
      "mean_token_accuracy": 0.20214680433273316,
      "num_tokens": 54804023.0,
      "step": 24555
    },
    {
      "entropy": 5.40861029624939,
      "epoch": 2.7583534565058683,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042505046618064314,
      "loss": 5.1336,
      "mean_token_accuracy": 0.20835502743721007,
      "num_tokens": 54813616.0,
      "step": 24560
    },
    {
      "entropy": 5.432675838470459,
      "epoch": 2.758915033413826,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042502054676052273,
      "loss": 5.1367,
      "mean_token_accuracy": 0.20537097603082657,
      "num_tokens": 54825876.0,
      "step": 24565
    },
    {
      "entropy": 5.537777423858643,
      "epoch": 2.7594766103217836,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042499062256368987,
      "loss": 5.2578,
      "mean_token_accuracy": 0.1987195372581482,
      "num_tokens": 54837207.0,
      "step": 24570
    },
    {
      "entropy": 5.469197559356689,
      "epoch": 2.7600381872297413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004249606935910975,
      "loss": 5.1601,
      "mean_token_accuracy": 0.20586324781179427,
      "num_tokens": 54849155.0,
      "step": 24575
    },
    {
      "entropy": 5.402039575576782,
      "epoch": 2.760599764137699,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004249307598436984,
      "loss": 5.1396,
      "mean_token_accuracy": 0.20521221607923507,
      "num_tokens": 54860345.0,
      "step": 24580
    },
    {
      "entropy": 5.354141473770142,
      "epoch": 2.761161341045656,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004249008213224458,
      "loss": 5.14,
      "mean_token_accuracy": 0.20254168808460235,
      "num_tokens": 54871676.0,
      "step": 24585
    },
    {
      "entropy": 5.460025930404663,
      "epoch": 2.761722917953614,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042487087802829275,
      "loss": 5.1635,
      "mean_token_accuracy": 0.2018424779176712,
      "num_tokens": 54882352.0,
      "step": 24590
    },
    {
      "entropy": 5.544378995895386,
      "epoch": 2.7622844948615715,
      "grad_norm": 1.125,
      "learning_rate": 0.0004248409299621931,
      "loss": 5.1874,
      "mean_token_accuracy": 0.2013700008392334,
      "num_tokens": 54893862.0,
      "step": 24595
    },
    {
      "entropy": 5.3760504722595215,
      "epoch": 2.7628460717695287,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004248109771251,
      "loss": 5.1472,
      "mean_token_accuracy": 0.2054835945367813,
      "num_tokens": 54904851.0,
      "step": 24600
    },
    {
      "entropy": 5.462527465820313,
      "epoch": 2.7634076486774863,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042478101951796716,
      "loss": 5.2431,
      "mean_token_accuracy": 0.20127769112586974,
      "num_tokens": 54916472.0,
      "step": 24605
    },
    {
      "entropy": 5.467825508117675,
      "epoch": 2.763969225585444,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004247510571417487,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20311488658189775,
      "num_tokens": 54927383.0,
      "step": 24610
    },
    {
      "entropy": 5.434236478805542,
      "epoch": 2.764530802493401,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004247210899973985,
      "loss": 5.1581,
      "mean_token_accuracy": 0.20460018813610076,
      "num_tokens": 54938157.0,
      "step": 24615
    },
    {
      "entropy": 5.448858165740967,
      "epoch": 2.765092379401359,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004246911180858707,
      "loss": 5.1867,
      "mean_token_accuracy": 0.2060574248433113,
      "num_tokens": 54949812.0,
      "step": 24620
    },
    {
      "entropy": 5.5344466209411625,
      "epoch": 2.7656539563093165,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004246611414081197,
      "loss": 5.2626,
      "mean_token_accuracy": 0.19349587261676787,
      "num_tokens": 54961201.0,
      "step": 24625
    },
    {
      "entropy": 5.4527263164520265,
      "epoch": 2.766215533217274,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042463115996509984,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20951994955539704,
      "num_tokens": 54971071.0,
      "step": 24630
    },
    {
      "entropy": 5.421306180953979,
      "epoch": 2.766777110125232,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004246011737577658,
      "loss": 5.2182,
      "mean_token_accuracy": 0.1998912960290909,
      "num_tokens": 54981294.0,
      "step": 24635
    },
    {
      "entropy": 5.47854471206665,
      "epoch": 2.767338687033189,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004245711827870724,
      "loss": 5.2554,
      "mean_token_accuracy": 0.20032831579446791,
      "num_tokens": 54994071.0,
      "step": 24640
    },
    {
      "entropy": 5.498867225646973,
      "epoch": 2.7679002639411467,
      "grad_norm": 1.125,
      "learning_rate": 0.00042454118705397454,
      "loss": 5.2455,
      "mean_token_accuracy": 0.20654832273721696,
      "num_tokens": 55005087.0,
      "step": 24645
    },
    {
      "entropy": 5.435654926300049,
      "epoch": 2.7684618408491044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004245111865594273,
      "loss": 5.091,
      "mean_token_accuracy": 0.20651579052209854,
      "num_tokens": 55016103.0,
      "step": 24650
    },
    {
      "entropy": 5.445204162597657,
      "epoch": 2.7690234177570616,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042448118130438583,
      "loss": 5.2424,
      "mean_token_accuracy": 0.1959297001361847,
      "num_tokens": 55028656.0,
      "step": 24655
    },
    {
      "entropy": 5.49377408027649,
      "epoch": 2.7695849946650193,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042445117128980553,
      "loss": 5.2916,
      "mean_token_accuracy": 0.19811236411333083,
      "num_tokens": 55038957.0,
      "step": 24660
    },
    {
      "entropy": 5.47159218788147,
      "epoch": 2.770146571572977,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042442115651664217,
      "loss": 5.1141,
      "mean_token_accuracy": 0.20781386643648148,
      "num_tokens": 55050407.0,
      "step": 24665
    },
    {
      "entropy": 5.427448463439942,
      "epoch": 2.7707081484809346,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042439113698585103,
      "loss": 5.1858,
      "mean_token_accuracy": 0.19719309806823732,
      "num_tokens": 55060119.0,
      "step": 24670
    },
    {
      "entropy": 5.427186822891235,
      "epoch": 2.7712697253888923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004243611126983882,
      "loss": 5.1951,
      "mean_token_accuracy": 0.20210651755332948,
      "num_tokens": 55072059.0,
      "step": 24675
    },
    {
      "entropy": 5.387882089614868,
      "epoch": 2.7718313022968495,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042433108365520967,
      "loss": 5.1569,
      "mean_token_accuracy": 0.20163421630859374,
      "num_tokens": 55083640.0,
      "step": 24680
    },
    {
      "entropy": 5.457586240768433,
      "epoch": 2.772392879204807,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042430104985727155,
      "loss": 5.1166,
      "mean_token_accuracy": 0.2119431287050247,
      "num_tokens": 55093014.0,
      "step": 24685
    },
    {
      "entropy": 5.38799786567688,
      "epoch": 2.772954456112765,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042427101130553,
      "loss": 5.1569,
      "mean_token_accuracy": 0.20743077993392944,
      "num_tokens": 55102795.0,
      "step": 24690
    },
    {
      "entropy": 5.420719480514526,
      "epoch": 2.773516033020722,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004242409680009416,
      "loss": 5.2346,
      "mean_token_accuracy": 0.19611083716154099,
      "num_tokens": 55114389.0,
      "step": 24695
    },
    {
      "entropy": 5.556324434280396,
      "epoch": 2.7740776099286797,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004242109199444629,
      "loss": 5.237,
      "mean_token_accuracy": 0.2050292655825615,
      "num_tokens": 55125078.0,
      "step": 24700
    },
    {
      "entropy": 5.4520574569702145,
      "epoch": 2.7746391868366374,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004241808671370506,
      "loss": 5.1607,
      "mean_token_accuracy": 0.196464604139328,
      "num_tokens": 55135151.0,
      "step": 24705
    },
    {
      "entropy": 5.427392053604126,
      "epoch": 2.7752007637445946,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042415080957966176,
      "loss": 5.1868,
      "mean_token_accuracy": 0.20725896805524827,
      "num_tokens": 55146330.0,
      "step": 24710
    },
    {
      "entropy": 5.440537643432617,
      "epoch": 2.7757623406525522,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004241207472732531,
      "loss": 5.1428,
      "mean_token_accuracy": 0.2053358346223831,
      "num_tokens": 55156636.0,
      "step": 24715
    },
    {
      "entropy": 5.469632482528686,
      "epoch": 2.77632391756051,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004240906802187821,
      "loss": 5.2356,
      "mean_token_accuracy": 0.19819016456604005,
      "num_tokens": 55168761.0,
      "step": 24720
    },
    {
      "entropy": 5.3847033977508545,
      "epoch": 2.7768854944684676,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004240606084172061,
      "loss": 5.1208,
      "mean_token_accuracy": 0.20449323952198029,
      "num_tokens": 55180747.0,
      "step": 24725
    },
    {
      "entropy": 5.516995143890381,
      "epoch": 2.777447071376425,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042403053186948237,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19401783496141434,
      "num_tokens": 55191645.0,
      "step": 24730
    },
    {
      "entropy": 5.450952529907227,
      "epoch": 2.7780086482843824,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004240004505765688,
      "loss": 5.143,
      "mean_token_accuracy": 0.1957264080643654,
      "num_tokens": 55203623.0,
      "step": 24735
    },
    {
      "entropy": 5.467602729797363,
      "epoch": 2.77857022519234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004239703645394231,
      "loss": 5.2687,
      "mean_token_accuracy": 0.1954276129603386,
      "num_tokens": 55214635.0,
      "step": 24740
    },
    {
      "entropy": 5.367819738388062,
      "epoch": 2.7791318021002978,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004239402737590032,
      "loss": 5.1233,
      "mean_token_accuracy": 0.2016369730234146,
      "num_tokens": 55225524.0,
      "step": 24745
    },
    {
      "entropy": 5.42220139503479,
      "epoch": 2.779693379008255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042391017823626716,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20959751754999162,
      "num_tokens": 55238360.0,
      "step": 24750
    },
    {
      "entropy": 5.53945689201355,
      "epoch": 2.7802549559162126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004238800779721734,
      "loss": 5.2535,
      "mean_token_accuracy": 0.19526060670614243,
      "num_tokens": 55248524.0,
      "step": 24755
    },
    {
      "entropy": 5.515944337844848,
      "epoch": 2.7808165328241703,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042384997296768007,
      "loss": 5.2762,
      "mean_token_accuracy": 0.20397339910268783,
      "num_tokens": 55260283.0,
      "step": 24760
    },
    {
      "entropy": 5.531018781661987,
      "epoch": 2.781378109732128,
      "grad_norm": 1.015625,
      "learning_rate": 0.000423819863223746,
      "loss": 5.1891,
      "mean_token_accuracy": 0.1965341955423355,
      "num_tokens": 55271928.0,
      "step": 24765
    },
    {
      "entropy": 5.462970781326294,
      "epoch": 2.7819396866400856,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042378974874132967,
      "loss": 5.2081,
      "mean_token_accuracy": 0.20109081268310547,
      "num_tokens": 55283015.0,
      "step": 24770
    },
    {
      "entropy": 5.423681545257568,
      "epoch": 2.782501263548043,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004237596295213901,
      "loss": 5.0746,
      "mean_token_accuracy": 0.2074076786637306,
      "num_tokens": 55293918.0,
      "step": 24775
    },
    {
      "entropy": 5.341319894790649,
      "epoch": 2.7830628404560005,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004237295055648862,
      "loss": 5.1051,
      "mean_token_accuracy": 0.1980103299021721,
      "num_tokens": 55305808.0,
      "step": 24780
    },
    {
      "entropy": 5.3905096530914305,
      "epoch": 2.783624417363958,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042369937687277716,
      "loss": 5.213,
      "mean_token_accuracy": 0.19633498042821884,
      "num_tokens": 55318128.0,
      "step": 24785
    },
    {
      "entropy": 5.502759122848511,
      "epoch": 2.7841859942719154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004236692434460223,
      "loss": 5.2413,
      "mean_token_accuracy": 0.20149529129266738,
      "num_tokens": 55329934.0,
      "step": 24790
    },
    {
      "entropy": 5.5992615699768065,
      "epoch": 2.784747571179873,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004236391052855811,
      "loss": 5.3615,
      "mean_token_accuracy": 0.19547825306653976,
      "num_tokens": 55340971.0,
      "step": 24795
    },
    {
      "entropy": 5.490921115875244,
      "epoch": 2.7853091480878307,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042360896239241306,
      "loss": 5.2128,
      "mean_token_accuracy": 0.19996569454669952,
      "num_tokens": 55352243.0,
      "step": 24800
    },
    {
      "entropy": 5.519829750061035,
      "epoch": 2.785870724995788,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042357881476747807,
      "loss": 5.1208,
      "mean_token_accuracy": 0.20832205414772034,
      "num_tokens": 55363694.0,
      "step": 24805
    },
    {
      "entropy": 5.451301383972168,
      "epoch": 2.7864323019037456,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000423548662411736,
      "loss": 5.3156,
      "mean_token_accuracy": 0.20221443474292755,
      "num_tokens": 55375720.0,
      "step": 24810
    },
    {
      "entropy": 5.496734142303467,
      "epoch": 2.7869938788117032,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004235185053261469,
      "loss": 5.1485,
      "mean_token_accuracy": 0.2062526300549507,
      "num_tokens": 55387493.0,
      "step": 24815
    },
    {
      "entropy": 5.519847202301025,
      "epoch": 2.787555455719661,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000423488343511671,
      "loss": 5.3306,
      "mean_token_accuracy": 0.19310317188501358,
      "num_tokens": 55399468.0,
      "step": 24820
    },
    {
      "entropy": 5.447811031341553,
      "epoch": 2.7881170326276186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004234581769692686,
      "loss": 5.1696,
      "mean_token_accuracy": 0.19211531728506087,
      "num_tokens": 55411526.0,
      "step": 24825
    },
    {
      "entropy": 5.483819770812988,
      "epoch": 2.788678609535576,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004234280056999003,
      "loss": 5.1236,
      "mean_token_accuracy": 0.20535043776035308,
      "num_tokens": 55422550.0,
      "step": 24830
    },
    {
      "entropy": 5.4230273246765135,
      "epoch": 2.7892401864435334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004233978297045267,
      "loss": 5.1067,
      "mean_token_accuracy": 0.20954055935144425,
      "num_tokens": 55433433.0,
      "step": 24835
    },
    {
      "entropy": 5.3980916976928714,
      "epoch": 2.789801763351491,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004233676489841087,
      "loss": 5.1269,
      "mean_token_accuracy": 0.21044438779354097,
      "num_tokens": 55444225.0,
      "step": 24840
    },
    {
      "entropy": 5.51431565284729,
      "epoch": 2.7903633402594483,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042333746353960723,
      "loss": 5.2903,
      "mean_token_accuracy": 0.19892845451831817,
      "num_tokens": 55456117.0,
      "step": 24845
    },
    {
      "entropy": 5.438743543624878,
      "epoch": 2.790924917167406,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004233072733719833,
      "loss": 5.1241,
      "mean_token_accuracy": 0.20516521036624907,
      "num_tokens": 55467267.0,
      "step": 24850
    },
    {
      "entropy": 5.476968145370483,
      "epoch": 2.7914864940753636,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004232770784821984,
      "loss": 5.1901,
      "mean_token_accuracy": 0.2051390990614891,
      "num_tokens": 55478520.0,
      "step": 24855
    },
    {
      "entropy": 5.473791837692261,
      "epoch": 2.7920480709833213,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042324687887121366,
      "loss": 5.2644,
      "mean_token_accuracy": 0.1945769667625427,
      "num_tokens": 55489704.0,
      "step": 24860
    },
    {
      "entropy": 5.43437180519104,
      "epoch": 2.792609647891279,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004232166745399909,
      "loss": 5.1219,
      "mean_token_accuracy": 0.20721485018730162,
      "num_tokens": 55500942.0,
      "step": 24865
    },
    {
      "entropy": 5.503784704208374,
      "epoch": 2.793171224799236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042318646548949177,
      "loss": 5.3195,
      "mean_token_accuracy": 0.18893152624368667,
      "num_tokens": 55512691.0,
      "step": 24870
    },
    {
      "entropy": 5.446723985671997,
      "epoch": 2.793732801707194,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042315625172067794,
      "loss": 5.1081,
      "mean_token_accuracy": 0.20965508818626405,
      "num_tokens": 55523827.0,
      "step": 24875
    },
    {
      "entropy": 5.516996383666992,
      "epoch": 2.7942943786151515,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042312603323451173,
      "loss": 5.2785,
      "mean_token_accuracy": 0.197563299536705,
      "num_tokens": 55534674.0,
      "step": 24880
    },
    {
      "entropy": 5.419544696807861,
      "epoch": 2.7948559555231087,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004230958100319551,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20707162618637084,
      "num_tokens": 55545504.0,
      "step": 24885
    },
    {
      "entropy": 5.448154258728027,
      "epoch": 2.7954175324310664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004230655821139705,
      "loss": 5.2915,
      "mean_token_accuracy": 0.19090787768363954,
      "num_tokens": 55557818.0,
      "step": 24890
    },
    {
      "entropy": 5.486776971817017,
      "epoch": 2.795979109339024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004230353494815203,
      "loss": 5.1838,
      "mean_token_accuracy": 0.20325971841812135,
      "num_tokens": 55568143.0,
      "step": 24895
    },
    {
      "entropy": 5.546880197525025,
      "epoch": 2.7965406862469813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004230051121355672,
      "loss": 5.2503,
      "mean_token_accuracy": 0.19601887911558152,
      "num_tokens": 55579358.0,
      "step": 24900
    },
    {
      "entropy": 5.50674877166748,
      "epoch": 2.797102263154939,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042297487007707387,
      "loss": 5.3473,
      "mean_token_accuracy": 0.19654758274555206,
      "num_tokens": 55590430.0,
      "step": 24905
    },
    {
      "entropy": 5.483537340164185,
      "epoch": 2.7976638400628966,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042294462330700336,
      "loss": 5.2428,
      "mean_token_accuracy": 0.20912528932094573,
      "num_tokens": 55602377.0,
      "step": 24910
    },
    {
      "entropy": 5.468552923202514,
      "epoch": 2.7982254169708543,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004229143718263186,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20369914919137955,
      "num_tokens": 55613261.0,
      "step": 24915
    },
    {
      "entropy": 5.327735805511475,
      "epoch": 2.798786993878812,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004228841156359829,
      "loss": 5.0938,
      "mean_token_accuracy": 0.2024439677596092,
      "num_tokens": 55624514.0,
      "step": 24920
    },
    {
      "entropy": 5.448320102691651,
      "epoch": 2.799348570786769,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004228538547369596,
      "loss": 5.1999,
      "mean_token_accuracy": 0.19834371358156205,
      "num_tokens": 55635899.0,
      "step": 24925
    },
    {
      "entropy": 5.513878583908081,
      "epoch": 2.799910147694727,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042282358913021224,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20321768671274185,
      "num_tokens": 55646807.0,
      "step": 24930
    },
    {
      "entropy": 5.523096418380737,
      "epoch": 2.8004717246026845,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042279331881670446,
      "loss": 5.1998,
      "mean_token_accuracy": 0.2060458019375801,
      "num_tokens": 55658937.0,
      "step": 24935
    },
    {
      "entropy": 5.356009674072266,
      "epoch": 2.8010333015106417,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004227630437974,
      "loss": 5.133,
      "mean_token_accuracy": 0.20822166949510573,
      "num_tokens": 55670216.0,
      "step": 24940
    },
    {
      "entropy": 5.531375932693481,
      "epoch": 2.8015948784185993,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042273276407326304,
      "loss": 5.2198,
      "mean_token_accuracy": 0.2031826853752136,
      "num_tokens": 55682743.0,
      "step": 24945
    },
    {
      "entropy": 5.441863584518432,
      "epoch": 2.802156455326557,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042270247964525755,
      "loss": 5.1648,
      "mean_token_accuracy": 0.19785788506269456,
      "num_tokens": 55694338.0,
      "step": 24950
    },
    {
      "entropy": 5.470470380783081,
      "epoch": 2.8027180322345147,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004226721905143477,
      "loss": 5.1742,
      "mean_token_accuracy": 0.19994216561317443,
      "num_tokens": 55704224.0,
      "step": 24955
    },
    {
      "entropy": 5.373088026046753,
      "epoch": 2.8032796091424723,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042264189668149817,
      "loss": 5.1347,
      "mean_token_accuracy": 0.19948941916227342,
      "num_tokens": 55714901.0,
      "step": 24960
    },
    {
      "entropy": 5.505374574661255,
      "epoch": 2.8038411860504295,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004226115981476733,
      "loss": 5.2881,
      "mean_token_accuracy": 0.1947023719549179,
      "num_tokens": 55725020.0,
      "step": 24965
    },
    {
      "entropy": 5.401102304458618,
      "epoch": 2.804402762958387,
      "grad_norm": 1.09375,
      "learning_rate": 0.000422581294913838,
      "loss": 5.0776,
      "mean_token_accuracy": 0.2135257124900818,
      "num_tokens": 55735643.0,
      "step": 24970
    },
    {
      "entropy": 5.4199991703033445,
      "epoch": 2.804964339866345,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000422550986980957,
      "loss": 5.1087,
      "mean_token_accuracy": 0.20003241300582886,
      "num_tokens": 55745778.0,
      "step": 24975
    },
    {
      "entropy": 5.362207269668579,
      "epoch": 2.805525916774302,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042252067434999524,
      "loss": 5.1466,
      "mean_token_accuracy": 0.20104152262210845,
      "num_tokens": 55757055.0,
      "step": 24980
    },
    {
      "entropy": 5.427798843383789,
      "epoch": 2.8060874936822597,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004224903570219181,
      "loss": 5.2101,
      "mean_token_accuracy": 0.2012692242860794,
      "num_tokens": 55768193.0,
      "step": 24985
    },
    {
      "entropy": 5.417451953887939,
      "epoch": 2.8066490705902174,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042246003499769077,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20490931570529938,
      "num_tokens": 55779382.0,
      "step": 24990
    },
    {
      "entropy": 5.422210693359375,
      "epoch": 2.8072106474981746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042242970827827874,
      "loss": 5.0795,
      "mean_token_accuracy": 0.2109823003411293,
      "num_tokens": 55790933.0,
      "step": 24995
    },
    {
      "entropy": 5.334620380401612,
      "epoch": 2.8077722244061323,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004223993768646476,
      "loss": 5.0594,
      "mean_token_accuracy": 0.2111021652817726,
      "num_tokens": 55803101.0,
      "step": 25000
    },
    {
      "entropy": 5.331802034378052,
      "epoch": 2.80833380131409,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004223690407577631,
      "loss": 5.0746,
      "mean_token_accuracy": 0.212496118247509,
      "num_tokens": 55814651.0,
      "step": 25005
    },
    {
      "entropy": 5.497640895843506,
      "epoch": 2.8088953782220476,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042233869995859127,
      "loss": 5.1834,
      "mean_token_accuracy": 0.20063014328479767,
      "num_tokens": 55825696.0,
      "step": 25010
    },
    {
      "entropy": 5.432696771621704,
      "epoch": 2.8094569551300053,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042230835446809803,
      "loss": 5.1707,
      "mean_token_accuracy": 0.200192953646183,
      "num_tokens": 55836679.0,
      "step": 25015
    },
    {
      "entropy": 5.407367849349976,
      "epoch": 2.8100185320379625,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004222780042872496,
      "loss": 5.1994,
      "mean_token_accuracy": 0.1965168073773384,
      "num_tokens": 55848013.0,
      "step": 25020
    },
    {
      "entropy": 5.3583556652069095,
      "epoch": 2.81058010894592,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004222476494170125,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20998100191354752,
      "num_tokens": 55859310.0,
      "step": 25025
    },
    {
      "entropy": 5.514933061599732,
      "epoch": 2.811141685853878,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004222172898583531,
      "loss": 5.2382,
      "mean_token_accuracy": 0.1895426630973816,
      "num_tokens": 55870484.0,
      "step": 25030
    },
    {
      "entropy": 5.513293933868408,
      "epoch": 2.811703262761835,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004221869256122381,
      "loss": 5.2151,
      "mean_token_accuracy": 0.19783167093992232,
      "num_tokens": 55881279.0,
      "step": 25035
    },
    {
      "entropy": 5.3786280155181885,
      "epoch": 2.8122648396697927,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004221565566796343,
      "loss": 5.0788,
      "mean_token_accuracy": 0.19985860735177993,
      "num_tokens": 55893029.0,
      "step": 25040
    },
    {
      "entropy": 5.315722179412842,
      "epoch": 2.8128264165777503,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004221261830615086,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20996863394975662,
      "num_tokens": 55904171.0,
      "step": 25045
    },
    {
      "entropy": 5.4315637111663815,
      "epoch": 2.813387993485708,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004220958047588282,
      "loss": 5.1348,
      "mean_token_accuracy": 0.20762409120798112,
      "num_tokens": 55916121.0,
      "step": 25050
    },
    {
      "entropy": 5.4339852809906,
      "epoch": 2.8139495703936657,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004220654217725603,
      "loss": 5.1923,
      "mean_token_accuracy": 0.19616778194904327,
      "num_tokens": 55925982.0,
      "step": 25055
    },
    {
      "entropy": 5.462462663650513,
      "epoch": 2.814511147301623,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042203503410367244,
      "loss": 5.146,
      "mean_token_accuracy": 0.20573732405900955,
      "num_tokens": 55935768.0,
      "step": 25060
    },
    {
      "entropy": 5.497834157943726,
      "epoch": 2.8150727242095805,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000422004641753132,
      "loss": 5.2097,
      "mean_token_accuracy": 0.20589301735162735,
      "num_tokens": 55946742.0,
      "step": 25065
    },
    {
      "entropy": 5.418813991546631,
      "epoch": 2.815634301117538,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004219742447219068,
      "loss": 5.1077,
      "mean_token_accuracy": 0.21268148571252823,
      "num_tokens": 55957133.0,
      "step": 25070
    },
    {
      "entropy": 5.433410930633545,
      "epoch": 2.8161958780254954,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042194384301096457,
      "loss": 5.1562,
      "mean_token_accuracy": 0.20142244249582292,
      "num_tokens": 55967179.0,
      "step": 25075
    },
    {
      "entropy": 5.360621500015259,
      "epoch": 2.816757454933453,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042191343662127337,
      "loss": 5.0507,
      "mean_token_accuracy": 0.2105631411075592,
      "num_tokens": 55977713.0,
      "step": 25080
    },
    {
      "entropy": 5.489274215698242,
      "epoch": 2.8173190318414107,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042188302555380144,
      "loss": 5.1948,
      "mean_token_accuracy": 0.20118847936391832,
      "num_tokens": 55988650.0,
      "step": 25085
    },
    {
      "entropy": 5.402718782424927,
      "epoch": 2.817880608749368,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000421852609809517,
      "loss": 5.0596,
      "mean_token_accuracy": 0.20768651068210603,
      "num_tokens": 55999745.0,
      "step": 25090
    },
    {
      "entropy": 5.366912603378296,
      "epoch": 2.8184421856573256,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004218221893893884,
      "loss": 5.099,
      "mean_token_accuracy": 0.2072529226541519,
      "num_tokens": 56011171.0,
      "step": 25095
    },
    {
      "entropy": 5.384856176376343,
      "epoch": 2.8190037625652833,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042179176429438454,
      "loss": 5.132,
      "mean_token_accuracy": 0.21273254007101058,
      "num_tokens": 56022121.0,
      "step": 25100
    },
    {
      "entropy": 5.404669284820557,
      "epoch": 2.819565339473241,
      "grad_norm": 1.125,
      "learning_rate": 0.0004217613345254739,
      "loss": 5.1446,
      "mean_token_accuracy": 0.21067192554473876,
      "num_tokens": 56033769.0,
      "step": 25105
    },
    {
      "entropy": 5.45750880241394,
      "epoch": 2.8201269163811986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042173090008362535,
      "loss": 5.0999,
      "mean_token_accuracy": 0.20771268904209136,
      "num_tokens": 56044691.0,
      "step": 25110
    },
    {
      "entropy": 5.367463684082031,
      "epoch": 2.820688493289156,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042170046096980807,
      "loss": 5.0686,
      "mean_token_accuracy": 0.20814388394355773,
      "num_tokens": 56055639.0,
      "step": 25115
    },
    {
      "entropy": 5.4927280902862545,
      "epoch": 2.8212500701971135,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042167001718499124,
      "loss": 5.2024,
      "mean_token_accuracy": 0.2019195944070816,
      "num_tokens": 56066579.0,
      "step": 25120
    },
    {
      "entropy": 5.459292221069336,
      "epoch": 2.821811647105071,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004216395687301442,
      "loss": 5.1769,
      "mean_token_accuracy": 0.19727346301078796,
      "num_tokens": 56076871.0,
      "step": 25125
    },
    {
      "entropy": 5.488662195205689,
      "epoch": 2.8223732240130284,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004216091156062364,
      "loss": 5.2656,
      "mean_token_accuracy": 0.196015864610672,
      "num_tokens": 56090022.0,
      "step": 25130
    },
    {
      "entropy": 5.302827835083008,
      "epoch": 2.822934800920986,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004215786578142375,
      "loss": 4.9554,
      "mean_token_accuracy": 0.21522179394960403,
      "num_tokens": 56100545.0,
      "step": 25135
    },
    {
      "entropy": 5.415987920761109,
      "epoch": 2.8234963778289437,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042154819535511733,
      "loss": 5.2572,
      "mean_token_accuracy": 0.19520153850317,
      "num_tokens": 56112059.0,
      "step": 25140
    },
    {
      "entropy": 5.5209368705749515,
      "epoch": 2.8240579547369014,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042151772822984567,
      "loss": 5.2019,
      "mean_token_accuracy": 0.20613405853509903,
      "num_tokens": 56123162.0,
      "step": 25145
    },
    {
      "entropy": 5.592617082595825,
      "epoch": 2.824619531644859,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004214872564393928,
      "loss": 5.2881,
      "mean_token_accuracy": 0.2005797505378723,
      "num_tokens": 56136008.0,
      "step": 25150
    },
    {
      "entropy": 5.370129632949829,
      "epoch": 2.8251811085528162,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004214567799847289,
      "loss": 5.1449,
      "mean_token_accuracy": 0.20419301241636276,
      "num_tokens": 56147169.0,
      "step": 25155
    },
    {
      "entropy": 5.4306810855865475,
      "epoch": 2.825742685460774,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004214262988668242,
      "loss": 5.1043,
      "mean_token_accuracy": 0.19814843237400054,
      "num_tokens": 56157644.0,
      "step": 25160
    },
    {
      "entropy": 5.543752861022949,
      "epoch": 2.8263042623687316,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004213958130866494,
      "loss": 5.2053,
      "mean_token_accuracy": 0.19652559608221054,
      "num_tokens": 56168248.0,
      "step": 25165
    },
    {
      "entropy": 5.474974918365478,
      "epoch": 2.8268658392766888,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042136532264517527,
      "loss": 5.2433,
      "mean_token_accuracy": 0.19865178614854812,
      "num_tokens": 56179140.0,
      "step": 25170
    },
    {
      "entropy": 5.4746476173400875,
      "epoch": 2.8274274161846464,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042133482754337236,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20108282566070557,
      "num_tokens": 56190184.0,
      "step": 25175
    },
    {
      "entropy": 5.524678134918213,
      "epoch": 2.827988993092604,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042130432778221186,
      "loss": 5.2202,
      "mean_token_accuracy": 0.2107068419456482,
      "num_tokens": 56201619.0,
      "step": 25180
    },
    {
      "entropy": 5.422843074798584,
      "epoch": 2.8285505700005613,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042127382336266467,
      "loss": 5.1184,
      "mean_token_accuracy": 0.2005790427327156,
      "num_tokens": 56212339.0,
      "step": 25185
    },
    {
      "entropy": 5.48595814704895,
      "epoch": 2.829112146908519,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004212433142857024,
      "loss": 5.1631,
      "mean_token_accuracy": 0.20588430762290955,
      "num_tokens": 56224734.0,
      "step": 25190
    },
    {
      "entropy": 5.470426607131958,
      "epoch": 2.8296737238164766,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042121280055229624,
      "loss": 5.1272,
      "mean_token_accuracy": 0.19953801333904267,
      "num_tokens": 56235594.0,
      "step": 25195
    },
    {
      "entropy": 5.384652137756348,
      "epoch": 2.8302353007244343,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004211822821634177,
      "loss": 5.0996,
      "mean_token_accuracy": 0.2068149372935295,
      "num_tokens": 56246932.0,
      "step": 25200
    },
    {
      "entropy": 5.361141967773437,
      "epoch": 2.830796877632392,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004211517591200387,
      "loss": 5.2142,
      "mean_token_accuracy": 0.2064760610461235,
      "num_tokens": 56258217.0,
      "step": 25205
    },
    {
      "entropy": 5.519026756286621,
      "epoch": 2.831358454540349,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004211212314231309,
      "loss": 5.1991,
      "mean_token_accuracy": 0.2025975003838539,
      "num_tokens": 56269226.0,
      "step": 25210
    },
    {
      "entropy": 5.473792028427124,
      "epoch": 2.831920031448307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042109069907366646,
      "loss": 5.1494,
      "mean_token_accuracy": 0.20363848358392717,
      "num_tokens": 56280411.0,
      "step": 25215
    },
    {
      "entropy": 5.436033010482788,
      "epoch": 2.8324816083562645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042106016207261754,
      "loss": 5.2681,
      "mean_token_accuracy": 0.19762004166841507,
      "num_tokens": 56292048.0,
      "step": 25220
    },
    {
      "entropy": 5.469280862808228,
      "epoch": 2.8330431852642217,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042102962042095643,
      "loss": 5.153,
      "mean_token_accuracy": 0.20751273185014724,
      "num_tokens": 56302470.0,
      "step": 25225
    },
    {
      "entropy": 5.442574882507325,
      "epoch": 2.8336047621721794,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042099907411965553,
      "loss": 5.1083,
      "mean_token_accuracy": 0.20420445054769515,
      "num_tokens": 56313871.0,
      "step": 25230
    },
    {
      "entropy": 5.322543144226074,
      "epoch": 2.834166339080137,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042096852316968756,
      "loss": 5.1009,
      "mean_token_accuracy": 0.20744802802801132,
      "num_tokens": 56325455.0,
      "step": 25235
    },
    {
      "entropy": 5.499881172180176,
      "epoch": 2.8347279159880947,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042093796757202513,
      "loss": 5.2495,
      "mean_token_accuracy": 0.20123881995677947,
      "num_tokens": 56336187.0,
      "step": 25240
    },
    {
      "entropy": 5.589601612091064,
      "epoch": 2.8352894928960524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004209074073276412,
      "loss": 5.2604,
      "mean_token_accuracy": 0.19995932877063752,
      "num_tokens": 56348175.0,
      "step": 25245
    },
    {
      "entropy": 5.465473556518555,
      "epoch": 2.8358510698040096,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004208768424375089,
      "loss": 5.1653,
      "mean_token_accuracy": 0.19941622763872147,
      "num_tokens": 56359668.0,
      "step": 25250
    },
    {
      "entropy": 5.425819015502929,
      "epoch": 2.8364126467119672,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042084627290260124,
      "loss": 5.148,
      "mean_token_accuracy": 0.20999141037464142,
      "num_tokens": 56371160.0,
      "step": 25255
    },
    {
      "entropy": 5.393160200119018,
      "epoch": 2.836974223619925,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004208156987238918,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20854933708906173,
      "num_tokens": 56382873.0,
      "step": 25260
    },
    {
      "entropy": 5.435856771469116,
      "epoch": 2.837535800527882,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004207851199023539,
      "loss": 5.1322,
      "mean_token_accuracy": 0.20391468554735184,
      "num_tokens": 56393562.0,
      "step": 25265
    },
    {
      "entropy": 5.489428091049194,
      "epoch": 2.83809737743584,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042075453643896125,
      "loss": 5.2649,
      "mean_token_accuracy": 0.19971169531345367,
      "num_tokens": 56405789.0,
      "step": 25270
    },
    {
      "entropy": 5.445730113983155,
      "epoch": 2.8386589543437974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042072394833468765,
      "loss": 5.1121,
      "mean_token_accuracy": 0.20937375277280806,
      "num_tokens": 56417621.0,
      "step": 25275
    },
    {
      "entropy": 5.495589256286621,
      "epoch": 2.8392205312517547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000420693355590507,
      "loss": 5.216,
      "mean_token_accuracy": 0.19677433967590333,
      "num_tokens": 56430167.0,
      "step": 25280
    },
    {
      "entropy": 5.486621999740601,
      "epoch": 2.8397821081597123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042066275820739333,
      "loss": 5.2214,
      "mean_token_accuracy": 0.19346835017204284,
      "num_tokens": 56441853.0,
      "step": 25285
    },
    {
      "entropy": 5.370355319976807,
      "epoch": 2.84034368506767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000420632156186321,
      "loss": 5.0917,
      "mean_token_accuracy": 0.20639448314905168,
      "num_tokens": 56454504.0,
      "step": 25290
    },
    {
      "entropy": 5.427287530899048,
      "epoch": 2.8409052619756276,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004206015495282644,
      "loss": 5.17,
      "mean_token_accuracy": 0.20759275704622268,
      "num_tokens": 56465503.0,
      "step": 25295
    },
    {
      "entropy": 5.455255460739136,
      "epoch": 2.8414668388835853,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042057093823419795,
      "loss": 5.2122,
      "mean_token_accuracy": 0.20222514122724533,
      "num_tokens": 56475310.0,
      "step": 25300
    },
    {
      "entropy": 5.464810132980347,
      "epoch": 2.8420284157915425,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042054032230509636,
      "loss": 5.1241,
      "mean_token_accuracy": 0.20564891397953033,
      "num_tokens": 56486575.0,
      "step": 25305
    },
    {
      "entropy": 5.493469905853272,
      "epoch": 2.8425899926995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042050970174193437,
      "loss": 5.2386,
      "mean_token_accuracy": 0.2009178340435028,
      "num_tokens": 56496542.0,
      "step": 25310
    },
    {
      "entropy": 5.530316495895386,
      "epoch": 2.843151569607458,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042047907654568724,
      "loss": 5.1863,
      "mean_token_accuracy": 0.20280291885137558,
      "num_tokens": 56507124.0,
      "step": 25315
    },
    {
      "entropy": 5.43949294090271,
      "epoch": 2.843713146515415,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004204484467173297,
      "loss": 5.0997,
      "mean_token_accuracy": 0.20389626026153565,
      "num_tokens": 56518635.0,
      "step": 25320
    },
    {
      "entropy": 5.374734210968017,
      "epoch": 2.8442747234233727,
      "grad_norm": 1.125,
      "learning_rate": 0.0004204178122578373,
      "loss": 5.1213,
      "mean_token_accuracy": 0.21008632928133011,
      "num_tokens": 56529499.0,
      "step": 25325
    },
    {
      "entropy": 5.472604751586914,
      "epoch": 2.8448363003313304,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042038717316818534,
      "loss": 5.1972,
      "mean_token_accuracy": 0.20285685062408448,
      "num_tokens": 56542432.0,
      "step": 25330
    },
    {
      "entropy": 5.445117568969726,
      "epoch": 2.845397877239288,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004203565294493494,
      "loss": 5.1778,
      "mean_token_accuracy": 0.20217007100582124,
      "num_tokens": 56552783.0,
      "step": 25335
    },
    {
      "entropy": 5.421955728530884,
      "epoch": 2.8459594541472457,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042032588110230525,
      "loss": 5.1306,
      "mean_token_accuracy": 0.20524280071258544,
      "num_tokens": 56563654.0,
      "step": 25340
    },
    {
      "entropy": 5.480859231948853,
      "epoch": 2.846521031055203,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004202952281280286,
      "loss": 5.1876,
      "mean_token_accuracy": 0.2014092981815338,
      "num_tokens": 56575541.0,
      "step": 25345
    },
    {
      "entropy": 5.548755025863647,
      "epoch": 2.8470826079631606,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004202645705274956,
      "loss": 5.2943,
      "mean_token_accuracy": 0.19491764158010483,
      "num_tokens": 56588397.0,
      "step": 25350
    },
    {
      "entropy": 5.4689349174499515,
      "epoch": 2.8476441848711183,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042023390830168224,
      "loss": 5.1671,
      "mean_token_accuracy": 0.2003230944275856,
      "num_tokens": 56599858.0,
      "step": 25355
    },
    {
      "entropy": 5.501519966125488,
      "epoch": 2.8482057617790755,
      "grad_norm": 1.046875,
      "learning_rate": 0.000420203241451565,
      "loss": 5.1366,
      "mean_token_accuracy": 0.21355710923671722,
      "num_tokens": 56610972.0,
      "step": 25360
    },
    {
      "entropy": 5.398740339279175,
      "epoch": 2.848767338687033,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042017256997812016,
      "loss": 5.0728,
      "mean_token_accuracy": 0.2101953685283661,
      "num_tokens": 56621619.0,
      "step": 25365
    },
    {
      "entropy": 5.3734790802001955,
      "epoch": 2.849328915594991,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042014189388232446,
      "loss": 5.1693,
      "mean_token_accuracy": 0.20445075184106826,
      "num_tokens": 56632955.0,
      "step": 25370
    },
    {
      "entropy": 5.49137396812439,
      "epoch": 2.849890492502948,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004201112131651545,
      "loss": 5.2366,
      "mean_token_accuracy": 0.20115977227687837,
      "num_tokens": 56645896.0,
      "step": 25375
    },
    {
      "entropy": 5.3593559741973875,
      "epoch": 2.8504520694109057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004200805278275873,
      "loss": 5.1046,
      "mean_token_accuracy": 0.2025277405977249,
      "num_tokens": 56656974.0,
      "step": 25380
    },
    {
      "entropy": 5.539211988449097,
      "epoch": 2.8510136463188633,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004200498378705997,
      "loss": 5.3335,
      "mean_token_accuracy": 0.19924675077199935,
      "num_tokens": 56669251.0,
      "step": 25385
    },
    {
      "entropy": 5.512341785430908,
      "epoch": 2.851575223226821,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004200191432951692,
      "loss": 5.0779,
      "mean_token_accuracy": 0.21129606962203978,
      "num_tokens": 56679526.0,
      "step": 25390
    },
    {
      "entropy": 5.382119226455688,
      "epoch": 2.8521368001347787,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004199884441022728,
      "loss": 5.1089,
      "mean_token_accuracy": 0.20841437131166457,
      "num_tokens": 56690276.0,
      "step": 25395
    },
    {
      "entropy": 5.3592352867126465,
      "epoch": 2.852698377042736,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041995774029288807,
      "loss": 5.052,
      "mean_token_accuracy": 0.21437742412090302,
      "num_tokens": 56702509.0,
      "step": 25400
    },
    {
      "entropy": 5.458844184875488,
      "epoch": 2.8532599539506935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041992703186799273,
      "loss": 5.1956,
      "mean_token_accuracy": 0.20444972813129425,
      "num_tokens": 56713562.0,
      "step": 25405
    },
    {
      "entropy": 5.400401973724366,
      "epoch": 2.853821530858651,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041989631882856447,
      "loss": 5.1325,
      "mean_token_accuracy": 0.20724885612726213,
      "num_tokens": 56724793.0,
      "step": 25410
    },
    {
      "entropy": 5.463048362731934,
      "epoch": 2.8543831077666084,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004198656011755812,
      "loss": 5.2352,
      "mean_token_accuracy": 0.1953496217727661,
      "num_tokens": 56736842.0,
      "step": 25415
    },
    {
      "entropy": 5.485711240768433,
      "epoch": 2.854944684674566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004198348789100211,
      "loss": 5.1274,
      "mean_token_accuracy": 0.19544288665056228,
      "num_tokens": 56748220.0,
      "step": 25420
    },
    {
      "entropy": 5.333998441696167,
      "epoch": 2.8555062615825237,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041980415203286223,
      "loss": 5.0426,
      "mean_token_accuracy": 0.2101687327027321,
      "num_tokens": 56758420.0,
      "step": 25425
    },
    {
      "entropy": 5.3409339427948,
      "epoch": 2.8560678384904814,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000419773420545083,
      "loss": 5.0682,
      "mean_token_accuracy": 0.21174260824918748,
      "num_tokens": 56768801.0,
      "step": 25430
    },
    {
      "entropy": 5.532502126693726,
      "epoch": 2.856629415398439,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041974268444766187,
      "loss": 5.2483,
      "mean_token_accuracy": 0.19456196278333665,
      "num_tokens": 56780756.0,
      "step": 25435
    },
    {
      "entropy": 5.415250062942505,
      "epoch": 2.8571909923063963,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004197119437415776,
      "loss": 5.0723,
      "mean_token_accuracy": 0.21114940345287322,
      "num_tokens": 56790679.0,
      "step": 25440
    },
    {
      "entropy": 5.429743528366089,
      "epoch": 2.857752569214354,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004196811984278089,
      "loss": 5.1603,
      "mean_token_accuracy": 0.20232051461935044,
      "num_tokens": 56802513.0,
      "step": 25445
    },
    {
      "entropy": 5.398722743988037,
      "epoch": 2.8583141461223116,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004196504485073347,
      "loss": 5.1572,
      "mean_token_accuracy": 0.2079310104250908,
      "num_tokens": 56813213.0,
      "step": 25450
    },
    {
      "entropy": 5.461503267288208,
      "epoch": 2.858875723030269,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041961969398113404,
      "loss": 5.1697,
      "mean_token_accuracy": 0.20354634523391724,
      "num_tokens": 56825097.0,
      "step": 25455
    },
    {
      "entropy": 5.454945135116577,
      "epoch": 2.8594372999382265,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041958893485018634,
      "loss": 5.2639,
      "mean_token_accuracy": 0.20158192962408067,
      "num_tokens": 56837185.0,
      "step": 25460
    },
    {
      "entropy": 5.4053099155426025,
      "epoch": 2.859998876846184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004195581711154709,
      "loss": 5.1807,
      "mean_token_accuracy": 0.19830115437507628,
      "num_tokens": 56848696.0,
      "step": 25465
    },
    {
      "entropy": 5.45564341545105,
      "epoch": 2.8605604537541414,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041952740277796717,
      "loss": 5.2142,
      "mean_token_accuracy": 0.1975526213645935,
      "num_tokens": 56860955.0,
      "step": 25470
    },
    {
      "entropy": 5.499543714523315,
      "epoch": 2.861122030662099,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041949662983865486,
      "loss": 5.2282,
      "mean_token_accuracy": 0.20718390494585037,
      "num_tokens": 56872375.0,
      "step": 25475
    },
    {
      "entropy": 5.468471097946167,
      "epoch": 2.8616836075700567,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004194658522985139,
      "loss": 5.2295,
      "mean_token_accuracy": 0.19644685685634614,
      "num_tokens": 56883644.0,
      "step": 25480
    },
    {
      "entropy": 5.4562958717346195,
      "epoch": 2.8622451844780143,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004194350701585241,
      "loss": 5.1938,
      "mean_token_accuracy": 0.19859348684549333,
      "num_tokens": 56894803.0,
      "step": 25485
    },
    {
      "entropy": 5.505408716201782,
      "epoch": 2.862806761385972,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004194042834196656,
      "loss": 5.1715,
      "mean_token_accuracy": 0.19894446283578873,
      "num_tokens": 56906475.0,
      "step": 25490
    },
    {
      "entropy": 5.503805828094483,
      "epoch": 2.8633683382939292,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004193734920829188,
      "loss": 5.264,
      "mean_token_accuracy": 0.20091774463653564,
      "num_tokens": 56918086.0,
      "step": 25495
    },
    {
      "entropy": 5.4378049850463865,
      "epoch": 2.863929915201887,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000419342696149264,
      "loss": 5.1402,
      "mean_token_accuracy": 0.2121339425444603,
      "num_tokens": 56929300.0,
      "step": 25500
    },
    {
      "entropy": 5.472341299057007,
      "epoch": 2.8644914921098446,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004193118956196818,
      "loss": 5.1931,
      "mean_token_accuracy": 0.20615252256393432,
      "num_tokens": 56941858.0,
      "step": 25505
    },
    {
      "entropy": 5.4952513694763185,
      "epoch": 2.8650530690178018,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004192810904951528,
      "loss": 5.2702,
      "mean_token_accuracy": 0.196147920191288,
      "num_tokens": 56953214.0,
      "step": 25510
    },
    {
      "entropy": 5.446628761291504,
      "epoch": 2.8656146459257594,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000419250280776658,
      "loss": 5.116,
      "mean_token_accuracy": 0.20575880259275436,
      "num_tokens": 56963437.0,
      "step": 25515
    },
    {
      "entropy": 5.442995405197143,
      "epoch": 2.866176222833717,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004192194664651783,
      "loss": 5.1585,
      "mean_token_accuracy": 0.20627513378858567,
      "num_tokens": 56975771.0,
      "step": 25520
    },
    {
      "entropy": 5.347312355041504,
      "epoch": 2.8667377997416748,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004191886475616947,
      "loss": 5.0379,
      "mean_token_accuracy": 0.2114684373140335,
      "num_tokens": 56986084.0,
      "step": 25525
    },
    {
      "entropy": 5.2878861904144285,
      "epoch": 2.8672993766496324,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041915782406718875,
      "loss": 4.9579,
      "mean_token_accuracy": 0.2166465699672699,
      "num_tokens": 56996616.0,
      "step": 25530
    },
    {
      "entropy": 5.3388489246368405,
      "epoch": 2.8678609535575896,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004191269959826418,
      "loss": 5.087,
      "mean_token_accuracy": 0.20653516948223113,
      "num_tokens": 57008431.0,
      "step": 25535
    },
    {
      "entropy": 5.435050344467163,
      "epoch": 2.8684225304655473,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041909616330903544,
      "loss": 5.2024,
      "mean_token_accuracy": 0.20243060439825059,
      "num_tokens": 57019265.0,
      "step": 25540
    },
    {
      "entropy": 5.4536457538604735,
      "epoch": 2.868984107373505,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041906532604735127,
      "loss": 5.1099,
      "mean_token_accuracy": 0.20648247748613358,
      "num_tokens": 57029869.0,
      "step": 25545
    },
    {
      "entropy": 5.406654262542725,
      "epoch": 2.869545684281462,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004190344841985713,
      "loss": 5.1151,
      "mean_token_accuracy": 0.2072426974773407,
      "num_tokens": 57041168.0,
      "step": 25550
    },
    {
      "entropy": 5.361726856231689,
      "epoch": 2.87010726118942,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041900363776367747,
      "loss": 5.0187,
      "mean_token_accuracy": 0.21476262509822847,
      "num_tokens": 57052190.0,
      "step": 25555
    },
    {
      "entropy": 5.46991605758667,
      "epoch": 2.8706688380973775,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004189727867436519,
      "loss": 5.2222,
      "mean_token_accuracy": 0.19721689373254775,
      "num_tokens": 57063817.0,
      "step": 25560
    },
    {
      "entropy": 5.540862417221069,
      "epoch": 2.8712304150053347,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041894193113947703,
      "loss": 5.2281,
      "mean_token_accuracy": 0.1990749180316925,
      "num_tokens": 57074998.0,
      "step": 25565
    },
    {
      "entropy": 5.352570867538452,
      "epoch": 2.8717919919132924,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041891107095213523,
      "loss": 5.0648,
      "mean_token_accuracy": 0.21047315001487732,
      "num_tokens": 57085377.0,
      "step": 25570
    },
    {
      "entropy": 5.369658994674682,
      "epoch": 2.87235356882125,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041888020618260913,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20440869629383088,
      "num_tokens": 57096698.0,
      "step": 25575
    },
    {
      "entropy": 5.3737434387207035,
      "epoch": 2.8729151457292077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004188493368318814,
      "loss": 5.1007,
      "mean_token_accuracy": 0.22189324349164963,
      "num_tokens": 57107559.0,
      "step": 25580
    },
    {
      "entropy": 5.4496818542480465,
      "epoch": 2.8734767226371654,
      "grad_norm": 1.09375,
      "learning_rate": 0.000418818462900935,
      "loss": 5.1665,
      "mean_token_accuracy": 0.2067941188812256,
      "num_tokens": 57118252.0,
      "step": 25585
    },
    {
      "entropy": 5.524903583526611,
      "epoch": 2.8740382995451226,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004187875843907531,
      "loss": 5.2313,
      "mean_token_accuracy": 0.19565823078155517,
      "num_tokens": 57128635.0,
      "step": 25590
    },
    {
      "entropy": 5.435476493835449,
      "epoch": 2.8745998764530802,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041875670130231857,
      "loss": 5.1397,
      "mean_token_accuracy": 0.19768817275762557,
      "num_tokens": 57138139.0,
      "step": 25595
    },
    {
      "entropy": 5.4453143119812015,
      "epoch": 2.875161453361038,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000418725813636615,
      "loss": 5.1959,
      "mean_token_accuracy": 0.20585262775421143,
      "num_tokens": 57149456.0,
      "step": 25600
    },
    {
      "entropy": 5.5206645965576175,
      "epoch": 2.875723030268995,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041869492139462585,
      "loss": 5.2292,
      "mean_token_accuracy": 0.20438657253980635,
      "num_tokens": 57160710.0,
      "step": 25605
    },
    {
      "entropy": 5.4862769603729244,
      "epoch": 2.876284607176953,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041866402457733454,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20441498160362243,
      "num_tokens": 57172006.0,
      "step": 25610
    },
    {
      "entropy": 5.411442232131958,
      "epoch": 2.8768461840849104,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041863312318572504,
      "loss": 5.137,
      "mean_token_accuracy": 0.20361126214265823,
      "num_tokens": 57183665.0,
      "step": 25615
    },
    {
      "entropy": 5.395540523529053,
      "epoch": 2.877407760992868,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004186022172207812,
      "loss": 5.1999,
      "mean_token_accuracy": 0.19711469411849974,
      "num_tokens": 57195711.0,
      "step": 25620
    },
    {
      "entropy": 5.479584455490112,
      "epoch": 2.8779693379008258,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041857130668348705,
      "loss": 5.1028,
      "mean_token_accuracy": 0.2091955617070198,
      "num_tokens": 57206664.0,
      "step": 25625
    },
    {
      "entropy": 5.403205871582031,
      "epoch": 2.878530914808783,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004185403915748268,
      "loss": 5.1877,
      "mean_token_accuracy": 0.19911823719739913,
      "num_tokens": 57217407.0,
      "step": 25630
    },
    {
      "entropy": 5.380570125579834,
      "epoch": 2.8790924917167406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004185094718957848,
      "loss": 5.111,
      "mean_token_accuracy": 0.20767218470573426,
      "num_tokens": 57228707.0,
      "step": 25635
    },
    {
      "entropy": 5.37411060333252,
      "epoch": 2.8796540686246983,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041847854764734556,
      "loss": 5.1145,
      "mean_token_accuracy": 0.20558747202157973,
      "num_tokens": 57240802.0,
      "step": 25640
    },
    {
      "entropy": 5.474435091018677,
      "epoch": 2.8802156455326555,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004184476188304937,
      "loss": 5.1899,
      "mean_token_accuracy": 0.1983725607395172,
      "num_tokens": 57251999.0,
      "step": 25645
    },
    {
      "entropy": 5.614864635467529,
      "epoch": 2.880777222440613,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041841668544621396,
      "loss": 5.3692,
      "mean_token_accuracy": 0.19171447157859803,
      "num_tokens": 57264199.0,
      "step": 25650
    },
    {
      "entropy": 5.414680242538452,
      "epoch": 2.881338799348571,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004183857474954915,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20323451161384581,
      "num_tokens": 57275234.0,
      "step": 25655
    },
    {
      "entropy": 5.461675691604614,
      "epoch": 2.881900376256528,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041835480497931114,
      "loss": 5.2163,
      "mean_token_accuracy": 0.19762876331806184,
      "num_tokens": 57286920.0,
      "step": 25660
    },
    {
      "entropy": 5.458273267745971,
      "epoch": 2.8824619531644857,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004183238578986582,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20384480506181718,
      "num_tokens": 57298148.0,
      "step": 25665
    },
    {
      "entropy": 5.397859811782837,
      "epoch": 2.8830235300724434,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041829290625451806,
      "loss": 5.1529,
      "mean_token_accuracy": 0.2048327222466469,
      "num_tokens": 57308717.0,
      "step": 25670
    },
    {
      "entropy": 5.42693247795105,
      "epoch": 2.883585106980401,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041826195004787616,
      "loss": 5.1249,
      "mean_token_accuracy": 0.1999645084142685,
      "num_tokens": 57319308.0,
      "step": 25675
    },
    {
      "entropy": 5.554058980941773,
      "epoch": 2.8841466838883587,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041823098927971827,
      "loss": 5.2762,
      "mean_token_accuracy": 0.19124944657087325,
      "num_tokens": 57330911.0,
      "step": 25680
    },
    {
      "entropy": 5.470970678329468,
      "epoch": 2.884708260796316,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004182000239510301,
      "loss": 5.1898,
      "mean_token_accuracy": 0.19733578860759735,
      "num_tokens": 57341624.0,
      "step": 25685
    },
    {
      "entropy": 5.432051515579223,
      "epoch": 2.8852698377042736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041816905406279756,
      "loss": 5.1384,
      "mean_token_accuracy": 0.21039615869522094,
      "num_tokens": 57353399.0,
      "step": 25690
    },
    {
      "entropy": 5.500779008865356,
      "epoch": 2.8858314146122312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041813807961600685,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1968797877430916,
      "num_tokens": 57364481.0,
      "step": 25695
    },
    {
      "entropy": 5.395797681808472,
      "epoch": 2.8863929915201885,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004181071006116442,
      "loss": 5.0625,
      "mean_token_accuracy": 0.21052238941192628,
      "num_tokens": 57375588.0,
      "step": 25700
    },
    {
      "entropy": 5.311108922958374,
      "epoch": 2.886954568428146,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041807611705069593,
      "loss": 5.1149,
      "mean_token_accuracy": 0.2019706577062607,
      "num_tokens": 57387048.0,
      "step": 25705
    },
    {
      "entropy": 5.4022002696990965,
      "epoch": 2.887516145336104,
      "grad_norm": 1.0,
      "learning_rate": 0.00041804512893414865,
      "loss": 5.0648,
      "mean_token_accuracy": 0.2137673020362854,
      "num_tokens": 57397946.0,
      "step": 25710
    },
    {
      "entropy": 5.435164976119995,
      "epoch": 2.8880777222440615,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041801413626298893,
      "loss": 5.2098,
      "mean_token_accuracy": 0.19829149842262267,
      "num_tokens": 57409505.0,
      "step": 25715
    },
    {
      "entropy": 5.367911720275879,
      "epoch": 2.888639299152019,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004179831390382037,
      "loss": 5.1481,
      "mean_token_accuracy": 0.20473384261131286,
      "num_tokens": 57420745.0,
      "step": 25720
    },
    {
      "entropy": 5.449914932250977,
      "epoch": 2.8892008760599763,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004179521372607798,
      "loss": 5.225,
      "mean_token_accuracy": 0.20241276770830155,
      "num_tokens": 57431557.0,
      "step": 25725
    },
    {
      "entropy": 5.398954296112061,
      "epoch": 2.889762452967934,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041792113093170447,
      "loss": 5.0775,
      "mean_token_accuracy": 0.20207086354494094,
      "num_tokens": 57442029.0,
      "step": 25730
    },
    {
      "entropy": 5.512173891067505,
      "epoch": 2.8903240298758917,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004178901200519649,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19664837419986725,
      "num_tokens": 57452921.0,
      "step": 25735
    },
    {
      "entropy": 5.375182485580444,
      "epoch": 2.890885606783849,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004178591046225485,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20179080814123154,
      "num_tokens": 57463068.0,
      "step": 25740
    },
    {
      "entropy": 5.482132387161255,
      "epoch": 2.8914471836918065,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041782808464444274,
      "loss": 5.1833,
      "mean_token_accuracy": 0.2064593493938446,
      "num_tokens": 57474064.0,
      "step": 25745
    },
    {
      "entropy": 5.37639970779419,
      "epoch": 2.892008760599764,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041779706011863546,
      "loss": 5.1387,
      "mean_token_accuracy": 0.2044037938117981,
      "num_tokens": 57485473.0,
      "step": 25750
    },
    {
      "entropy": 5.345814752578735,
      "epoch": 2.8925703375077214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041776603104611427,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21522606313228607,
      "num_tokens": 57495562.0,
      "step": 25755
    },
    {
      "entropy": 5.4127197265625,
      "epoch": 2.893131914415679,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041773499742786736,
      "loss": 5.1452,
      "mean_token_accuracy": 0.20925923734903334,
      "num_tokens": 57506322.0,
      "step": 25760
    },
    {
      "entropy": 5.437736654281617,
      "epoch": 2.8936934913236367,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041770395926488276,
      "loss": 5.1589,
      "mean_token_accuracy": 0.2075704365968704,
      "num_tokens": 57517353.0,
      "step": 25765
    },
    {
      "entropy": 5.481371927261352,
      "epoch": 2.8942550682315944,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004176729165581488,
      "loss": 5.1887,
      "mean_token_accuracy": 0.20795637965202332,
      "num_tokens": 57527273.0,
      "step": 25770
    },
    {
      "entropy": 5.3552697658538815,
      "epoch": 2.894816645139552,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041764186930865377,
      "loss": 5.0783,
      "mean_token_accuracy": 0.2094851016998291,
      "num_tokens": 57538486.0,
      "step": 25775
    },
    {
      "entropy": 5.431337165832519,
      "epoch": 2.8953782220475093,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004176108175173863,
      "loss": 5.1365,
      "mean_token_accuracy": 0.20532081574201583,
      "num_tokens": 57549682.0,
      "step": 25780
    },
    {
      "entropy": 5.461119890213013,
      "epoch": 2.895939798955467,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004175797611853351,
      "loss": 5.1165,
      "mean_token_accuracy": 0.203577484190464,
      "num_tokens": 57559397.0,
      "step": 25785
    },
    {
      "entropy": 5.4301341533660885,
      "epoch": 2.8965013758634246,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000417548700313489,
      "loss": 5.1716,
      "mean_token_accuracy": 0.20445359498262405,
      "num_tokens": 57570568.0,
      "step": 25790
    },
    {
      "entropy": 5.402664470672607,
      "epoch": 2.897062952771382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000417517634902837,
      "loss": 5.0589,
      "mean_token_accuracy": 0.21207818388938904,
      "num_tokens": 57581152.0,
      "step": 25795
    },
    {
      "entropy": 5.455234336853027,
      "epoch": 2.8976245296793395,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041748656495436834,
      "loss": 5.2218,
      "mean_token_accuracy": 0.2009818211197853,
      "num_tokens": 57592227.0,
      "step": 25800
    },
    {
      "entropy": 5.428123474121094,
      "epoch": 2.898186106587297,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041745549046907205,
      "loss": 5.1591,
      "mean_token_accuracy": 0.19946266561746598,
      "num_tokens": 57604037.0,
      "step": 25805
    },
    {
      "entropy": 5.530929517745972,
      "epoch": 2.898747683495255,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041742441144793774,
      "loss": 5.1842,
      "mean_token_accuracy": 0.209214286506176,
      "num_tokens": 57613937.0,
      "step": 25810
    },
    {
      "entropy": 5.426222515106201,
      "epoch": 2.8993092604032125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004173933278919549,
      "loss": 5.0755,
      "mean_token_accuracy": 0.20047360211610793,
      "num_tokens": 57624287.0,
      "step": 25815
    },
    {
      "entropy": 5.425559425354004,
      "epoch": 2.8998708373111697,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004173622398021133,
      "loss": 5.2075,
      "mean_token_accuracy": 0.19660639762878418,
      "num_tokens": 57636677.0,
      "step": 25820
    },
    {
      "entropy": 5.30368185043335,
      "epoch": 2.9004324142191273,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041733114717940273,
      "loss": 5.0677,
      "mean_token_accuracy": 0.21388312578201293,
      "num_tokens": 57647848.0,
      "step": 25825
    },
    {
      "entropy": 5.321092510223389,
      "epoch": 2.900993991127085,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004173000500248133,
      "loss": 4.9894,
      "mean_token_accuracy": 0.21329567581415176,
      "num_tokens": 57658946.0,
      "step": 25830
    },
    {
      "entropy": 5.370088529586792,
      "epoch": 2.901555568035042,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004172689483393351,
      "loss": 5.1798,
      "mean_token_accuracy": 0.20348377525806427,
      "num_tokens": 57670690.0,
      "step": 25835
    },
    {
      "entropy": 5.616034412384034,
      "epoch": 2.902117144943,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041723784212395834,
      "loss": 5.2648,
      "mean_token_accuracy": 0.1973352164030075,
      "num_tokens": 57682309.0,
      "step": 25840
    },
    {
      "entropy": 5.48683443069458,
      "epoch": 2.9026787218509575,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041720673137967356,
      "loss": 5.1953,
      "mean_token_accuracy": 0.20093556940555574,
      "num_tokens": 57692804.0,
      "step": 25845
    },
    {
      "entropy": 5.520909309387207,
      "epoch": 2.9032402987589148,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041717561610747124,
      "loss": 5.2552,
      "mean_token_accuracy": 0.19937456846237184,
      "num_tokens": 57705824.0,
      "step": 25850
    },
    {
      "entropy": 5.471019554138183,
      "epoch": 2.9038018756668724,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041714449630834223,
      "loss": 5.1476,
      "mean_token_accuracy": 0.20346733629703523,
      "num_tokens": 57718983.0,
      "step": 25855
    },
    {
      "entropy": 5.4206592559814455,
      "epoch": 2.90436345257483,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004171133719832774,
      "loss": 5.111,
      "mean_token_accuracy": 0.2079310417175293,
      "num_tokens": 57730727.0,
      "step": 25860
    },
    {
      "entropy": 5.416716051101685,
      "epoch": 2.9049250294827877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041708224313326747,
      "loss": 5.2258,
      "mean_token_accuracy": 0.19479139447212218,
      "num_tokens": 57742474.0,
      "step": 25865
    },
    {
      "entropy": 5.526018285751343,
      "epoch": 2.9054866063907454,
      "grad_norm": 0.96875,
      "learning_rate": 0.000417051109759304,
      "loss": 5.1942,
      "mean_token_accuracy": 0.20101891756057738,
      "num_tokens": 57755094.0,
      "step": 25870
    },
    {
      "entropy": 5.490170288085937,
      "epoch": 2.9060481832987026,
      "grad_norm": 1.109375,
      "learning_rate": 0.000417019971862378,
      "loss": 5.2067,
      "mean_token_accuracy": 0.19423471987247468,
      "num_tokens": 57765820.0,
      "step": 25875
    },
    {
      "entropy": 5.453639698028565,
      "epoch": 2.9066097602066603,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041698882944348105,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1975415378808975,
      "num_tokens": 57776749.0,
      "step": 25880
    },
    {
      "entropy": 5.3526660919189455,
      "epoch": 2.907171337114618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041695768250360466,
      "loss": 5.0502,
      "mean_token_accuracy": 0.20913441926240922,
      "num_tokens": 57787734.0,
      "step": 25885
    },
    {
      "entropy": 5.408602905273438,
      "epoch": 2.907732914022575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004169265310437407,
      "loss": 5.1781,
      "mean_token_accuracy": 0.20336141139268876,
      "num_tokens": 57799799.0,
      "step": 25890
    },
    {
      "entropy": 5.324546241760254,
      "epoch": 2.908294490930533,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004168953750648809,
      "loss": 5.0406,
      "mean_token_accuracy": 0.20447193533182145,
      "num_tokens": 57810446.0,
      "step": 25895
    },
    {
      "entropy": 5.439510488510132,
      "epoch": 2.9088560678384905,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041686421456801724,
      "loss": 5.2753,
      "mean_token_accuracy": 0.20082124024629594,
      "num_tokens": 57822931.0,
      "step": 25900
    },
    {
      "entropy": 5.472529315948487,
      "epoch": 2.909417644746448,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004168330495541421,
      "loss": 5.2135,
      "mean_token_accuracy": 0.19524312913417816,
      "num_tokens": 57835250.0,
      "step": 25905
    },
    {
      "entropy": 5.4400389194488525,
      "epoch": 2.909979221654406,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041680188002424755,
      "loss": 5.2249,
      "mean_token_accuracy": 0.20385906249284744,
      "num_tokens": 57847726.0,
      "step": 25910
    },
    {
      "entropy": 5.462716865539551,
      "epoch": 2.910540798562363,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004167707059793262,
      "loss": 5.1794,
      "mean_token_accuracy": 0.20232041776180268,
      "num_tokens": 57858484.0,
      "step": 25915
    },
    {
      "entropy": 5.377426671981811,
      "epoch": 2.9111023754703207,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041673952742037057,
      "loss": 5.1023,
      "mean_token_accuracy": 0.21010929644107817,
      "num_tokens": 57870469.0,
      "step": 25920
    },
    {
      "entropy": 5.41539740562439,
      "epoch": 2.9116639523782784,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004167083443483734,
      "loss": 5.1841,
      "mean_token_accuracy": 0.20243118405342103,
      "num_tokens": 57880088.0,
      "step": 25925
    },
    {
      "entropy": 5.440696668624878,
      "epoch": 2.9122255292862356,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004166771567643276,
      "loss": 5.1553,
      "mean_token_accuracy": 0.19611647129058837,
      "num_tokens": 57890521.0,
      "step": 25930
    },
    {
      "entropy": 5.53588285446167,
      "epoch": 2.9127871061941932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004166459646692262,
      "loss": 5.2975,
      "mean_token_accuracy": 0.19729715436697007,
      "num_tokens": 57904111.0,
      "step": 25935
    },
    {
      "entropy": 5.5199450016021725,
      "epoch": 2.913348683102151,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004166147680640623,
      "loss": 5.2049,
      "mean_token_accuracy": 0.20056531429290772,
      "num_tokens": 57915854.0,
      "step": 25940
    },
    {
      "entropy": 5.381930351257324,
      "epoch": 2.913910260010108,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041658356694982936,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20999300330877305,
      "num_tokens": 57926822.0,
      "step": 25945
    },
    {
      "entropy": 5.408369731903076,
      "epoch": 2.9144718369180658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004165523613275207,
      "loss": 5.1875,
      "mean_token_accuracy": 0.20637202113866807,
      "num_tokens": 57939029.0,
      "step": 25950
    },
    {
      "entropy": 5.393961715698242,
      "epoch": 2.9150334138260234,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004165211511981299,
      "loss": 5.1139,
      "mean_token_accuracy": 0.20553070604801177,
      "num_tokens": 57950053.0,
      "step": 25955
    },
    {
      "entropy": 5.408486270904541,
      "epoch": 2.915594990733981,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004164899365626507,
      "loss": 5.051,
      "mean_token_accuracy": 0.21321532428264617,
      "num_tokens": 57960925.0,
      "step": 25960
    },
    {
      "entropy": 5.350430059432983,
      "epoch": 2.9161565676419388,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004164587174220771,
      "loss": 5.0915,
      "mean_token_accuracy": 0.2021890789270401,
      "num_tokens": 57971948.0,
      "step": 25965
    },
    {
      "entropy": 5.39500846862793,
      "epoch": 2.916718144549896,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041642749377740316,
      "loss": 5.1929,
      "mean_token_accuracy": 0.1953263282775879,
      "num_tokens": 57984310.0,
      "step": 25970
    },
    {
      "entropy": 5.423575401306152,
      "epoch": 2.9172797214578536,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041639626562962293,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21097016632556914,
      "num_tokens": 57995194.0,
      "step": 25975
    },
    {
      "entropy": 5.461473560333252,
      "epoch": 2.9178412983658113,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041636503297973074,
      "loss": 5.2019,
      "mean_token_accuracy": 0.19940125793218613,
      "num_tokens": 58006370.0,
      "step": 25980
    },
    {
      "entropy": 5.435816669464112,
      "epoch": 2.9184028752737685,
      "grad_norm": 1.03125,
      "learning_rate": 0.000416333795828721,
      "loss": 5.148,
      "mean_token_accuracy": 0.20874685198068618,
      "num_tokens": 58017635.0,
      "step": 25985
    },
    {
      "entropy": 5.434745979309082,
      "epoch": 2.918964452181726,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004163025541775885,
      "loss": 5.1396,
      "mean_token_accuracy": 0.2090197592973709,
      "num_tokens": 58028110.0,
      "step": 25990
    },
    {
      "entropy": 5.433812522888184,
      "epoch": 2.919526029089684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004162713080273278,
      "loss": 5.2306,
      "mean_token_accuracy": 0.19314567595720292,
      "num_tokens": 58039715.0,
      "step": 25995
    },
    {
      "entropy": 5.508810758590698,
      "epoch": 2.9200876059976415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041624005737893387,
      "loss": 5.1691,
      "mean_token_accuracy": 0.20778415501117706,
      "num_tokens": 58051052.0,
      "step": 26000
    },
    {
      "entropy": 5.4035248279571535,
      "epoch": 2.920649182905599,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004162088022334018,
      "loss": 5.1755,
      "mean_token_accuracy": 0.20375409126281738,
      "num_tokens": 58062789.0,
      "step": 26005
    },
    {
      "entropy": 5.357011795043945,
      "epoch": 2.9212107598135564,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041617754259172665,
      "loss": 5.0842,
      "mean_token_accuracy": 0.2069661021232605,
      "num_tokens": 58073896.0,
      "step": 26010
    },
    {
      "entropy": 5.541758298873901,
      "epoch": 2.921772336721514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041614627845490377,
      "loss": 5.3534,
      "mean_token_accuracy": 0.19384508579969406,
      "num_tokens": 58084529.0,
      "step": 26015
    },
    {
      "entropy": 5.4695415019989015,
      "epoch": 2.9223339136294717,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041611500982392867,
      "loss": 5.2037,
      "mean_token_accuracy": 0.20239120423793794,
      "num_tokens": 58095989.0,
      "step": 26020
    },
    {
      "entropy": 5.3880956172943115,
      "epoch": 2.922895490537429,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041608373669979696,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21667951643466948,
      "num_tokens": 58107001.0,
      "step": 26025
    },
    {
      "entropy": 5.394635486602783,
      "epoch": 2.9234570674453866,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004160524590835042,
      "loss": 5.116,
      "mean_token_accuracy": 0.20562432557344437,
      "num_tokens": 58118224.0,
      "step": 26030
    },
    {
      "entropy": 5.410785293579101,
      "epoch": 2.9240186443533442,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041602117697604664,
      "loss": 5.0757,
      "mean_token_accuracy": 0.21118640452623366,
      "num_tokens": 58129675.0,
      "step": 26035
    },
    {
      "entropy": 5.442865467071533,
      "epoch": 2.9245802212613015,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041598989037842,
      "loss": 5.1454,
      "mean_token_accuracy": 0.20115061104297638,
      "num_tokens": 58140409.0,
      "step": 26040
    },
    {
      "entropy": 5.408319568634033,
      "epoch": 2.925141798169259,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004159585992916206,
      "loss": 5.0988,
      "mean_token_accuracy": 0.21264615952968596,
      "num_tokens": 58151139.0,
      "step": 26045
    },
    {
      "entropy": 5.492136859893799,
      "epoch": 2.925703375077217,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004159273037166447,
      "loss": 5.2448,
      "mean_token_accuracy": 0.19388846009969712,
      "num_tokens": 58162953.0,
      "step": 26050
    },
    {
      "entropy": 5.43334469795227,
      "epoch": 2.9262649519851744,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041589600365448884,
      "loss": 5.0889,
      "mean_token_accuracy": 0.2060546264052391,
      "num_tokens": 58173480.0,
      "step": 26055
    },
    {
      "entropy": 5.479301881790161,
      "epoch": 2.926826528893132,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004158646991061495,
      "loss": 5.225,
      "mean_token_accuracy": 0.19901593029499054,
      "num_tokens": 58185392.0,
      "step": 26060
    },
    {
      "entropy": 5.419336223602295,
      "epoch": 2.9273881058010893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004158333900726235,
      "loss": 5.0468,
      "mean_token_accuracy": 0.2121843382716179,
      "num_tokens": 58195984.0,
      "step": 26065
    },
    {
      "entropy": 5.44150071144104,
      "epoch": 2.927949682709047,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041580207655490776,
      "loss": 5.202,
      "mean_token_accuracy": 0.20273322016000747,
      "num_tokens": 58207529.0,
      "step": 26070
    },
    {
      "entropy": 5.421343040466309,
      "epoch": 2.9285112596170046,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004157707585539994,
      "loss": 5.1461,
      "mean_token_accuracy": 0.2040032297372818,
      "num_tokens": 58217936.0,
      "step": 26075
    },
    {
      "entropy": 5.411030864715576,
      "epoch": 2.929072836524962,
      "grad_norm": 1.125,
      "learning_rate": 0.00041573943607089536,
      "loss": 5.1213,
      "mean_token_accuracy": 0.2078385517001152,
      "num_tokens": 58228144.0,
      "step": 26080
    },
    {
      "entropy": 5.369383192062378,
      "epoch": 2.9296344134329195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004157081091065932,
      "loss": 5.2021,
      "mean_token_accuracy": 0.20769353359937667,
      "num_tokens": 58239130.0,
      "step": 26085
    },
    {
      "entropy": 5.379394578933716,
      "epoch": 2.930195990340877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004156767776620902,
      "loss": 5.1495,
      "mean_token_accuracy": 0.20469617694616318,
      "num_tokens": 58253119.0,
      "step": 26090
    },
    {
      "entropy": 5.493002796173096,
      "epoch": 2.930757567248835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000415645441738384,
      "loss": 5.1711,
      "mean_token_accuracy": 0.2003135547041893,
      "num_tokens": 58265549.0,
      "step": 26095
    },
    {
      "entropy": 5.4172422885894775,
      "epoch": 2.9313191441567925,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041561410133647244,
      "loss": 5.0698,
      "mean_token_accuracy": 0.2199552074074745,
      "num_tokens": 58275021.0,
      "step": 26100
    },
    {
      "entropy": 5.38364610671997,
      "epoch": 2.9318807210647497,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041558275645735336,
      "loss": 5.159,
      "mean_token_accuracy": 0.20382583290338516,
      "num_tokens": 58286191.0,
      "step": 26105
    },
    {
      "entropy": 5.322482013702393,
      "epoch": 2.9324422979727074,
      "grad_norm": 1.125,
      "learning_rate": 0.00041555140710202484,
      "loss": 5.032,
      "mean_token_accuracy": 0.21036772131919862,
      "num_tokens": 58297066.0,
      "step": 26110
    },
    {
      "entropy": 5.446857500076294,
      "epoch": 2.933003874880665,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041552005327148494,
      "loss": 5.1088,
      "mean_token_accuracy": 0.2117781326174736,
      "num_tokens": 58308008.0,
      "step": 26115
    },
    {
      "entropy": 5.42616868019104,
      "epoch": 2.9335654517886223,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004154886949667321,
      "loss": 5.2326,
      "mean_token_accuracy": 0.20192758589982987,
      "num_tokens": 58319584.0,
      "step": 26120
    },
    {
      "entropy": 5.454682397842407,
      "epoch": 2.93412702869658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041545733218876466,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20412180125713347,
      "num_tokens": 58328964.0,
      "step": 26125
    },
    {
      "entropy": 5.4964745998382565,
      "epoch": 2.9346886056045376,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004154259649385813,
      "loss": 5.1662,
      "mean_token_accuracy": 0.20751436352729796,
      "num_tokens": 58339673.0,
      "step": 26130
    },
    {
      "entropy": 5.46304759979248,
      "epoch": 2.935250182512495,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041539459321718076,
      "loss": 5.2175,
      "mean_token_accuracy": 0.20141832530498505,
      "num_tokens": 58351453.0,
      "step": 26135
    },
    {
      "entropy": 5.472506523132324,
      "epoch": 2.9358117594204525,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004153632170255619,
      "loss": 5.2308,
      "mean_token_accuracy": 0.19904376566410065,
      "num_tokens": 58362792.0,
      "step": 26140
    },
    {
      "entropy": 5.442188549041748,
      "epoch": 2.93637333632841,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041533183636472385,
      "loss": 5.1938,
      "mean_token_accuracy": 0.2084762841463089,
      "num_tokens": 58372628.0,
      "step": 26145
    },
    {
      "entropy": 5.351198005676269,
      "epoch": 2.936934913236368,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004153004512356656,
      "loss": 5.0557,
      "mean_token_accuracy": 0.20982459783554078,
      "num_tokens": 58383723.0,
      "step": 26150
    },
    {
      "entropy": 5.3353804588317875,
      "epoch": 2.9374964901443255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004152690616393866,
      "loss": 5.1208,
      "mean_token_accuracy": 0.2098701700568199,
      "num_tokens": 58394547.0,
      "step": 26155
    },
    {
      "entropy": 5.424770212173462,
      "epoch": 2.9380580670522827,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041523766757688634,
      "loss": 5.1298,
      "mean_token_accuracy": 0.2060968443751335,
      "num_tokens": 58405460.0,
      "step": 26160
    },
    {
      "entropy": 5.465103721618652,
      "epoch": 2.9386196439602403,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004152062690491643,
      "loss": 5.1623,
      "mean_token_accuracy": 0.20847226530313492,
      "num_tokens": 58416436.0,
      "step": 26165
    },
    {
      "entropy": 5.430757808685303,
      "epoch": 2.939181220868198,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004151748660572203,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20920496433973312,
      "num_tokens": 58426920.0,
      "step": 26170
    },
    {
      "entropy": 5.375385808944702,
      "epoch": 2.939742797776155,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004151434586020541,
      "loss": 4.9965,
      "mean_token_accuracy": 0.2141570433974266,
      "num_tokens": 58437525.0,
      "step": 26175
    },
    {
      "entropy": 5.44410457611084,
      "epoch": 2.940304374684113,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041511204668466593,
      "loss": 5.2395,
      "mean_token_accuracy": 0.20362237989902496,
      "num_tokens": 58448955.0,
      "step": 26180
    },
    {
      "entropy": 5.319963788986206,
      "epoch": 2.9408659515920705,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041508063030605585,
      "loss": 5.0262,
      "mean_token_accuracy": 0.21264876127243043,
      "num_tokens": 58459941.0,
      "step": 26185
    },
    {
      "entropy": 5.42093186378479,
      "epoch": 2.941427528500028,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041504920946722413,
      "loss": 5.0902,
      "mean_token_accuracy": 0.2079539641737938,
      "num_tokens": 58470499.0,
      "step": 26190
    },
    {
      "entropy": 5.412980556488037,
      "epoch": 2.941989105407986,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004150177841691713,
      "loss": 5.1397,
      "mean_token_accuracy": 0.2061356484889984,
      "num_tokens": 58482969.0,
      "step": 26195
    },
    {
      "entropy": 5.517292070388794,
      "epoch": 2.942550682315943,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004149863544128979,
      "loss": 5.1511,
      "mean_token_accuracy": 0.20663608610630035,
      "num_tokens": 58492929.0,
      "step": 26200
    },
    {
      "entropy": 5.406018924713135,
      "epoch": 2.9431122592239007,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041495492019940467,
      "loss": 5.1045,
      "mean_token_accuracy": 0.2094804584980011,
      "num_tokens": 58503932.0,
      "step": 26205
    },
    {
      "entropy": 5.310173034667969,
      "epoch": 2.9436738361318584,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004149234815296925,
      "loss": 5.0215,
      "mean_token_accuracy": 0.21466903686523436,
      "num_tokens": 58515282.0,
      "step": 26210
    },
    {
      "entropy": 5.451882934570312,
      "epoch": 2.9442354130398156,
      "grad_norm": 1.25,
      "learning_rate": 0.00041489203840476243,
      "loss": 5.2589,
      "mean_token_accuracy": 0.19951150715351104,
      "num_tokens": 58527218.0,
      "step": 26215
    },
    {
      "entropy": 5.4922027587890625,
      "epoch": 2.9447969899477733,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041486059082561556,
      "loss": 5.1048,
      "mean_token_accuracy": 0.20603868812322618,
      "num_tokens": 58538622.0,
      "step": 26220
    },
    {
      "entropy": 5.434462642669677,
      "epoch": 2.945358566855731,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004148291387932533,
      "loss": 5.1454,
      "mean_token_accuracy": 0.2069733738899231,
      "num_tokens": 58549747.0,
      "step": 26225
    },
    {
      "entropy": 5.383919382095337,
      "epoch": 2.9459201437636886,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041479768230867696,
      "loss": 5.1748,
      "mean_token_accuracy": 0.20045768320560456,
      "num_tokens": 58560771.0,
      "step": 26230
    },
    {
      "entropy": 5.417341041564941,
      "epoch": 2.946481720671646,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041476622137288823,
      "loss": 5.0512,
      "mean_token_accuracy": 0.20584377497434617,
      "num_tokens": 58571184.0,
      "step": 26235
    },
    {
      "entropy": 5.497204446792603,
      "epoch": 2.9470432975796035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041473475598688886,
      "loss": 5.1922,
      "mean_token_accuracy": 0.2010476619005203,
      "num_tokens": 58583075.0,
      "step": 26240
    },
    {
      "entropy": 5.472446489334106,
      "epoch": 2.947604874487561,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004147032861516805,
      "loss": 5.1181,
      "mean_token_accuracy": 0.20338730216026307,
      "num_tokens": 58593978.0,
      "step": 26245
    },
    {
      "entropy": 5.3990837097167965,
      "epoch": 2.948166451395519,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041467181186826546,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20300608724355698,
      "num_tokens": 58604525.0,
      "step": 26250
    },
    {
      "entropy": 5.332387256622314,
      "epoch": 2.948728028303476,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041464033313764575,
      "loss": 5.111,
      "mean_token_accuracy": 0.2076411470770836,
      "num_tokens": 58614965.0,
      "step": 26255
    },
    {
      "entropy": 5.417438220977783,
      "epoch": 2.9492896052114337,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041460884996082363,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20631330013275145,
      "num_tokens": 58626254.0,
      "step": 26260
    },
    {
      "entropy": 5.445356607437134,
      "epoch": 2.9498511821193913,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041457736233880166,
      "loss": 5.1855,
      "mean_token_accuracy": 0.20529866516590117,
      "num_tokens": 58637474.0,
      "step": 26265
    },
    {
      "entropy": 5.471877861022949,
      "epoch": 2.9504127590273486,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004145458702725823,
      "loss": 5.165,
      "mean_token_accuracy": 0.20356644839048385,
      "num_tokens": 58649560.0,
      "step": 26270
    },
    {
      "entropy": 5.394941425323486,
      "epoch": 2.9509743359353062,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041451437376316825,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20584533363580704,
      "num_tokens": 58659944.0,
      "step": 26275
    },
    {
      "entropy": 5.5028317928314205,
      "epoch": 2.951535912843264,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004144828728115624,
      "loss": 5.1385,
      "mean_token_accuracy": 0.20217933356761933,
      "num_tokens": 58670636.0,
      "step": 26280
    },
    {
      "entropy": 5.461650705337524,
      "epoch": 2.9520974897512215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004144513674187679,
      "loss": 5.1811,
      "mean_token_accuracy": 0.20394921451807022,
      "num_tokens": 58682832.0,
      "step": 26285
    },
    {
      "entropy": 5.385468769073486,
      "epoch": 2.952659066659179,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004144198575857878,
      "loss": 5.1594,
      "mean_token_accuracy": 0.20358292013406754,
      "num_tokens": 58694675.0,
      "step": 26290
    },
    {
      "entropy": 5.34437141418457,
      "epoch": 2.9532206435671364,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004143883433136252,
      "loss": 5.1295,
      "mean_token_accuracy": 0.20645448714494705,
      "num_tokens": 58705613.0,
      "step": 26295
    },
    {
      "entropy": 5.51003851890564,
      "epoch": 2.953782220475094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004143568246032838,
      "loss": 5.1937,
      "mean_token_accuracy": 0.20135696232318878,
      "num_tokens": 58716951.0,
      "step": 26300
    },
    {
      "entropy": 5.435044670104981,
      "epoch": 2.9543437973830518,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004143253014557671,
      "loss": 5.2202,
      "mean_token_accuracy": 0.19886878728866578,
      "num_tokens": 58727730.0,
      "step": 26305
    },
    {
      "entropy": 5.368337345123291,
      "epoch": 2.954905374291009,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041429377387207865,
      "loss": 5.0964,
      "mean_token_accuracy": 0.20425185114145278,
      "num_tokens": 58739183.0,
      "step": 26310
    },
    {
      "entropy": 5.535157585144043,
      "epoch": 2.9554669511989666,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041426224185322255,
      "loss": 5.1986,
      "mean_token_accuracy": 0.20431118905544282,
      "num_tokens": 58750031.0,
      "step": 26315
    },
    {
      "entropy": 5.491509246826172,
      "epoch": 2.9560285281069243,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004142307054002025,
      "loss": 5.1745,
      "mean_token_accuracy": 0.20514926165342331,
      "num_tokens": 58760488.0,
      "step": 26320
    },
    {
      "entropy": 5.486062049865723,
      "epoch": 2.956590105014882,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004141991645140228,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1931058406829834,
      "num_tokens": 58771484.0,
      "step": 26325
    },
    {
      "entropy": 5.382993507385254,
      "epoch": 2.957151681922839,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041416761919568784,
      "loss": 5.0907,
      "mean_token_accuracy": 0.20423207879066468,
      "num_tokens": 58782208.0,
      "step": 26330
    },
    {
      "entropy": 5.452815246582031,
      "epoch": 2.957713258830797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041413606944620183,
      "loss": 5.1613,
      "mean_token_accuracy": 0.20244159549474716,
      "num_tokens": 58792556.0,
      "step": 26335
    },
    {
      "entropy": 5.387005949020386,
      "epoch": 2.9582748357387545,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004141045152665695,
      "loss": 5.1026,
      "mean_token_accuracy": 0.21215709447860717,
      "num_tokens": 58803531.0,
      "step": 26340
    },
    {
      "entropy": 5.476835680007935,
      "epoch": 2.958836412646712,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041407295665779533,
      "loss": 5.149,
      "mean_token_accuracy": 0.20485242307186127,
      "num_tokens": 58813639.0,
      "step": 26345
    },
    {
      "entropy": 5.469878005981445,
      "epoch": 2.9593979895546694,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041404139362088434,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20671990215778352,
      "num_tokens": 58824830.0,
      "step": 26350
    },
    {
      "entropy": 5.418271684646607,
      "epoch": 2.959959566462627,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004140098261568414,
      "loss": 5.1659,
      "mean_token_accuracy": 0.20456338673830032,
      "num_tokens": 58835664.0,
      "step": 26355
    },
    {
      "entropy": 5.43869833946228,
      "epoch": 2.9605211433705847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041397825426667175,
      "loss": 5.1353,
      "mean_token_accuracy": 0.19986433833837508,
      "num_tokens": 58845872.0,
      "step": 26360
    },
    {
      "entropy": 5.339887237548828,
      "epoch": 2.961082720278542,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004139466779513805,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21431145071983337,
      "num_tokens": 58857090.0,
      "step": 26365
    },
    {
      "entropy": 5.445224571228027,
      "epoch": 2.9616442971864996,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004139150972119732,
      "loss": 5.1529,
      "mean_token_accuracy": 0.20400341004133224,
      "num_tokens": 58868814.0,
      "step": 26370
    },
    {
      "entropy": 5.3369225025177,
      "epoch": 2.9622058740944572,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041388351204945534,
      "loss": 5.0742,
      "mean_token_accuracy": 0.20559915155172348,
      "num_tokens": 58879270.0,
      "step": 26375
    },
    {
      "entropy": 5.490208530426026,
      "epoch": 2.962767451002415,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041385192246483257,
      "loss": 5.1964,
      "mean_token_accuracy": 0.20113086551427842,
      "num_tokens": 58891861.0,
      "step": 26380
    },
    {
      "entropy": 5.446350622177124,
      "epoch": 2.9633290279103726,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041382032845911073,
      "loss": 5.1607,
      "mean_token_accuracy": 0.2029400497674942,
      "num_tokens": 58903779.0,
      "step": 26385
    },
    {
      "entropy": 5.517849922180176,
      "epoch": 2.9638906048183298,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041378873003329575,
      "loss": 5.2115,
      "mean_token_accuracy": 0.20550186783075333,
      "num_tokens": 58915773.0,
      "step": 26390
    },
    {
      "entropy": 5.40208888053894,
      "epoch": 2.9644521817262874,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004137571271883938,
      "loss": 5.1084,
      "mean_token_accuracy": 0.20732554495334626,
      "num_tokens": 58926701.0,
      "step": 26395
    },
    {
      "entropy": 5.335906648635865,
      "epoch": 2.965013758634245,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004137255199254112,
      "loss": 5.0467,
      "mean_token_accuracy": 0.2102498471736908,
      "num_tokens": 58937606.0,
      "step": 26400
    },
    {
      "entropy": 5.440891361236572,
      "epoch": 2.9655753355422023,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041369390824535417,
      "loss": 5.1936,
      "mean_token_accuracy": 0.20925533026456833,
      "num_tokens": 58948982.0,
      "step": 26405
    },
    {
      "entropy": 5.4642637252807615,
      "epoch": 2.96613691245016,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041366229214922926,
      "loss": 5.1154,
      "mean_token_accuracy": 0.21148009449243546,
      "num_tokens": 58961607.0,
      "step": 26410
    },
    {
      "entropy": 5.35354585647583,
      "epoch": 2.9666984893581176,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004136306716380432,
      "loss": 5.0923,
      "mean_token_accuracy": 0.20871664881706237,
      "num_tokens": 58973072.0,
      "step": 26415
    },
    {
      "entropy": 5.366648721694946,
      "epoch": 2.9672600662660753,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041359904671280285,
      "loss": 5.0914,
      "mean_token_accuracy": 0.2050260439515114,
      "num_tokens": 58983330.0,
      "step": 26420
    },
    {
      "entropy": 5.536860227584839,
      "epoch": 2.9678216431740325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004135674173745151,
      "loss": 5.2932,
      "mean_token_accuracy": 0.19546388387680053,
      "num_tokens": 58995912.0,
      "step": 26425
    },
    {
      "entropy": 5.533572149276734,
      "epoch": 2.96838322008199,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041353578362418695,
      "loss": 5.2463,
      "mean_token_accuracy": 0.20119615495204926,
      "num_tokens": 59005964.0,
      "step": 26430
    },
    {
      "entropy": 5.530998563766479,
      "epoch": 2.968944796989948,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004135041454628258,
      "loss": 5.2656,
      "mean_token_accuracy": 0.1933860018849373,
      "num_tokens": 59017285.0,
      "step": 26435
    },
    {
      "entropy": 5.316849994659424,
      "epoch": 2.9695063738979055,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004134725028914389,
      "loss": 5.0379,
      "mean_token_accuracy": 0.21500612199306487,
      "num_tokens": 59027751.0,
      "step": 26440
    },
    {
      "entropy": 5.412247991561889,
      "epoch": 2.9700679508058627,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041344085591103383,
      "loss": 5.2041,
      "mean_token_accuracy": 0.20224321037530898,
      "num_tokens": 59039568.0,
      "step": 26445
    },
    {
      "entropy": 5.441375303268432,
      "epoch": 2.9706295277138204,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004134092045226182,
      "loss": 5.1332,
      "mean_token_accuracy": 0.20378589183092116,
      "num_tokens": 59051162.0,
      "step": 26450
    },
    {
      "entropy": 5.371971559524536,
      "epoch": 2.971191104621778,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041337754872719984,
      "loss": 5.0214,
      "mean_token_accuracy": 0.21234232932329178,
      "num_tokens": 59061381.0,
      "step": 26455
    },
    {
      "entropy": 5.348218727111816,
      "epoch": 2.9717526815297353,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004133458885257866,
      "loss": 5.0227,
      "mean_token_accuracy": 0.208959698677063,
      "num_tokens": 59071026.0,
      "step": 26460
    },
    {
      "entropy": 5.346969127655029,
      "epoch": 2.972314258437693,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041331422391938663,
      "loss": 5.0858,
      "mean_token_accuracy": 0.20463314950466155,
      "num_tokens": 59082793.0,
      "step": 26465
    },
    {
      "entropy": 5.543031501770019,
      "epoch": 2.9728758353456506,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004132825549090082,
      "loss": 5.3085,
      "mean_token_accuracy": 0.20023994892835617,
      "num_tokens": 59094247.0,
      "step": 26470
    },
    {
      "entropy": 5.355016994476318,
      "epoch": 2.9734374122536082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004132508814956595,
      "loss": 5.0812,
      "mean_token_accuracy": 0.20555794090032578,
      "num_tokens": 59104936.0,
      "step": 26475
    },
    {
      "entropy": 5.44595365524292,
      "epoch": 2.973998989161566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004132192036803491,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20822259336709975,
      "num_tokens": 59116181.0,
      "step": 26480
    },
    {
      "entropy": 5.305372047424316,
      "epoch": 2.974560566069523,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041318752146408575,
      "loss": 5.0505,
      "mean_token_accuracy": 0.20972964614629747,
      "num_tokens": 59126953.0,
      "step": 26485
    },
    {
      "entropy": 5.3992119312286375,
      "epoch": 2.975122142977481,
      "grad_norm": 1.0,
      "learning_rate": 0.0004131558348478781,
      "loss": 5.2425,
      "mean_token_accuracy": 0.20022059679031373,
      "num_tokens": 59138460.0,
      "step": 26490
    },
    {
      "entropy": 5.518276214599609,
      "epoch": 2.9756837198854384,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041312414383273507,
      "loss": 5.1373,
      "mean_token_accuracy": 0.20661238431930543,
      "num_tokens": 59149003.0,
      "step": 26495
    },
    {
      "entropy": 5.462990808486938,
      "epoch": 2.9762452967933957,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041309244841966565,
      "loss": 5.1416,
      "mean_token_accuracy": 0.20664554685354233,
      "num_tokens": 59160582.0,
      "step": 26500
    },
    {
      "entropy": 5.440999698638916,
      "epoch": 2.9768068737013533,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041306074860967916,
      "loss": 5.2259,
      "mean_token_accuracy": 0.19604205191135407,
      "num_tokens": 59172115.0,
      "step": 26505
    },
    {
      "entropy": 5.388048839569092,
      "epoch": 2.977368450609311,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041302904440378494,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20857212990522384,
      "num_tokens": 59183044.0,
      "step": 26510
    },
    {
      "entropy": 5.497510719299316,
      "epoch": 2.9779300275172687,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041299733580299236,
      "loss": 5.2497,
      "mean_token_accuracy": 0.19283446818590164,
      "num_tokens": 59194791.0,
      "step": 26515
    },
    {
      "entropy": 5.354038190841675,
      "epoch": 2.978491604425226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041296562280831106,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20381662100553513,
      "num_tokens": 59206285.0,
      "step": 26520
    },
    {
      "entropy": 5.453002834320069,
      "epoch": 2.9790531813331835,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041293390542075095,
      "loss": 5.2027,
      "mean_token_accuracy": 0.1985051915049553,
      "num_tokens": 59217906.0,
      "step": 26525
    },
    {
      "entropy": 5.442254495620728,
      "epoch": 2.979614758241141,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004129021836413217,
      "loss": 5.1732,
      "mean_token_accuracy": 0.2017549991607666,
      "num_tokens": 59228163.0,
      "step": 26530
    },
    {
      "entropy": 5.3813752174377445,
      "epoch": 2.980176335149099,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004128704574710334,
      "loss": 5.1395,
      "mean_token_accuracy": 0.20656674802303315,
      "num_tokens": 59238633.0,
      "step": 26535
    },
    {
      "entropy": 5.463999032974243,
      "epoch": 2.980737912057056,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004128387269108963,
      "loss": 5.1628,
      "mean_token_accuracy": 0.19960794299840928,
      "num_tokens": 59249523.0,
      "step": 26540
    },
    {
      "entropy": 5.48050217628479,
      "epoch": 2.9812994889650137,
      "grad_norm": 1.0,
      "learning_rate": 0.00041280699196192074,
      "loss": 5.2325,
      "mean_token_accuracy": 0.1989982932806015,
      "num_tokens": 59259636.0,
      "step": 26545
    },
    {
      "entropy": 5.420747709274292,
      "epoch": 2.9818610658729714,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041277525262511705,
      "loss": 5.1549,
      "mean_token_accuracy": 0.20476855784654618,
      "num_tokens": 59270963.0,
      "step": 26550
    },
    {
      "entropy": 5.47378191947937,
      "epoch": 2.9824226427809286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041274350890149584,
      "loss": 5.2243,
      "mean_token_accuracy": 0.2068357452750206,
      "num_tokens": 59282085.0,
      "step": 26555
    },
    {
      "entropy": 5.520145082473755,
      "epoch": 2.9829842196888863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000412711760792068,
      "loss": 5.1793,
      "mean_token_accuracy": 0.2068144276738167,
      "num_tokens": 59292210.0,
      "step": 26560
    },
    {
      "entropy": 5.418028020858765,
      "epoch": 2.983545796596844,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004126800082978442,
      "loss": 5.1875,
      "mean_token_accuracy": 0.20605515092611312,
      "num_tokens": 59303727.0,
      "step": 26565
    },
    {
      "entropy": 5.402276563644409,
      "epoch": 2.9841073735048016,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041264825141983547,
      "loss": 5.0929,
      "mean_token_accuracy": 0.20997599512338638,
      "num_tokens": 59313922.0,
      "step": 26570
    },
    {
      "entropy": 5.392837190628052,
      "epoch": 2.9846689504127593,
      "grad_norm": 1.0,
      "learning_rate": 0.00041261649015905314,
      "loss": 5.113,
      "mean_token_accuracy": 0.20775419175624849,
      "num_tokens": 59324974.0,
      "step": 26575
    },
    {
      "entropy": 5.473950481414795,
      "epoch": 2.9852305273207165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004125847245165084,
      "loss": 5.1666,
      "mean_token_accuracy": 0.19924530237913132,
      "num_tokens": 59335538.0,
      "step": 26580
    },
    {
      "entropy": 5.475880193710327,
      "epoch": 2.985792104228674,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004125529544932126,
      "loss": 5.2309,
      "mean_token_accuracy": 0.2036566212773323,
      "num_tokens": 59347116.0,
      "step": 26585
    },
    {
      "entropy": 5.470913124084473,
      "epoch": 2.986353681136632,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004125211800901774,
      "loss": 5.1472,
      "mean_token_accuracy": 0.20641808062791825,
      "num_tokens": 59358505.0,
      "step": 26590
    },
    {
      "entropy": 5.3709053039550785,
      "epoch": 2.986915258044589,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004124894013084145,
      "loss": 5.074,
      "mean_token_accuracy": 0.21068450957536697,
      "num_tokens": 59369897.0,
      "step": 26595
    },
    {
      "entropy": 5.3752206325531,
      "epoch": 2.9874768349525467,
      "grad_norm": 1.0,
      "learning_rate": 0.00041245761814893576,
      "loss": 5.1521,
      "mean_token_accuracy": 0.20931680500507355,
      "num_tokens": 59381372.0,
      "step": 26600
    },
    {
      "entropy": 5.406676340103149,
      "epoch": 2.9880384118605043,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004124258306127531,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20757926553487777,
      "num_tokens": 59391543.0,
      "step": 26605
    },
    {
      "entropy": 5.504978847503662,
      "epoch": 2.988599988768462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004123940387008787,
      "loss": 5.205,
      "mean_token_accuracy": 0.20452597588300706,
      "num_tokens": 59402474.0,
      "step": 26610
    },
    {
      "entropy": 5.399457311630249,
      "epoch": 2.9891615656764197,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004123622424143248,
      "loss": 5.09,
      "mean_token_accuracy": 0.20751651376485825,
      "num_tokens": 59413970.0,
      "step": 26615
    },
    {
      "entropy": 5.348721599578857,
      "epoch": 2.989723142584377,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041233044175410387,
      "loss": 5.0866,
      "mean_token_accuracy": 0.2083508014678955,
      "num_tokens": 59424958.0,
      "step": 26620
    },
    {
      "entropy": 5.465299272537232,
      "epoch": 2.9902847194923345,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041229863672122844,
      "loss": 5.1563,
      "mean_token_accuracy": 0.2053896278142929,
      "num_tokens": 59436226.0,
      "step": 26625
    },
    {
      "entropy": 5.5227179527282715,
      "epoch": 2.990846296400292,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004122668273167111,
      "loss": 5.2417,
      "mean_token_accuracy": 0.1991884171962738,
      "num_tokens": 59447607.0,
      "step": 26630
    },
    {
      "entropy": 5.340977048873901,
      "epoch": 2.9914078733082494,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004122350135415648,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21271868348121642,
      "num_tokens": 59458120.0,
      "step": 26635
    },
    {
      "entropy": 5.425700378417969,
      "epoch": 2.991969450216207,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004122031953968024,
      "loss": 5.1822,
      "mean_token_accuracy": 0.20534602850675582,
      "num_tokens": 59468646.0,
      "step": 26640
    },
    {
      "entropy": 5.4211506843566895,
      "epoch": 2.9925310271241647,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041217137288343704,
      "loss": 5.1524,
      "mean_token_accuracy": 0.20261449813842775,
      "num_tokens": 59478887.0,
      "step": 26645
    },
    {
      "entropy": 5.445552968978882,
      "epoch": 2.993092604032122,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000412139546002482,
      "loss": 5.1825,
      "mean_token_accuracy": 0.2055753469467163,
      "num_tokens": 59490501.0,
      "step": 26650
    },
    {
      "entropy": 5.453135776519775,
      "epoch": 2.9936541809400796,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041210771475495064,
      "loss": 5.2553,
      "mean_token_accuracy": 0.19832976311445236,
      "num_tokens": 59500873.0,
      "step": 26655
    },
    {
      "entropy": 5.377845764160156,
      "epoch": 2.9942157578480373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004120758791418564,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20698791295289992,
      "num_tokens": 59511793.0,
      "step": 26660
    },
    {
      "entropy": 5.450450420379639,
      "epoch": 2.994777334755995,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004120440391642131,
      "loss": 5.2154,
      "mean_token_accuracy": 0.19895641058683394,
      "num_tokens": 59523978.0,
      "step": 26665
    },
    {
      "entropy": 5.4931426525115965,
      "epoch": 2.9953389116639526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004120121948230343,
      "loss": 5.2236,
      "mean_token_accuracy": 0.20332878530025483,
      "num_tokens": 59534575.0,
      "step": 26670
    },
    {
      "entropy": 5.4301379203796385,
      "epoch": 2.99590048857191,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004119803461193342,
      "loss": 5.1725,
      "mean_token_accuracy": 0.19445961713790894,
      "num_tokens": 59545542.0,
      "step": 26675
    },
    {
      "entropy": 5.407687568664551,
      "epoch": 2.9964620654798675,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004119484930541267,
      "loss": 5.125,
      "mean_token_accuracy": 0.21505854278802872,
      "num_tokens": 59556482.0,
      "step": 26680
    },
    {
      "entropy": 5.454925394058227,
      "epoch": 2.997023642387825,
      "grad_norm": 1.0,
      "learning_rate": 0.00041191663562842605,
      "loss": 5.1403,
      "mean_token_accuracy": 0.20555537194013596,
      "num_tokens": 59568941.0,
      "step": 26685
    },
    {
      "entropy": 5.425735378265381,
      "epoch": 2.9975852192957824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004118847738432467,
      "loss": 5.2084,
      "mean_token_accuracy": 0.19801675975322724,
      "num_tokens": 59579690.0,
      "step": 26690
    },
    {
      "entropy": 5.448099660873413,
      "epoch": 2.99814679620374,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004118529076996029,
      "loss": 5.1752,
      "mean_token_accuracy": 0.20091814994812013,
      "num_tokens": 59589831.0,
      "step": 26695
    },
    {
      "entropy": 5.40655927658081,
      "epoch": 2.9987083731116977,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041182103719850967,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20042402297258377,
      "num_tokens": 59600939.0,
      "step": 26700
    },
    {
      "entropy": 5.327128887176514,
      "epoch": 2.9992699500196554,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041178916234098134,
      "loss": 5.0121,
      "mean_token_accuracy": 0.20884857028722764,
      "num_tokens": 59612084.0,
      "step": 26705
    },
    {
      "entropy": 5.431169271469116,
      "epoch": 2.999831526927613,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004117572831280332,
      "loss": 5.1714,
      "mean_token_accuracy": 0.1992751881480217,
      "num_tokens": 59624359.0,
      "step": 26710
    },
    {
      "entropy": 5.526124954223633,
      "epoch": 3.0003369461447744,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041172539956068,
      "loss": 5.1564,
      "mean_token_accuracy": 0.20234054492579567,
      "num_tokens": 59633030.0,
      "step": 26715
    },
    {
      "entropy": 5.324075937271118,
      "epoch": 3.000898523052732,
      "grad_norm": 1.0,
      "learning_rate": 0.00041169351163993705,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20862768590450287,
      "num_tokens": 59644107.0,
      "step": 26720
    },
    {
      "entropy": 5.323019647598267,
      "epoch": 3.0014600999606897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041166161936681973,
      "loss": 4.9294,
      "mean_token_accuracy": 0.21860850602388382,
      "num_tokens": 59654166.0,
      "step": 26725
    },
    {
      "entropy": 5.3719439029693605,
      "epoch": 3.002021676868647,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004116297227423435,
      "loss": 5.0493,
      "mean_token_accuracy": 0.2064652293920517,
      "num_tokens": 59664785.0,
      "step": 26730
    },
    {
      "entropy": 5.283218574523926,
      "epoch": 3.0025832537766046,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041159782176752375,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22047033607959748,
      "num_tokens": 59675634.0,
      "step": 26735
    },
    {
      "entropy": 5.421010875701905,
      "epoch": 3.0031448306845623,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004115659164433765,
      "loss": 5.1035,
      "mean_token_accuracy": 0.21313134729862213,
      "num_tokens": 59687636.0,
      "step": 26740
    },
    {
      "entropy": 5.452748203277588,
      "epoch": 3.00370640759252,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041153400677091745,
      "loss": 5.0731,
      "mean_token_accuracy": 0.20327839851379395,
      "num_tokens": 59700389.0,
      "step": 26745
    },
    {
      "entropy": 5.352976274490357,
      "epoch": 3.004267984500477,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041150209275116266,
      "loss": 4.9969,
      "mean_token_accuracy": 0.21862302124500274,
      "num_tokens": 59710898.0,
      "step": 26750
    },
    {
      "entropy": 5.407136392593384,
      "epoch": 3.004829561408435,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041147017438512834,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20468493551015854,
      "num_tokens": 59722398.0,
      "step": 26755
    },
    {
      "entropy": 5.366084861755371,
      "epoch": 3.0053911383163925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041143825167383073,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20729355067014693,
      "num_tokens": 59732504.0,
      "step": 26760
    },
    {
      "entropy": 5.421009969711304,
      "epoch": 3.00595271522435,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004114063246182863,
      "loss": 5.0787,
      "mean_token_accuracy": 0.21264821588993071,
      "num_tokens": 59742919.0,
      "step": 26765
    },
    {
      "entropy": 5.397217416763306,
      "epoch": 3.0065142921323074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041137439321951147,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20624478459358214,
      "num_tokens": 59754316.0,
      "step": 26770
    },
    {
      "entropy": 5.434725952148438,
      "epoch": 3.007075869040265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041134245747852314,
      "loss": 5.149,
      "mean_token_accuracy": 0.20038966089487076,
      "num_tokens": 59765031.0,
      "step": 26775
    },
    {
      "entropy": 5.414833688735962,
      "epoch": 3.0076374459482227,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004113105173963382,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20650243759155273,
      "num_tokens": 59776338.0,
      "step": 26780
    },
    {
      "entropy": 5.422388648986816,
      "epoch": 3.0081990228561803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004112785729739733,
      "loss": 5.0609,
      "mean_token_accuracy": 0.21055942475795747,
      "num_tokens": 59788544.0,
      "step": 26785
    },
    {
      "entropy": 5.373544216156006,
      "epoch": 3.0087605997641376,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041124662421244593,
      "loss": 5.0287,
      "mean_token_accuracy": 0.20555375814437865,
      "num_tokens": 59800131.0,
      "step": 26790
    },
    {
      "entropy": 5.4171277523040775,
      "epoch": 3.0093221766720952,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004112146711127732,
      "loss": 5.0819,
      "mean_token_accuracy": 0.21017909348011016,
      "num_tokens": 59811986.0,
      "step": 26795
    },
    {
      "entropy": 5.477831649780273,
      "epoch": 3.009883753580053,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004111827136759725,
      "loss": 5.1226,
      "mean_token_accuracy": 0.20790442824363708,
      "num_tokens": 59824758.0,
      "step": 26800
    },
    {
      "entropy": 5.302941131591797,
      "epoch": 3.0104453304880106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004111507519030613,
      "loss": 5.0368,
      "mean_token_accuracy": 0.20788172632455826,
      "num_tokens": 59836317.0,
      "step": 26805
    },
    {
      "entropy": 5.403493452072143,
      "epoch": 3.0110069073959678,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004111187857950574,
      "loss": 5.1235,
      "mean_token_accuracy": 0.20668591260910035,
      "num_tokens": 59847743.0,
      "step": 26810
    },
    {
      "entropy": 5.439643335342407,
      "epoch": 3.0115684843039254,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004110868153529786,
      "loss": 5.0758,
      "mean_token_accuracy": 0.21108705103397368,
      "num_tokens": 59859442.0,
      "step": 26815
    },
    {
      "entropy": 5.327371120452881,
      "epoch": 3.012130061211883,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004110548405778428,
      "loss": 4.9989,
      "mean_token_accuracy": 0.21206795573234558,
      "num_tokens": 59869949.0,
      "step": 26820
    },
    {
      "entropy": 5.331804513931274,
      "epoch": 3.0126916381198403,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004110228614706681,
      "loss": 5.0324,
      "mean_token_accuracy": 0.21683990508317946,
      "num_tokens": 59881818.0,
      "step": 26825
    },
    {
      "entropy": 5.396168851852417,
      "epoch": 3.013253215027798,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041099087803247276,
      "loss": 4.9921,
      "mean_token_accuracy": 0.20632669180631638,
      "num_tokens": 59893152.0,
      "step": 26830
    },
    {
      "entropy": 5.450803136825561,
      "epoch": 3.0138147919357556,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004109588902642751,
      "loss": 5.1509,
      "mean_token_accuracy": 0.19558384418487548,
      "num_tokens": 59904082.0,
      "step": 26835
    },
    {
      "entropy": 5.380934238433838,
      "epoch": 3.0143763688437133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004109268981670936,
      "loss": 5.0856,
      "mean_token_accuracy": 0.21097348332405091,
      "num_tokens": 59916116.0,
      "step": 26840
    },
    {
      "entropy": 5.271137762069702,
      "epoch": 3.0149379457516705,
      "grad_norm": 1.03125,
      "learning_rate": 0.000410894901741947,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2152872383594513,
      "num_tokens": 59926399.0,
      "step": 26845
    },
    {
      "entropy": 5.357283020019532,
      "epoch": 3.015499522659628,
      "grad_norm": 1.046875,
      "learning_rate": 0.000410862900989854,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21239547431468964,
      "num_tokens": 59936986.0,
      "step": 26850
    },
    {
      "entropy": 5.480962800979614,
      "epoch": 3.016061099567586,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004108308959118336,
      "loss": 5.1264,
      "mean_token_accuracy": 0.1977434828877449,
      "num_tokens": 59948926.0,
      "step": 26855
    },
    {
      "entropy": 5.373425483703613,
      "epoch": 3.0166226764755435,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004107988865089047,
      "loss": 4.9913,
      "mean_token_accuracy": 0.21494581550359726,
      "num_tokens": 59960468.0,
      "step": 26860
    },
    {
      "entropy": 5.400493049621582,
      "epoch": 3.0171842533835007,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041076687278208664,
      "loss": 5.066,
      "mean_token_accuracy": 0.20581958442926407,
      "num_tokens": 59971151.0,
      "step": 26865
    },
    {
      "entropy": 5.367842340469361,
      "epoch": 3.0177458302914584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004107348547323987,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2123366117477417,
      "num_tokens": 59982205.0,
      "step": 26870
    },
    {
      "entropy": 5.36891736984253,
      "epoch": 3.018307407199416,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041070283236086024,
      "loss": 5.0158,
      "mean_token_accuracy": 0.2128363236784935,
      "num_tokens": 59993612.0,
      "step": 26875
    },
    {
      "entropy": 5.477448320388794,
      "epoch": 3.0188689841073737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004106708056684911,
      "loss": 5.1867,
      "mean_token_accuracy": 0.20023497492074965,
      "num_tokens": 60004284.0,
      "step": 26880
    },
    {
      "entropy": 5.4639973640441895,
      "epoch": 3.019430561015331,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041063877465631084,
      "loss": 5.1721,
      "mean_token_accuracy": 0.19982555359601975,
      "num_tokens": 60016276.0,
      "step": 26885
    },
    {
      "entropy": 5.36993260383606,
      "epoch": 3.0199921379232886,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004106067393253394,
      "loss": 5.0942,
      "mean_token_accuracy": 0.2127436712384224,
      "num_tokens": 60029155.0,
      "step": 26890
    },
    {
      "entropy": 5.37736873626709,
      "epoch": 3.0205537148312462,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004105746996765968,
      "loss": 5.0632,
      "mean_token_accuracy": 0.20029204934835435,
      "num_tokens": 60041042.0,
      "step": 26895
    },
    {
      "entropy": 5.483259868621826,
      "epoch": 3.021115291739204,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004105426557111031,
      "loss": 5.1109,
      "mean_token_accuracy": 0.20519065111875534,
      "num_tokens": 60052025.0,
      "step": 26900
    },
    {
      "entropy": 5.388881635665894,
      "epoch": 3.021676868647161,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004105106074298787,
      "loss": 5.053,
      "mean_token_accuracy": 0.21102722287178038,
      "num_tokens": 60063260.0,
      "step": 26905
    },
    {
      "entropy": 5.3546959400177006,
      "epoch": 3.022238445555119,
      "grad_norm": 1.046875,
      "learning_rate": 0.000410478554833944,
      "loss": 5.0987,
      "mean_token_accuracy": 0.20078408122062683,
      "num_tokens": 60074721.0,
      "step": 26910
    },
    {
      "entropy": 5.369453048706054,
      "epoch": 3.0228000224630764,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041044649792431963,
      "loss": 5.0223,
      "mean_token_accuracy": 0.21459574550390242,
      "num_tokens": 60085028.0,
      "step": 26915
    },
    {
      "entropy": 5.345967245101929,
      "epoch": 3.0233615993710337,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041041443670202616,
      "loss": 4.9129,
      "mean_token_accuracy": 0.21686313599348067,
      "num_tokens": 60095587.0,
      "step": 26920
    },
    {
      "entropy": 5.328643560409546,
      "epoch": 3.0239231762789913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004103823711680846,
      "loss": 4.9378,
      "mean_token_accuracy": 0.215905624628067,
      "num_tokens": 60106574.0,
      "step": 26925
    },
    {
      "entropy": 5.300217485427856,
      "epoch": 3.024484753186949,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041035030132351575,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21330203413963317,
      "num_tokens": 60116821.0,
      "step": 26930
    },
    {
      "entropy": 5.407117080688477,
      "epoch": 3.0250463300949066,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041031822716934086,
      "loss": 5.1146,
      "mean_token_accuracy": 0.20793962031602858,
      "num_tokens": 60128229.0,
      "step": 26935
    },
    {
      "entropy": 5.395570611953735,
      "epoch": 3.025607907002864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041028614870658106,
      "loss": 5.0235,
      "mean_token_accuracy": 0.20196650177240372,
      "num_tokens": 60139416.0,
      "step": 26940
    },
    {
      "entropy": 5.373397588729858,
      "epoch": 3.0261694839108215,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004102540659362579,
      "loss": 5.013,
      "mean_token_accuracy": 0.21117731630802156,
      "num_tokens": 60150225.0,
      "step": 26945
    },
    {
      "entropy": 5.337496662139893,
      "epoch": 3.026731060818779,
      "grad_norm": 1.0,
      "learning_rate": 0.00041022197885939285,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20414439141750335,
      "num_tokens": 60160065.0,
      "step": 26950
    },
    {
      "entropy": 5.389242172241211,
      "epoch": 3.027292637726737,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041018988747700745,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2205623283982277,
      "num_tokens": 60170396.0,
      "step": 26955
    },
    {
      "entropy": 5.345180559158325,
      "epoch": 3.027854214634694,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004101577917901236,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20904203057289122,
      "num_tokens": 60181938.0,
      "step": 26960
    },
    {
      "entropy": 5.380049037933349,
      "epoch": 3.0284157915426517,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041012569179976326,
      "loss": 5.0055,
      "mean_token_accuracy": 0.20925343930721282,
      "num_tokens": 60192388.0,
      "step": 26965
    },
    {
      "entropy": 5.447871303558349,
      "epoch": 3.0289773684506094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004100935875069485,
      "loss": 5.0828,
      "mean_token_accuracy": 0.20731501430273055,
      "num_tokens": 60203195.0,
      "step": 26970
    },
    {
      "entropy": 5.365634632110596,
      "epoch": 3.029538945358567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004100614789127015,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20791214555501938,
      "num_tokens": 60213659.0,
      "step": 26975
    },
    {
      "entropy": 5.3228692531585695,
      "epoch": 3.0301005222665243,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004100293660180447,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20935701578855515,
      "num_tokens": 60224795.0,
      "step": 26980
    },
    {
      "entropy": 5.3245758533477785,
      "epoch": 3.030662099174482,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040999724882400046,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21754882037639617,
      "num_tokens": 60236188.0,
      "step": 26985
    },
    {
      "entropy": 5.379591178894043,
      "epoch": 3.0312236760824396,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004099651273315915,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20981260091066362,
      "num_tokens": 60247003.0,
      "step": 26990
    },
    {
      "entropy": 5.356689357757569,
      "epoch": 3.0317852529903973,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004099330015418405,
      "loss": 5.0553,
      "mean_token_accuracy": 0.20446816384792327,
      "num_tokens": 60258317.0,
      "step": 26995
    },
    {
      "entropy": 5.456135177612305,
      "epoch": 3.0323468298983545,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040990087145577044,
      "loss": 5.0644,
      "mean_token_accuracy": 0.213719180226326,
      "num_tokens": 60269000.0,
      "step": 27000
    },
    {
      "epoch": 3.0323468298983545,
      "eval_entropy": 5.22835329423284,
      "eval_loss": 5.234411239624023,
      "eval_mean_token_accuracy": 0.20840384364550896,
      "eval_num_tokens": 60269000.0,
      "eval_runtime": 23.863,
      "eval_samples_per_second": 1299.627,
      "eval_steps_per_second": 162.469,
      "step": 27000
    },
    {
      "entropy": 5.31757116317749,
      "epoch": 3.032908406806312,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004098687370744042,
      "loss": 4.9534,
      "mean_token_accuracy": 0.21355899572372436,
      "num_tokens": 60280411.0,
      "step": 27005
    },
    {
      "entropy": 5.3899437427520756,
      "epoch": 3.03346998371427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004098365983987652,
      "loss": 5.1123,
      "mean_token_accuracy": 0.19954788833856582,
      "num_tokens": 60290197.0,
      "step": 27010
    },
    {
      "entropy": 5.47722110748291,
      "epoch": 3.034031560622227,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040980445542987653,
      "loss": 4.995,
      "mean_token_accuracy": 0.21232680678367616,
      "num_tokens": 60300927.0,
      "step": 27015
    },
    {
      "entropy": 5.35021743774414,
      "epoch": 3.0345931375301847,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040977230816876176,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21355928778648375,
      "num_tokens": 60312697.0,
      "step": 27020
    },
    {
      "entropy": 5.26846342086792,
      "epoch": 3.0351547144381423,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040974015661644433,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2186899110674858,
      "num_tokens": 60323779.0,
      "step": 27025
    },
    {
      "entropy": 5.371620082855225,
      "epoch": 3.0357162913461,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004097080007739481,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20789124369621276,
      "num_tokens": 60335036.0,
      "step": 27030
    },
    {
      "entropy": 5.352286720275879,
      "epoch": 3.036277868254057,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040967584064229695,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20215086340904237,
      "num_tokens": 60345600.0,
      "step": 27035
    },
    {
      "entropy": 5.467558145523071,
      "epoch": 3.036839445162015,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040964367622251473,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20349147468805312,
      "num_tokens": 60357241.0,
      "step": 27040
    },
    {
      "entropy": 5.3810282230377195,
      "epoch": 3.0374010220699725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004096115075156257,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20575074553489686,
      "num_tokens": 60369036.0,
      "step": 27045
    },
    {
      "entropy": 5.4619745254516605,
      "epoch": 3.03796259897793,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004095793345226541,
      "loss": 5.1619,
      "mean_token_accuracy": 0.20736043602228166,
      "num_tokens": 60380310.0,
      "step": 27050
    },
    {
      "entropy": 5.382735967636108,
      "epoch": 3.0385241758858874,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004095471572446242,
      "loss": 4.9534,
      "mean_token_accuracy": 0.21890883445739745,
      "num_tokens": 60390962.0,
      "step": 27055
    },
    {
      "entropy": 5.467870092391967,
      "epoch": 3.039085752793845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004095149756825606,
      "loss": 5.0813,
      "mean_token_accuracy": 0.21038102656602858,
      "num_tokens": 60400342.0,
      "step": 27060
    },
    {
      "entropy": 5.3418288230896,
      "epoch": 3.0396473297018027,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004094827898374881,
      "loss": 5.006,
      "mean_token_accuracy": 0.2114434450864792,
      "num_tokens": 60411921.0,
      "step": 27065
    },
    {
      "entropy": 5.499903392791748,
      "epoch": 3.0402089066097604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004094505997104314,
      "loss": 5.1556,
      "mean_token_accuracy": 0.2031136706471443,
      "num_tokens": 60424185.0,
      "step": 27070
    },
    {
      "entropy": 5.234940910339356,
      "epoch": 3.0407704835177176,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040941840530241537,
      "loss": 4.9505,
      "mean_token_accuracy": 0.212970469892025,
      "num_tokens": 60433729.0,
      "step": 27075
    },
    {
      "entropy": 5.4151482582092285,
      "epoch": 3.0413320604256753,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040938620661446526,
      "loss": 5.1072,
      "mean_token_accuracy": 0.2074386328458786,
      "num_tokens": 60443846.0,
      "step": 27080
    },
    {
      "entropy": 5.437692594528198,
      "epoch": 3.041893637333633,
      "grad_norm": 1.0,
      "learning_rate": 0.0004093540036476062,
      "loss": 5.0961,
      "mean_token_accuracy": 0.20617983788251876,
      "num_tokens": 60455183.0,
      "step": 27085
    },
    {
      "entropy": 5.333357095718384,
      "epoch": 3.0424552142415906,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040932179640286356,
      "loss": 5.0328,
      "mean_token_accuracy": 0.20447436720132828,
      "num_tokens": 60465769.0,
      "step": 27090
    },
    {
      "entropy": 5.367208814620971,
      "epoch": 3.043016791149548,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004092895848812628,
      "loss": 4.9248,
      "mean_token_accuracy": 0.21281246542930604,
      "num_tokens": 60476252.0,
      "step": 27095
    },
    {
      "entropy": 5.354694128036499,
      "epoch": 3.0435783680575055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040925736908382953,
      "loss": 5.0392,
      "mean_token_accuracy": 0.21310027539730073,
      "num_tokens": 60485939.0,
      "step": 27100
    },
    {
      "entropy": 5.310129404067993,
      "epoch": 3.044139944965463,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004092251490115896,
      "loss": 5.0649,
      "mean_token_accuracy": 0.20587736070156099,
      "num_tokens": 60497994.0,
      "step": 27105
    },
    {
      "entropy": 5.578674697875977,
      "epoch": 3.0447015218734204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040919292466556895,
      "loss": 5.2031,
      "mean_token_accuracy": 0.20687660723924636,
      "num_tokens": 60508944.0,
      "step": 27110
    },
    {
      "entropy": 5.471115207672119,
      "epoch": 3.045263098781378,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004091606960467934,
      "loss": 5.1173,
      "mean_token_accuracy": 0.21451826393604279,
      "num_tokens": 60519928.0,
      "step": 27115
    },
    {
      "entropy": 5.378223657608032,
      "epoch": 3.0458246756893357,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040912846315628934,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21164528876543046,
      "num_tokens": 60530417.0,
      "step": 27120
    },
    {
      "entropy": 5.354965496063232,
      "epoch": 3.0463862525972933,
      "grad_norm": 1.0,
      "learning_rate": 0.000409096225995083,
      "loss": 5.1058,
      "mean_token_accuracy": 0.20679617077112197,
      "num_tokens": 60542140.0,
      "step": 27125
    },
    {
      "entropy": 5.358968734741211,
      "epoch": 3.0469478295052506,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040906398456420077,
      "loss": 4.9903,
      "mean_token_accuracy": 0.21610727161169052,
      "num_tokens": 60553032.0,
      "step": 27130
    },
    {
      "entropy": 5.406117153167725,
      "epoch": 3.047509406413208,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040903173886466934,
      "loss": 5.1095,
      "mean_token_accuracy": 0.21434491723775864,
      "num_tokens": 60563773.0,
      "step": 27135
    },
    {
      "entropy": 5.321988534927368,
      "epoch": 3.048070983321166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040899948889751525,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21136685609817504,
      "num_tokens": 60574354.0,
      "step": 27140
    },
    {
      "entropy": 5.416130876541137,
      "epoch": 3.0486325602291235,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040896723466376557,
      "loss": 5.0284,
      "mean_token_accuracy": 0.21526232063770295,
      "num_tokens": 60584760.0,
      "step": 27145
    },
    {
      "entropy": 5.215230417251587,
      "epoch": 3.0491941371370808,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040893497616444714,
      "loss": 4.912,
      "mean_token_accuracy": 0.22041582465171813,
      "num_tokens": 60595322.0,
      "step": 27150
    },
    {
      "entropy": 5.486225605010986,
      "epoch": 3.0497557140450384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004089027134005871,
      "loss": 5.2162,
      "mean_token_accuracy": 0.19851610958576202,
      "num_tokens": 60607897.0,
      "step": 27155
    },
    {
      "entropy": 5.333057832717896,
      "epoch": 3.050317290952996,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004088704463732127,
      "loss": 5.0182,
      "mean_token_accuracy": 0.2097821593284607,
      "num_tokens": 60618882.0,
      "step": 27160
    },
    {
      "entropy": 5.380187511444092,
      "epoch": 3.0508788678609537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004088381750833514,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20405630618333817,
      "num_tokens": 60630431.0,
      "step": 27165
    },
    {
      "entropy": 5.481782484054565,
      "epoch": 3.051440444768911,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004088058995320307,
      "loss": 5.109,
      "mean_token_accuracy": 0.20438822358846664,
      "num_tokens": 60642479.0,
      "step": 27170
    },
    {
      "entropy": 5.358634424209595,
      "epoch": 3.0520020216768686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004087736197202784,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2123665601015091,
      "num_tokens": 60652922.0,
      "step": 27175
    },
    {
      "entropy": 5.363733291625977,
      "epoch": 3.0525635985848263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040874133564912195,
      "loss": 5.0342,
      "mean_token_accuracy": 0.21142424792051315,
      "num_tokens": 60665125.0,
      "step": 27180
    },
    {
      "entropy": 5.527137899398804,
      "epoch": 3.053125175492784,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040870904731958965,
      "loss": 5.1686,
      "mean_token_accuracy": 0.20281098634004593,
      "num_tokens": 60675807.0,
      "step": 27185
    },
    {
      "entropy": 5.3713212490081785,
      "epoch": 3.053686752400741,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004086767547327094,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21260489970445634,
      "num_tokens": 60686841.0,
      "step": 27190
    },
    {
      "entropy": 5.365980863571167,
      "epoch": 3.054248329308699,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040864445788950935,
      "loss": 5.0778,
      "mean_token_accuracy": 0.2069576010107994,
      "num_tokens": 60698965.0,
      "step": 27195
    },
    {
      "entropy": 5.429429292678833,
      "epoch": 3.0548099062166565,
      "grad_norm": 1.015625,
      "learning_rate": 0.000408612156791018,
      "loss": 5.1614,
      "mean_token_accuracy": 0.21075122356414794,
      "num_tokens": 60709962.0,
      "step": 27200
    },
    {
      "entropy": 5.400036954879761,
      "epoch": 3.0553714831246137,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004085798514382637,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21083596497774124,
      "num_tokens": 60721267.0,
      "step": 27205
    },
    {
      "entropy": 5.349150705337524,
      "epoch": 3.0559330600325714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040854754183227516,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21718464642763138,
      "num_tokens": 60731892.0,
      "step": 27210
    },
    {
      "entropy": 5.36124095916748,
      "epoch": 3.056494636940529,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004085152279740811,
      "loss": 5.0676,
      "mean_token_accuracy": 0.20305193364620208,
      "num_tokens": 60742875.0,
      "step": 27215
    },
    {
      "entropy": 5.448908042907715,
      "epoch": 3.0570562138484867,
      "grad_norm": 1.125,
      "learning_rate": 0.00040848290986471037,
      "loss": 5.0523,
      "mean_token_accuracy": 0.20990779548883437,
      "num_tokens": 60752716.0,
      "step": 27220
    },
    {
      "entropy": 5.382467651367188,
      "epoch": 3.057617790756444,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040845058750519205,
      "loss": 5.1564,
      "mean_token_accuracy": 0.2002987876534462,
      "num_tokens": 60765466.0,
      "step": 27225
    },
    {
      "entropy": 5.415820932388305,
      "epoch": 3.0581793676644016,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040841826089655525,
      "loss": 5.0999,
      "mean_token_accuracy": 0.1995697170495987,
      "num_tokens": 60776488.0,
      "step": 27230
    },
    {
      "entropy": 5.460047197341919,
      "epoch": 3.0587409445723592,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040838593003982926,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20870957374572754,
      "num_tokens": 60786510.0,
      "step": 27235
    },
    {
      "entropy": 5.343446826934814,
      "epoch": 3.059302521480317,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004083535949360436,
      "loss": 5.0169,
      "mean_token_accuracy": 0.21967872530221938,
      "num_tokens": 60798300.0,
      "step": 27240
    },
    {
      "entropy": 5.227676010131836,
      "epoch": 3.059864098388274,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004083212555862277,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21240556389093398,
      "num_tokens": 60809949.0,
      "step": 27245
    },
    {
      "entropy": 5.34961724281311,
      "epoch": 3.0604256752962318,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004082889119914113,
      "loss": 5.005,
      "mean_token_accuracy": 0.21123767346143724,
      "num_tokens": 60819960.0,
      "step": 27250
    },
    {
      "entropy": 5.458117437362671,
      "epoch": 3.0609872522041894,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004082565641526242,
      "loss": 5.0493,
      "mean_token_accuracy": 0.21555231809616088,
      "num_tokens": 60830934.0,
      "step": 27255
    },
    {
      "entropy": 5.27258448600769,
      "epoch": 3.061548829112147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040822421207089654,
      "loss": 5.025,
      "mean_token_accuracy": 0.20771117210388185,
      "num_tokens": 60842161.0,
      "step": 27260
    },
    {
      "entropy": 5.4284507751464846,
      "epoch": 3.0621104060201043,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004081918557472582,
      "loss": 5.0978,
      "mean_token_accuracy": 0.2092292621731758,
      "num_tokens": 60854220.0,
      "step": 27265
    },
    {
      "entropy": 5.396103763580323,
      "epoch": 3.062671982928062,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004081594951827396,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2209744930267334,
      "num_tokens": 60864769.0,
      "step": 27270
    },
    {
      "entropy": 5.324406290054322,
      "epoch": 3.0632335598360196,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040812713037837096,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21664815843105317,
      "num_tokens": 60874865.0,
      "step": 27275
    },
    {
      "entropy": 5.392212820053101,
      "epoch": 3.0637951367439773,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040809476133518284,
      "loss": 5.0602,
      "mean_token_accuracy": 0.20880553126335144,
      "num_tokens": 60885099.0,
      "step": 27280
    },
    {
      "entropy": 5.405401563644409,
      "epoch": 3.0643567136519345,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040806238805420595,
      "loss": 5.1312,
      "mean_token_accuracy": 0.2052501082420349,
      "num_tokens": 60896071.0,
      "step": 27285
    },
    {
      "entropy": 5.43503360748291,
      "epoch": 3.064918290559892,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000408030010536471,
      "loss": 5.1515,
      "mean_token_accuracy": 0.2068452075123787,
      "num_tokens": 60907255.0,
      "step": 27290
    },
    {
      "entropy": 5.3690601825714115,
      "epoch": 3.06547986746785,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040799762878300895,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21929231137037278,
      "num_tokens": 60918740.0,
      "step": 27295
    },
    {
      "entropy": 5.323395109176635,
      "epoch": 3.066041444375807,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040796524279485084,
      "loss": 4.953,
      "mean_token_accuracy": 0.2155454784631729,
      "num_tokens": 60929327.0,
      "step": 27300
    },
    {
      "entropy": 5.3457330703735355,
      "epoch": 3.0666030212837647,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00040793285257302776,
      "loss": 5.016,
      "mean_token_accuracy": 0.21312355101108552,
      "num_tokens": 60939987.0,
      "step": 27305
    },
    {
      "entropy": 5.485849761962891,
      "epoch": 3.0671645981917224,
      "grad_norm": 1.0625,
      "learning_rate": 0.000407900458118571,
      "loss": 5.1747,
      "mean_token_accuracy": 0.2073729529976845,
      "num_tokens": 60951924.0,
      "step": 27310
    },
    {
      "entropy": 5.4338137149810795,
      "epoch": 3.06772617509968,
      "grad_norm": 1.0,
      "learning_rate": 0.0004078680594325123,
      "loss": 5.087,
      "mean_token_accuracy": 0.208570396900177,
      "num_tokens": 60963706.0,
      "step": 27315
    },
    {
      "entropy": 5.40734167098999,
      "epoch": 3.0682877520076373,
      "grad_norm": 1.09375,
      "learning_rate": 0.000407835656515883,
      "loss": 5.0759,
      "mean_token_accuracy": 0.2131690964102745,
      "num_tokens": 60974842.0,
      "step": 27320
    },
    {
      "entropy": 5.348116683959961,
      "epoch": 3.068849328915595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040780324936971486,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2179832562804222,
      "num_tokens": 60984921.0,
      "step": 27325
    },
    {
      "entropy": 5.3936262130737305,
      "epoch": 3.0694109058235526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004077708379950397,
      "loss": 5.0727,
      "mean_token_accuracy": 0.2088018536567688,
      "num_tokens": 60995648.0,
      "step": 27330
    },
    {
      "entropy": 5.322894954681397,
      "epoch": 3.0699724827315102,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004077384223928896,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20863904654979706,
      "num_tokens": 61008020.0,
      "step": 27335
    },
    {
      "entropy": 5.3846851825714115,
      "epoch": 3.0705340596394675,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004077060025642967,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21295994371175767,
      "num_tokens": 61017735.0,
      "step": 27340
    },
    {
      "entropy": 5.37728214263916,
      "epoch": 3.071095636547425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004076735785102931,
      "loss": 5.1417,
      "mean_token_accuracy": 0.20318266600370408,
      "num_tokens": 61029111.0,
      "step": 27345
    },
    {
      "entropy": 5.3966254711151125,
      "epoch": 3.071657213455383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004076411502319114,
      "loss": 5.0282,
      "mean_token_accuracy": 0.2106887772679329,
      "num_tokens": 61039392.0,
      "step": 27350
    },
    {
      "entropy": 5.344965171813965,
      "epoch": 3.0722187903633404,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040760871773018405,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21763253062963486,
      "num_tokens": 61049681.0,
      "step": 27355
    },
    {
      "entropy": 5.387428617477417,
      "epoch": 3.0727803672712977,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004075762810061437,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20756817162036895,
      "num_tokens": 61059910.0,
      "step": 27360
    },
    {
      "entropy": 5.436712980270386,
      "epoch": 3.0733419441792553,
      "grad_norm": 1.25,
      "learning_rate": 0.000407543840060823,
      "loss": 5.1231,
      "mean_token_accuracy": 0.21505847722291946,
      "num_tokens": 61071074.0,
      "step": 27365
    },
    {
      "entropy": 5.350608396530151,
      "epoch": 3.073903521087213,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004075113948952551,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2164027914404869,
      "num_tokens": 61082051.0,
      "step": 27370
    },
    {
      "entropy": 5.391592597961425,
      "epoch": 3.0744650979951706,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004074789455104731,
      "loss": 5.0711,
      "mean_token_accuracy": 0.21173403561115264,
      "num_tokens": 61093578.0,
      "step": 27375
    },
    {
      "entropy": 5.3749778270721436,
      "epoch": 3.075026674903128,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040744649190750995,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20764901489019394,
      "num_tokens": 61105192.0,
      "step": 27380
    },
    {
      "entropy": 5.430095386505127,
      "epoch": 3.0755882518110855,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004074140340873992,
      "loss": 5.1099,
      "mean_token_accuracy": 0.20671942383050917,
      "num_tokens": 61116659.0,
      "step": 27385
    },
    {
      "entropy": 5.4345602035522464,
      "epoch": 3.076149828719043,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004073815720511743,
      "loss": 5.0638,
      "mean_token_accuracy": 0.20779457241296767,
      "num_tokens": 61127761.0,
      "step": 27390
    },
    {
      "entropy": 5.4210714340209964,
      "epoch": 3.0767114056270004,
      "grad_norm": 1.0,
      "learning_rate": 0.00040734910579986875,
      "loss": 4.9816,
      "mean_token_accuracy": 0.2167563334107399,
      "num_tokens": 61139039.0,
      "step": 27395
    },
    {
      "entropy": 5.430883836746216,
      "epoch": 3.077272982534958,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040731663533451624,
      "loss": 5.0578,
      "mean_token_accuracy": 0.2053838327527046,
      "num_tokens": 61149877.0,
      "step": 27400
    },
    {
      "entropy": 5.451173400878906,
      "epoch": 3.0778345594429157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040728416065615087,
      "loss": 5.0674,
      "mean_token_accuracy": 0.20616116672754287,
      "num_tokens": 61160806.0,
      "step": 27405
    },
    {
      "entropy": 5.35833511352539,
      "epoch": 3.0783961363508734,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004072516817658065,
      "loss": 5.17,
      "mean_token_accuracy": 0.20267827361822127,
      "num_tokens": 61171471.0,
      "step": 27410
    },
    {
      "entropy": 5.406922292709351,
      "epoch": 3.0789577132588306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040721919866451724,
      "loss": 5.0693,
      "mean_token_accuracy": 0.20906611979007722,
      "num_tokens": 61182140.0,
      "step": 27415
    },
    {
      "entropy": 5.406250095367431,
      "epoch": 3.0795192901667883,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004071867113533174,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21201249212026596,
      "num_tokens": 61192532.0,
      "step": 27420
    },
    {
      "entropy": 5.400935220718384,
      "epoch": 3.080080867074746,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040715421983324147,
      "loss": 5.1283,
      "mean_token_accuracy": 0.2066509634256363,
      "num_tokens": 61205207.0,
      "step": 27425
    },
    {
      "entropy": 5.487467098236084,
      "epoch": 3.0806424439827036,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040712172410532397,
      "loss": 5.164,
      "mean_token_accuracy": 0.20032101422548293,
      "num_tokens": 61218039.0,
      "step": 27430
    },
    {
      "entropy": 5.491283178329468,
      "epoch": 3.081204020890661,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004070892241705994,
      "loss": 5.1171,
      "mean_token_accuracy": 0.2054124265909195,
      "num_tokens": 61230343.0,
      "step": 27435
    },
    {
      "entropy": 5.295750570297241,
      "epoch": 3.0817655977986185,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040705672003010275,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2111295446753502,
      "num_tokens": 61240781.0,
      "step": 27440
    },
    {
      "entropy": 5.332730102539062,
      "epoch": 3.082327174706576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000407024211684869,
      "loss": 5.0426,
      "mean_token_accuracy": 0.21263851821422577,
      "num_tokens": 61250638.0,
      "step": 27445
    },
    {
      "entropy": 5.44442892074585,
      "epoch": 3.082888751614534,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040699169913593306,
      "loss": 5.067,
      "mean_token_accuracy": 0.20637549757957457,
      "num_tokens": 61261431.0,
      "step": 27450
    },
    {
      "entropy": 5.467871141433716,
      "epoch": 3.083450328522491,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004069591823843302,
      "loss": 5.1645,
      "mean_token_accuracy": 0.20015959441661835,
      "num_tokens": 61272726.0,
      "step": 27455
    },
    {
      "entropy": 5.450448656082154,
      "epoch": 3.0840119054304487,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004069266614310959,
      "loss": 5.0847,
      "mean_token_accuracy": 0.2108982339501381,
      "num_tokens": 61283779.0,
      "step": 27460
    },
    {
      "entropy": 5.419009828567505,
      "epoch": 3.0845734823384063,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040689413627726544,
      "loss": 5.1221,
      "mean_token_accuracy": 0.2117372676730156,
      "num_tokens": 61294477.0,
      "step": 27465
    },
    {
      "entropy": 5.444853353500366,
      "epoch": 3.085135059246364,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040686160692387455,
      "loss": 5.1907,
      "mean_token_accuracy": 0.20476775318384172,
      "num_tokens": 61305302.0,
      "step": 27470
    },
    {
      "entropy": 5.448298835754395,
      "epoch": 3.085696636154321,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040682907337195894,
      "loss": 5.0288,
      "mean_token_accuracy": 0.212177774310112,
      "num_tokens": 61315499.0,
      "step": 27475
    },
    {
      "entropy": 5.444242238998413,
      "epoch": 3.086258213062279,
      "grad_norm": 1.125,
      "learning_rate": 0.0004067965356225545,
      "loss": 5.0723,
      "mean_token_accuracy": 0.19936156123876572,
      "num_tokens": 61326654.0,
      "step": 27480
    },
    {
      "entropy": 5.4075723648071286,
      "epoch": 3.0868197899702365,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040676399367669733,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20307598561048507,
      "num_tokens": 61338475.0,
      "step": 27485
    },
    {
      "entropy": 5.377256441116333,
      "epoch": 3.0873813668781938,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004067314475354234,
      "loss": 5.1425,
      "mean_token_accuracy": 0.20647278428077698,
      "num_tokens": 61348748.0,
      "step": 27490
    },
    {
      "entropy": 5.350716543197632,
      "epoch": 3.0879429437861514,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040669889719976916,
      "loss": 5.0506,
      "mean_token_accuracy": 0.2181146949529648,
      "num_tokens": 61359466.0,
      "step": 27495
    },
    {
      "entropy": 5.38816556930542,
      "epoch": 3.088504520694109,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004066663426707709,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21696633994579315,
      "num_tokens": 61370136.0,
      "step": 27500
    },
    {
      "entropy": 5.357815599441528,
      "epoch": 3.0890660976020667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040663378394946534,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20907026827335357,
      "num_tokens": 61381012.0,
      "step": 27505
    },
    {
      "entropy": 5.383496904373169,
      "epoch": 3.089627674510024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004066012210368889,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21088650226593017,
      "num_tokens": 61393530.0,
      "step": 27510
    },
    {
      "entropy": 5.429406452178955,
      "epoch": 3.0901892514179816,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004065686539340786,
      "loss": 5.0695,
      "mean_token_accuracy": 0.21019048541784285,
      "num_tokens": 61405753.0,
      "step": 27515
    },
    {
      "entropy": 5.309476137161255,
      "epoch": 3.0907508283259393,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040653608264207134,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21437268406152726,
      "num_tokens": 61416688.0,
      "step": 27520
    },
    {
      "entropy": 5.444396257400513,
      "epoch": 3.091312405233897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040650350716190417,
      "loss": 5.1165,
      "mean_token_accuracy": 0.20811243653297423,
      "num_tokens": 61427377.0,
      "step": 27525
    },
    {
      "entropy": 5.364966154098511,
      "epoch": 3.091873982141854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004064709274946143,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21048704236745835,
      "num_tokens": 61438457.0,
      "step": 27530
    },
    {
      "entropy": 5.270884943008423,
      "epoch": 3.092435559049812,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004064383436412392,
      "loss": 5.0302,
      "mean_token_accuracy": 0.21374015808105468,
      "num_tokens": 61449380.0,
      "step": 27535
    },
    {
      "entropy": 5.4113071918487545,
      "epoch": 3.0929971359577695,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004064057556028162,
      "loss": 5.0902,
      "mean_token_accuracy": 0.2105550065636635,
      "num_tokens": 61460846.0,
      "step": 27540
    },
    {
      "entropy": 5.461072015762329,
      "epoch": 3.093558712865727,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000406373163380383,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20711333006620408,
      "num_tokens": 61471593.0,
      "step": 27545
    },
    {
      "entropy": 5.365269041061401,
      "epoch": 3.0941202897736844,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040634056697497726,
      "loss": 5.0875,
      "mean_token_accuracy": 0.20533518493175507,
      "num_tokens": 61483643.0,
      "step": 27550
    },
    {
      "entropy": 5.343554592132568,
      "epoch": 3.094681866681642,
      "grad_norm": 1.03125,
      "learning_rate": 0.000406307966387637,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2083800569176674,
      "num_tokens": 61493220.0,
      "step": 27555
    },
    {
      "entropy": 5.408980512619019,
      "epoch": 3.0952434435895997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040627536161940016,
      "loss": 5.0603,
      "mean_token_accuracy": 0.20979873985052108,
      "num_tokens": 61504372.0,
      "step": 27560
    },
    {
      "entropy": 5.311625242233276,
      "epoch": 3.0958050204975573,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004062427526713049,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20713423043489457,
      "num_tokens": 61515335.0,
      "step": 27565
    },
    {
      "entropy": 5.366089153289795,
      "epoch": 3.0963665974055146,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004062101395443894,
      "loss": 4.9964,
      "mean_token_accuracy": 0.2107372611761093,
      "num_tokens": 61526265.0,
      "step": 27570
    },
    {
      "entropy": 5.475622272491455,
      "epoch": 3.0969281743134722,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004061775222396922,
      "loss": 5.0785,
      "mean_token_accuracy": 0.20455131083726882,
      "num_tokens": 61538075.0,
      "step": 27575
    },
    {
      "entropy": 5.360054159164429,
      "epoch": 3.09748975122143,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004061449007582519,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20688650906085967,
      "num_tokens": 61548612.0,
      "step": 27580
    },
    {
      "entropy": 5.308639144897461,
      "epoch": 3.098051328129387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000406112275101107,
      "loss": 5.034,
      "mean_token_accuracy": 0.2149777352809906,
      "num_tokens": 61559702.0,
      "step": 27585
    },
    {
      "entropy": 5.404494571685791,
      "epoch": 3.0986129050373448,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040607964526929655,
      "loss": 5.0333,
      "mean_token_accuracy": 0.20723124593496323,
      "num_tokens": 61569815.0,
      "step": 27590
    },
    {
      "entropy": 5.404392862319947,
      "epoch": 3.0991744819453024,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040604701126385924,
      "loss": 5.0957,
      "mean_token_accuracy": 0.21203310340642928,
      "num_tokens": 61580050.0,
      "step": 27595
    },
    {
      "entropy": 5.438221216201782,
      "epoch": 3.09973605885326,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004060143730858342,
      "loss": 5.1013,
      "mean_token_accuracy": 0.20926881432533265,
      "num_tokens": 61590885.0,
      "step": 27600
    },
    {
      "entropy": 5.394190406799316,
      "epoch": 3.1002976357612173,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040598173073626086,
      "loss": 5.0715,
      "mean_token_accuracy": 0.20808927416801454,
      "num_tokens": 61601633.0,
      "step": 27605
    },
    {
      "entropy": 5.466530799865723,
      "epoch": 3.100859212669175,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004059490842161783,
      "loss": 5.1925,
      "mean_token_accuracy": 0.1990865021944046,
      "num_tokens": 61613244.0,
      "step": 27610
    },
    {
      "entropy": 5.408761310577392,
      "epoch": 3.1014207895771326,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040591643352662615,
      "loss": 5.049,
      "mean_token_accuracy": 0.21945154070854186,
      "num_tokens": 61624939.0,
      "step": 27615
    },
    {
      "entropy": 5.321692848205567,
      "epoch": 3.1019823664850903,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000405883778668644,
      "loss": 4.906,
      "mean_token_accuracy": 0.21740760356187822,
      "num_tokens": 61635480.0,
      "step": 27620
    },
    {
      "entropy": 5.369289255142212,
      "epoch": 3.1025439433930475,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040585111964327157,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21377938538789748,
      "num_tokens": 61645500.0,
      "step": 27625
    },
    {
      "entropy": 5.286838579177856,
      "epoch": 3.103105520301005,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040581845645154877,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21444351673126222,
      "num_tokens": 61656880.0,
      "step": 27630
    },
    {
      "entropy": 5.443379211425781,
      "epoch": 3.103667097208963,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004057857890945155,
      "loss": 5.1129,
      "mean_token_accuracy": 0.210970900952816,
      "num_tokens": 61667451.0,
      "step": 27635
    },
    {
      "entropy": 5.5185082912445065,
      "epoch": 3.1042286741169205,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040575311757321203,
      "loss": 5.2228,
      "mean_token_accuracy": 0.20185371339321137,
      "num_tokens": 61679356.0,
      "step": 27640
    },
    {
      "entropy": 5.395617294311523,
      "epoch": 3.1047902510248777,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040572044188867863,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20510266721248627,
      "num_tokens": 61690154.0,
      "step": 27645
    },
    {
      "entropy": 5.402671432495117,
      "epoch": 3.1053518279328354,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004056877620419556,
      "loss": 5.1504,
      "mean_token_accuracy": 0.2036595106124878,
      "num_tokens": 61702321.0,
      "step": 27650
    },
    {
      "entropy": 5.418680953979492,
      "epoch": 3.105913404840793,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004056550780340835,
      "loss": 5.0629,
      "mean_token_accuracy": 0.20417182594537736,
      "num_tokens": 61713498.0,
      "step": 27655
    },
    {
      "entropy": 5.540579795837402,
      "epoch": 3.1064749817487507,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004056223898661031,
      "loss": 5.1966,
      "mean_token_accuracy": 0.20398214608430862,
      "num_tokens": 61725009.0,
      "step": 27660
    },
    {
      "entropy": 5.3587456226348875,
      "epoch": 3.107036558656708,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040558969753905514,
      "loss": 4.9748,
      "mean_token_accuracy": 0.22180883586406708,
      "num_tokens": 61735652.0,
      "step": 27665
    },
    {
      "entropy": 5.226569414138794,
      "epoch": 3.1075981355646656,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004055570010539806,
      "loss": 4.9294,
      "mean_token_accuracy": 0.22017083466053008,
      "num_tokens": 61746435.0,
      "step": 27670
    },
    {
      "entropy": 5.317531442642212,
      "epoch": 3.1081597124726232,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004055243004119204,
      "loss": 4.9616,
      "mean_token_accuracy": 0.21567502468824387,
      "num_tokens": 61757523.0,
      "step": 27675
    },
    {
      "entropy": 5.474256372451782,
      "epoch": 3.108721289380581,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040549159561391595,
      "loss": 5.1086,
      "mean_token_accuracy": 0.20113369226455688,
      "num_tokens": 61769309.0,
      "step": 27680
    },
    {
      "entropy": 5.408780288696289,
      "epoch": 3.109282866288538,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004054588866610085,
      "loss": 5.1154,
      "mean_token_accuracy": 0.20833078473806382,
      "num_tokens": 61781093.0,
      "step": 27685
    },
    {
      "entropy": 5.419935894012451,
      "epoch": 3.1098444431964958,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040542617355423947,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21725351214408875,
      "num_tokens": 61792837.0,
      "step": 27690
    },
    {
      "entropy": 5.411913728713989,
      "epoch": 3.1104060201044534,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040539345629465043,
      "loss": 5.1003,
      "mean_token_accuracy": 0.20807561576366423,
      "num_tokens": 61804212.0,
      "step": 27695
    },
    {
      "entropy": 5.370340871810913,
      "epoch": 3.1109675970124107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040536073488328325,
      "loss": 5.0821,
      "mean_token_accuracy": 0.20318736433982848,
      "num_tokens": 61815436.0,
      "step": 27700
    },
    {
      "entropy": 5.462411022186279,
      "epoch": 3.1115291739203683,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004053280093211796,
      "loss": 5.1318,
      "mean_token_accuracy": 0.2065327674150467,
      "num_tokens": 61829292.0,
      "step": 27705
    },
    {
      "entropy": 5.47971544265747,
      "epoch": 3.112090750828326,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004052952796093817,
      "loss": 5.1509,
      "mean_token_accuracy": 0.21345512568950653,
      "num_tokens": 61841440.0,
      "step": 27710
    },
    {
      "entropy": 5.324461889266968,
      "epoch": 3.1126523277362836,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040526254574893154,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21012418568134308,
      "num_tokens": 61853864.0,
      "step": 27715
    },
    {
      "entropy": 5.417181539535522,
      "epoch": 3.113213904644241,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040522980774087143,
      "loss": 5.0703,
      "mean_token_accuracy": 0.20945813804864882,
      "num_tokens": 61864739.0,
      "step": 27720
    },
    {
      "entropy": 5.367512845993042,
      "epoch": 3.1137754815521985,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004051970655862437,
      "loss": 5.0859,
      "mean_token_accuracy": 0.2135436400771141,
      "num_tokens": 61875940.0,
      "step": 27725
    },
    {
      "entropy": 5.4020284652709964,
      "epoch": 3.114337058460156,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004051643192860908,
      "loss": 5.034,
      "mean_token_accuracy": 0.2102968618273735,
      "num_tokens": 61887650.0,
      "step": 27730
    },
    {
      "entropy": 5.43201642036438,
      "epoch": 3.114898635368114,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040513156884145564,
      "loss": 5.1605,
      "mean_token_accuracy": 0.20244670212268828,
      "num_tokens": 61898624.0,
      "step": 27735
    },
    {
      "entropy": 5.410088109970093,
      "epoch": 3.115460212276071,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004050988142533808,
      "loss": 5.1283,
      "mean_token_accuracy": 0.20314429104328155,
      "num_tokens": 61911017.0,
      "step": 27740
    },
    {
      "entropy": 5.4241943359375,
      "epoch": 3.1160217891840287,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004050660555229093,
      "loss": 5.1176,
      "mean_token_accuracy": 0.2031803771853447,
      "num_tokens": 61922381.0,
      "step": 27745
    },
    {
      "entropy": 5.539089870452881,
      "epoch": 3.1165833660919864,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004050332926510841,
      "loss": 5.1592,
      "mean_token_accuracy": 0.2067386254668236,
      "num_tokens": 61932876.0,
      "step": 27750
    },
    {
      "entropy": 5.3301208972930905,
      "epoch": 3.117144942999944,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004050005256389484,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2183411478996277,
      "num_tokens": 61943121.0,
      "step": 27755
    },
    {
      "entropy": 5.2945556640625,
      "epoch": 3.1177065199079013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004049677544875456,
      "loss": 4.938,
      "mean_token_accuracy": 0.2080162599682808,
      "num_tokens": 61953988.0,
      "step": 27760
    },
    {
      "entropy": 5.419624090194702,
      "epoch": 3.118268096815859,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040493497919791906,
      "loss": 5.0938,
      "mean_token_accuracy": 0.21016569584608077,
      "num_tokens": 61965111.0,
      "step": 27765
    },
    {
      "entropy": 5.505664396286011,
      "epoch": 3.1188296737238166,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040490219977111237,
      "loss": 5.1422,
      "mean_token_accuracy": 0.21375453174114228,
      "num_tokens": 61976193.0,
      "step": 27770
    },
    {
      "entropy": 5.458707714080811,
      "epoch": 3.1193912506317742,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004048694162081693,
      "loss": 5.1248,
      "mean_token_accuracy": 0.2114075243473053,
      "num_tokens": 61987554.0,
      "step": 27775
    },
    {
      "entropy": 5.517564392089843,
      "epoch": 3.1199528275397315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004048366285101336,
      "loss": 5.1587,
      "mean_token_accuracy": 0.1945674881339073,
      "num_tokens": 61998472.0,
      "step": 27780
    },
    {
      "entropy": 5.395972824096679,
      "epoch": 3.120514404447689,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004048038366780494,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20793312340974807,
      "num_tokens": 62009368.0,
      "step": 27785
    },
    {
      "entropy": 5.3373970031738285,
      "epoch": 3.121075981355647,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004047710407129606,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20725186467170714,
      "num_tokens": 62021510.0,
      "step": 27790
    },
    {
      "entropy": 5.451146030426026,
      "epoch": 3.121637558263604,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004047382406159115,
      "loss": 5.1319,
      "mean_token_accuracy": 0.2058703362941742,
      "num_tokens": 62032057.0,
      "step": 27795
    },
    {
      "entropy": 5.3722936630249025,
      "epoch": 3.1221991351715617,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004047054363879466,
      "loss": 5.0255,
      "mean_token_accuracy": 0.22270579040050506,
      "num_tokens": 62043185.0,
      "step": 27800
    },
    {
      "entropy": 5.488694190979004,
      "epoch": 3.1227607120795193,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040467262803011025,
      "loss": 5.1169,
      "mean_token_accuracy": 0.2046107530593872,
      "num_tokens": 62053483.0,
      "step": 27805
    },
    {
      "entropy": 5.424950695037841,
      "epoch": 3.123322288987477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040463981554344716,
      "loss": 5.177,
      "mean_token_accuracy": 0.19927919507026673,
      "num_tokens": 62065721.0,
      "step": 27810
    },
    {
      "entropy": 5.369228649139404,
      "epoch": 3.123883865895434,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040460699892900197,
      "loss": 5.0256,
      "mean_token_accuracy": 0.21313521713018418,
      "num_tokens": 62076251.0,
      "step": 27815
    },
    {
      "entropy": 5.394562673568726,
      "epoch": 3.124445442803392,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004045741781878198,
      "loss": 5.0483,
      "mean_token_accuracy": 0.21053803712129593,
      "num_tokens": 62086791.0,
      "step": 27820
    },
    {
      "entropy": 5.357765150070191,
      "epoch": 3.1250070197113495,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040454135332094543,
      "loss": 4.922,
      "mean_token_accuracy": 0.21566542685031892,
      "num_tokens": 62097054.0,
      "step": 27825
    },
    {
      "entropy": 5.3928991794586185,
      "epoch": 3.125568596619307,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004045085243294242,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21483320742845535,
      "num_tokens": 62108304.0,
      "step": 27830
    },
    {
      "entropy": 5.408805894851684,
      "epoch": 3.1261301735272644,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004044756912143013,
      "loss": 5.1303,
      "mean_token_accuracy": 0.2051289603114128,
      "num_tokens": 62119724.0,
      "step": 27835
    },
    {
      "entropy": 5.365106248855591,
      "epoch": 3.126691750435222,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040444285397662225,
      "loss": 4.9813,
      "mean_token_accuracy": 0.22066470384597778,
      "num_tokens": 62131021.0,
      "step": 27840
    },
    {
      "entropy": 5.439765405654907,
      "epoch": 3.1272533273431797,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004044100126174324,
      "loss": 5.1239,
      "mean_token_accuracy": 0.20138310194015502,
      "num_tokens": 62141846.0,
      "step": 27845
    },
    {
      "entropy": 5.492084884643555,
      "epoch": 3.1278149042511374,
      "grad_norm": 1.359375,
      "learning_rate": 0.00040437716713777756,
      "loss": 5.2009,
      "mean_token_accuracy": 0.20362530946731566,
      "num_tokens": 62152922.0,
      "step": 27850
    },
    {
      "entropy": 5.454377269744873,
      "epoch": 3.1283764811590946,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004043443175387036,
      "loss": 5.1517,
      "mean_token_accuracy": 0.20615992248058318,
      "num_tokens": 62164645.0,
      "step": 27855
    },
    {
      "entropy": 5.372118663787842,
      "epoch": 3.1289380580670523,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040431146382125644,
      "loss": 5.1391,
      "mean_token_accuracy": 0.2011812373995781,
      "num_tokens": 62175910.0,
      "step": 27860
    },
    {
      "entropy": 5.3320629596710205,
      "epoch": 3.12949963497501,
      "grad_norm": 0.984375,
      "learning_rate": 0.000404278605986482,
      "loss": 4.965,
      "mean_token_accuracy": 0.21617646962404252,
      "num_tokens": 62187411.0,
      "step": 27865
    },
    {
      "entropy": 5.319237852096558,
      "epoch": 3.1300612118829676,
      "grad_norm": 1.0,
      "learning_rate": 0.0004042457440354267,
      "loss": 4.999,
      "mean_token_accuracy": 0.21538869738578797,
      "num_tokens": 62199316.0,
      "step": 27870
    },
    {
      "entropy": 5.481736087799073,
      "epoch": 3.130622788790925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004042128779691368,
      "loss": 5.1559,
      "mean_token_accuracy": 0.2084455206990242,
      "num_tokens": 62211864.0,
      "step": 27875
    },
    {
      "entropy": 5.397811937332153,
      "epoch": 3.1311843656988825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004041800077886587,
      "loss": 4.9891,
      "mean_token_accuracy": 0.22080040276050567,
      "num_tokens": 62221715.0,
      "step": 27880
    },
    {
      "entropy": 5.4069263458251955,
      "epoch": 3.13174594260684,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040414713349503907,
      "loss": 5.1343,
      "mean_token_accuracy": 0.20926571190357207,
      "num_tokens": 62233198.0,
      "step": 27885
    },
    {
      "entropy": 5.387363719940185,
      "epoch": 3.1323075195147974,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040411425508932457,
      "loss": 5.04,
      "mean_token_accuracy": 0.20948956906795502,
      "num_tokens": 62244149.0,
      "step": 27890
    },
    {
      "entropy": 5.479681205749512,
      "epoch": 3.132869096422755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040408137257256225,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20748914033174515,
      "num_tokens": 62254994.0,
      "step": 27895
    },
    {
      "entropy": 5.434012794494629,
      "epoch": 3.1334306733307127,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004040484859457988,
      "loss": 5.1449,
      "mean_token_accuracy": 0.206869937479496,
      "num_tokens": 62265685.0,
      "step": 27900
    },
    {
      "entropy": 5.339555883407593,
      "epoch": 3.1339922502386703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040401559521008155,
      "loss": 5.0354,
      "mean_token_accuracy": 0.2136625826358795,
      "num_tokens": 62275765.0,
      "step": 27905
    },
    {
      "entropy": 5.306742382049561,
      "epoch": 3.1345538271466276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040398270036645775,
      "loss": 5.0393,
      "mean_token_accuracy": 0.20739591270685195,
      "num_tokens": 62286020.0,
      "step": 27910
    },
    {
      "entropy": 5.422234392166137,
      "epoch": 3.135115404054585,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004039498014159748,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21015482246875763,
      "num_tokens": 62296231.0,
      "step": 27915
    },
    {
      "entropy": 5.388809251785278,
      "epoch": 3.135676980962543,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040391689835968004,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20927929878234863,
      "num_tokens": 62307243.0,
      "step": 27920
    },
    {
      "entropy": 5.392269325256348,
      "epoch": 3.1362385578705005,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004038839911986213,
      "loss": 5.0688,
      "mean_token_accuracy": 0.2043734982609749,
      "num_tokens": 62319061.0,
      "step": 27925
    },
    {
      "entropy": 5.438141393661499,
      "epoch": 3.1368001347784578,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004038510799338462,
      "loss": 5.124,
      "mean_token_accuracy": 0.21011329889297486,
      "num_tokens": 62329477.0,
      "step": 27930
    },
    {
      "entropy": 5.39131350517273,
      "epoch": 3.1373617116864154,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004038181645664029,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2149667575955391,
      "num_tokens": 62339170.0,
      "step": 27935
    },
    {
      "entropy": 5.467151641845703,
      "epoch": 3.137923288594373,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040378524509733916,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20648448914289474,
      "num_tokens": 62350548.0,
      "step": 27940
    },
    {
      "entropy": 5.374819374084472,
      "epoch": 3.1384848655023307,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004037523215277033,
      "loss": 4.9895,
      "mean_token_accuracy": 0.2139353483915329,
      "num_tokens": 62361105.0,
      "step": 27945
    },
    {
      "entropy": 5.370611047744751,
      "epoch": 3.139046442410288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004037193938585436,
      "loss": 5.0828,
      "mean_token_accuracy": 0.2113683730363846,
      "num_tokens": 62372048.0,
      "step": 27950
    },
    {
      "entropy": 5.378255701065063,
      "epoch": 3.1396080193182456,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004036864620909085,
      "loss": 5.0872,
      "mean_token_accuracy": 0.2041005551815033,
      "num_tokens": 62382936.0,
      "step": 27955
    },
    {
      "entropy": 5.3852311134338375,
      "epoch": 3.1401695962262033,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004036535262258465,
      "loss": 4.9798,
      "mean_token_accuracy": 0.2139171227812767,
      "num_tokens": 62393656.0,
      "step": 27960
    },
    {
      "entropy": 5.394815826416016,
      "epoch": 3.140731173134161,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040362058626440637,
      "loss": 5.1021,
      "mean_token_accuracy": 0.21242287904024124,
      "num_tokens": 62404234.0,
      "step": 27965
    },
    {
      "entropy": 5.4129030227661135,
      "epoch": 3.141292750042118,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040358764220763685,
      "loss": 5.0466,
      "mean_token_accuracy": 0.20921226292848588,
      "num_tokens": 62416694.0,
      "step": 27970
    },
    {
      "entropy": 5.381912326812744,
      "epoch": 3.141854326950076,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00040355469405658687,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20664192885160446,
      "num_tokens": 62429437.0,
      "step": 27975
    },
    {
      "entropy": 5.429514455795288,
      "epoch": 3.1424159038580335,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004035217418123056,
      "loss": 5.0304,
      "mean_token_accuracy": 0.21849615424871444,
      "num_tokens": 62440085.0,
      "step": 27980
    },
    {
      "entropy": 5.41820502281189,
      "epoch": 3.1429774807659907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040348878547584227,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20813782811164855,
      "num_tokens": 62451285.0,
      "step": 27985
    },
    {
      "entropy": 5.373146629333496,
      "epoch": 3.1435390576739484,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040345582504824624,
      "loss": 5.0325,
      "mean_token_accuracy": 0.21408046633005143,
      "num_tokens": 62461777.0,
      "step": 27990
    },
    {
      "entropy": 5.404027652740479,
      "epoch": 3.144100634581906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004034228605305668,
      "loss": 5.0939,
      "mean_token_accuracy": 0.21634501218795776,
      "num_tokens": 62472463.0,
      "step": 27995
    },
    {
      "entropy": 5.290450763702393,
      "epoch": 3.1446622114898637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040338989192385377,
      "loss": 5.0221,
      "mean_token_accuracy": 0.21135189086198808,
      "num_tokens": 62484168.0,
      "step": 28000
    },
    {
      "entropy": 5.434889221191407,
      "epoch": 3.145223788397821,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004033569192291567,
      "loss": 5.0949,
      "mean_token_accuracy": 0.21173793077468872,
      "num_tokens": 62495187.0,
      "step": 28005
    },
    {
      "entropy": 5.428685474395752,
      "epoch": 3.1457853653057786,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004033239424475256,
      "loss": 5.1353,
      "mean_token_accuracy": 0.20254740118980408,
      "num_tokens": 62505936.0,
      "step": 28010
    },
    {
      "entropy": 5.488847446441651,
      "epoch": 3.1463469422137362,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040329096158001033,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20842781960964202,
      "num_tokens": 62516486.0,
      "step": 28015
    },
    {
      "entropy": 5.361659193038941,
      "epoch": 3.146908519121694,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004032579766276612,
      "loss": 5.0124,
      "mean_token_accuracy": 0.215682153403759,
      "num_tokens": 62528349.0,
      "step": 28020
    },
    {
      "entropy": 5.266450023651123,
      "epoch": 3.147470096029651,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040322498759152824,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21442334949970246,
      "num_tokens": 62538698.0,
      "step": 28025
    },
    {
      "entropy": 5.490656042098999,
      "epoch": 3.1480316729376088,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040319199447266205,
      "loss": 5.1947,
      "mean_token_accuracy": 0.2042116791009903,
      "num_tokens": 62550891.0,
      "step": 28030
    },
    {
      "entropy": 5.453319644927978,
      "epoch": 3.1485932498455664,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004031589972721129,
      "loss": 5.0462,
      "mean_token_accuracy": 0.210380257666111,
      "num_tokens": 62560553.0,
      "step": 28035
    },
    {
      "entropy": 5.443738603591919,
      "epoch": 3.149154826753524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004031259959909317,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20956064611673356,
      "num_tokens": 62571281.0,
      "step": 28040
    },
    {
      "entropy": 5.2905693531036375,
      "epoch": 3.1497164036614813,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000403092990630169,
      "loss": 5.0353,
      "mean_token_accuracy": 0.2130683869123459,
      "num_tokens": 62583489.0,
      "step": 28045
    },
    {
      "entropy": 5.324041366577148,
      "epoch": 3.150277980569439,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004030599811908758,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21763916462659835,
      "num_tokens": 62594155.0,
      "step": 28050
    },
    {
      "entropy": 5.358155584335327,
      "epoch": 3.1508395574773966,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004030269676741032,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21360611021518708,
      "num_tokens": 62605464.0,
      "step": 28055
    },
    {
      "entropy": 5.400675344467163,
      "epoch": 3.1514011343853543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004029939500809023,
      "loss": 5.071,
      "mean_token_accuracy": 0.21409667730331422,
      "num_tokens": 62617343.0,
      "step": 28060
    },
    {
      "entropy": 5.365217876434326,
      "epoch": 3.1519627112933115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004029609284123243,
      "loss": 5.051,
      "mean_token_accuracy": 0.21621219515800477,
      "num_tokens": 62628461.0,
      "step": 28065
    },
    {
      "entropy": 5.379937028884887,
      "epoch": 3.152524288201269,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040292790266942074,
      "loss": 5.0255,
      "mean_token_accuracy": 0.21413537710905076,
      "num_tokens": 62638947.0,
      "step": 28070
    },
    {
      "entropy": 5.280916547775268,
      "epoch": 3.153085865109227,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040289487285324324,
      "loss": 4.8767,
      "mean_token_accuracy": 0.21882817894220352,
      "num_tokens": 62648840.0,
      "step": 28075
    },
    {
      "entropy": 5.443727922439575,
      "epoch": 3.153647442017184,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004028618389648433,
      "loss": 5.1523,
      "mean_token_accuracy": 0.20561097413301468,
      "num_tokens": 62659491.0,
      "step": 28080
    },
    {
      "entropy": 5.457854461669922,
      "epoch": 3.1542090189251417,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004028288010052728,
      "loss": 5.12,
      "mean_token_accuracy": 0.20365190207958223,
      "num_tokens": 62669038.0,
      "step": 28085
    },
    {
      "entropy": 5.340935087203979,
      "epoch": 3.1547705958330994,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004027957589755837,
      "loss": 5.0087,
      "mean_token_accuracy": 0.21984484195709228,
      "num_tokens": 62680280.0,
      "step": 28090
    },
    {
      "entropy": 5.411754894256592,
      "epoch": 3.155332172741057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040276271287682813,
      "loss": 5.0733,
      "mean_token_accuracy": 0.2121509537100792,
      "num_tokens": 62691767.0,
      "step": 28095
    },
    {
      "entropy": 5.453503751754761,
      "epoch": 3.1558937496490143,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040272966271005816,
      "loss": 5.0637,
      "mean_token_accuracy": 0.20902395993471146,
      "num_tokens": 62701721.0,
      "step": 28100
    },
    {
      "entropy": 5.419015121459961,
      "epoch": 3.156455326556972,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040269660847632626,
      "loss": 5.0974,
      "mean_token_accuracy": 0.2068915143609047,
      "num_tokens": 62713238.0,
      "step": 28105
    },
    {
      "entropy": 5.306504106521606,
      "epoch": 3.1570169034649296,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00040266355017668475,
      "loss": 4.95,
      "mean_token_accuracy": 0.2190881311893463,
      "num_tokens": 62724457.0,
      "step": 28110
    },
    {
      "entropy": 5.386609315872192,
      "epoch": 3.1575784803728872,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004026304878121864,
      "loss": 5.08,
      "mean_token_accuracy": 0.21441061347723006,
      "num_tokens": 62735178.0,
      "step": 28115
    },
    {
      "entropy": 5.4241656303405765,
      "epoch": 3.1581400572808445,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040259742138388366,
      "loss": 5.0596,
      "mean_token_accuracy": 0.2138619527220726,
      "num_tokens": 62746767.0,
      "step": 28120
    },
    {
      "entropy": 5.399972724914551,
      "epoch": 3.158701634188802,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040256435089282966,
      "loss": 5.0594,
      "mean_token_accuracy": 0.2143801122903824,
      "num_tokens": 62757963.0,
      "step": 28125
    },
    {
      "entropy": 5.424853086471558,
      "epoch": 3.15926321109676,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004025312763400772,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20117214769124986,
      "num_tokens": 62769763.0,
      "step": 28130
    },
    {
      "entropy": 5.2908038139343265,
      "epoch": 3.1598247880047174,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004024981977266794,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2196861118078232,
      "num_tokens": 62780947.0,
      "step": 28135
    },
    {
      "entropy": 5.411683225631714,
      "epoch": 3.1603863649126747,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004024651150536896,
      "loss": 5.159,
      "mean_token_accuracy": 0.19475050419569015,
      "num_tokens": 62793003.0,
      "step": 28140
    },
    {
      "entropy": 5.4262713432312015,
      "epoch": 3.1609479418206323,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004024320283221611,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20729157477617263,
      "num_tokens": 62804467.0,
      "step": 28145
    },
    {
      "entropy": 5.357975816726684,
      "epoch": 3.16150951872859,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040239893753314735,
      "loss": 5.0102,
      "mean_token_accuracy": 0.22001030445098876,
      "num_tokens": 62815834.0,
      "step": 28150
    },
    {
      "entropy": 5.364781665802002,
      "epoch": 3.1620710956365476,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004023658426877021,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20624707341194154,
      "num_tokens": 62827277.0,
      "step": 28155
    },
    {
      "entropy": 5.433741331100464,
      "epoch": 3.162632672544505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040233274378687897,
      "loss": 4.9778,
      "mean_token_accuracy": 0.22276339828968048,
      "num_tokens": 62838652.0,
      "step": 28160
    },
    {
      "entropy": 5.443546867370605,
      "epoch": 3.1631942494524625,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040229964083173187,
      "loss": 5.123,
      "mean_token_accuracy": 0.2052655965089798,
      "num_tokens": 62849448.0,
      "step": 28165
    },
    {
      "entropy": 5.406308317184449,
      "epoch": 3.16375582636042,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040226653382331484,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21640822291374207,
      "num_tokens": 62860840.0,
      "step": 28170
    },
    {
      "entropy": 5.457951641082763,
      "epoch": 3.1643174032683774,
      "grad_norm": 1.09375,
      "learning_rate": 0.000402233422762682,
      "loss": 5.0737,
      "mean_token_accuracy": 0.21558486372232438,
      "num_tokens": 62871321.0,
      "step": 28175
    },
    {
      "entropy": 5.299130868911743,
      "epoch": 3.164878980176335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040220030765088764,
      "loss": 4.9855,
      "mean_token_accuracy": 0.21723635494709015,
      "num_tokens": 62882865.0,
      "step": 28180
    },
    {
      "entropy": 5.341673564910889,
      "epoch": 3.1654405570842927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004021671884889862,
      "loss": 5.0515,
      "mean_token_accuracy": 0.2077361285686493,
      "num_tokens": 62894122.0,
      "step": 28185
    },
    {
      "entropy": 5.415648460388184,
      "epoch": 3.1660021339922504,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004021340652780321,
      "loss": 5.1234,
      "mean_token_accuracy": 0.20084903985261918,
      "num_tokens": 62904797.0,
      "step": 28190
    },
    {
      "entropy": 5.427558708190918,
      "epoch": 3.1665637109002076,
      "grad_norm": 1.0,
      "learning_rate": 0.0004021009380190801,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2106771722435951,
      "num_tokens": 62914739.0,
      "step": 28195
    },
    {
      "entropy": 5.356130456924438,
      "epoch": 3.1671252878081653,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040206780671318495,
      "loss": 5.0561,
      "mean_token_accuracy": 0.2133734792470932,
      "num_tokens": 62925183.0,
      "step": 28200
    },
    {
      "entropy": 5.358009910583496,
      "epoch": 3.167686864716123,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004020346713614015,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20596429258584975,
      "num_tokens": 62936197.0,
      "step": 28205
    },
    {
      "entropy": 5.395763111114502,
      "epoch": 3.1682484416240806,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040200153196478486,
      "loss": 5.0925,
      "mean_token_accuracy": 0.2067570447921753,
      "num_tokens": 62947571.0,
      "step": 28210
    },
    {
      "entropy": 5.380672597885132,
      "epoch": 3.168810018532038,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004019683885243902,
      "loss": 5.0785,
      "mean_token_accuracy": 0.21214079856872559,
      "num_tokens": 62957614.0,
      "step": 28215
    },
    {
      "entropy": 5.353319311141968,
      "epoch": 3.1693715954399955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004019352410412728,
      "loss": 4.9834,
      "mean_token_accuracy": 0.21977371722459793,
      "num_tokens": 62968840.0,
      "step": 28220
    },
    {
      "entropy": 5.387166929244995,
      "epoch": 3.169933172347953,
      "grad_norm": 1.0,
      "learning_rate": 0.00040190208951648807,
      "loss": 5.0854,
      "mean_token_accuracy": 0.20981072187423705,
      "num_tokens": 62979956.0,
      "step": 28225
    },
    {
      "entropy": 5.391596126556396,
      "epoch": 3.170494749255911,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004018689339510916,
      "loss": 5.017,
      "mean_token_accuracy": 0.2134726330637932,
      "num_tokens": 62991352.0,
      "step": 28230
    },
    {
      "entropy": 5.539651823043823,
      "epoch": 3.171056326163868,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040183577434613904,
      "loss": 5.2219,
      "mean_token_accuracy": 0.201230089366436,
      "num_tokens": 63002738.0,
      "step": 28235
    },
    {
      "entropy": 5.46684718132019,
      "epoch": 3.1716179030718257,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004018026107026863,
      "loss": 5.143,
      "mean_token_accuracy": 0.20960735976696016,
      "num_tokens": 63013989.0,
      "step": 28240
    },
    {
      "entropy": 5.348393201828003,
      "epoch": 3.1721794799797833,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004017694430217892,
      "loss": 5.0042,
      "mean_token_accuracy": 0.20906526744365692,
      "num_tokens": 63026158.0,
      "step": 28245
    },
    {
      "entropy": 5.276983785629272,
      "epoch": 3.172741056887741,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004017362713045039,
      "loss": 4.9587,
      "mean_token_accuracy": 0.21678174138069153,
      "num_tokens": 63036432.0,
      "step": 28250
    },
    {
      "entropy": 5.438896131515503,
      "epoch": 3.173302633795698,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040170309555188655,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20916943848133088,
      "num_tokens": 63047151.0,
      "step": 28255
    },
    {
      "entropy": 5.342494201660156,
      "epoch": 3.173864210703656,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004016699157649936,
      "loss": 5.0614,
      "mean_token_accuracy": 0.2089928060770035,
      "num_tokens": 63059514.0,
      "step": 28260
    },
    {
      "entropy": 5.475094127655029,
      "epoch": 3.1744257876116135,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004016367319448812,
      "loss": 5.2785,
      "mean_token_accuracy": 0.20070800930261612,
      "num_tokens": 63071037.0,
      "step": 28265
    },
    {
      "entropy": 5.445153617858887,
      "epoch": 3.1749873645195708,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040160354409260625,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20755532383918762,
      "num_tokens": 63083615.0,
      "step": 28270
    },
    {
      "entropy": 5.390350198745727,
      "epoch": 3.1755489414275284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004015703522092254,
      "loss": 5.0918,
      "mean_token_accuracy": 0.21055447012186052,
      "num_tokens": 63094377.0,
      "step": 28275
    },
    {
      "entropy": 5.328070402145386,
      "epoch": 3.176110518335486,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004015371562957954,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21093183010816574,
      "num_tokens": 63105108.0,
      "step": 28280
    },
    {
      "entropy": 5.481532192230224,
      "epoch": 3.1766720952434437,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004015039563533732,
      "loss": 5.179,
      "mean_token_accuracy": 0.20188043117523194,
      "num_tokens": 63116576.0,
      "step": 28285
    },
    {
      "entropy": 5.467678260803223,
      "epoch": 3.177233672151401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004014707523830161,
      "loss": 5.1376,
      "mean_token_accuracy": 0.20718589425086975,
      "num_tokens": 63128343.0,
      "step": 28290
    },
    {
      "entropy": 5.4789817333221436,
      "epoch": 3.1777952490593586,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004014375443857811,
      "loss": 5.2628,
      "mean_token_accuracy": 0.20177952349185943,
      "num_tokens": 63140309.0,
      "step": 28295
    },
    {
      "entropy": 5.461684131622315,
      "epoch": 3.1783568259673163,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004014043323627257,
      "loss": 5.1217,
      "mean_token_accuracy": 0.21024358868598939,
      "num_tokens": 63150825.0,
      "step": 28300
    },
    {
      "entropy": 5.427690601348877,
      "epoch": 3.178918402875274,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004013711163149072,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21917017847299575,
      "num_tokens": 63161296.0,
      "step": 28305
    },
    {
      "entropy": 5.3100550174713135,
      "epoch": 3.179479979783231,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040133789624338353,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21330492645502092,
      "num_tokens": 63171447.0,
      "step": 28310
    },
    {
      "entropy": 5.332453060150146,
      "epoch": 3.180041556691189,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040130467214921216,
      "loss": 5.095,
      "mean_token_accuracy": 0.19951386898756027,
      "num_tokens": 63182860.0,
      "step": 28315
    },
    {
      "entropy": 5.468055629730225,
      "epoch": 3.1806031335991465,
      "grad_norm": 1.09375,
      "learning_rate": 0.000401271444033451,
      "loss": 5.1018,
      "mean_token_accuracy": 0.20608717799186707,
      "num_tokens": 63192894.0,
      "step": 28320
    },
    {
      "entropy": 5.4533641815185545,
      "epoch": 3.181164710507104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004012382118971582,
      "loss": 5.1335,
      "mean_token_accuracy": 0.2035526528954506,
      "num_tokens": 63204013.0,
      "step": 28325
    },
    {
      "entropy": 5.358470106124878,
      "epoch": 3.1817262874150614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004012049757413917,
      "loss": 5.0244,
      "mean_token_accuracy": 0.2130138710141182,
      "num_tokens": 63215079.0,
      "step": 28330
    },
    {
      "entropy": 5.348422527313232,
      "epoch": 3.182287864323019,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040117173556720976,
      "loss": 5.062,
      "mean_token_accuracy": 0.20759831368923187,
      "num_tokens": 63225185.0,
      "step": 28335
    },
    {
      "entropy": 5.47285532951355,
      "epoch": 3.1828494412309767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040113849137567094,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20581945180892944,
      "num_tokens": 63237490.0,
      "step": 28340
    },
    {
      "entropy": 5.363579559326172,
      "epoch": 3.1834110181389343,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004011052431678336,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21477397829294204,
      "num_tokens": 63247774.0,
      "step": 28345
    },
    {
      "entropy": 5.317459392547607,
      "epoch": 3.1839725950468916,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004010719909447564,
      "loss": 5.0641,
      "mean_token_accuracy": 0.21031411737203598,
      "num_tokens": 63258953.0,
      "step": 28350
    },
    {
      "entropy": 5.38041672706604,
      "epoch": 3.1845341719548492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004010387347074981,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20946240276098252,
      "num_tokens": 63269670.0,
      "step": 28355
    },
    {
      "entropy": 5.3769121170043945,
      "epoch": 3.185095748862807,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040100547445711765,
      "loss": 5.0615,
      "mean_token_accuracy": 0.21452366858720778,
      "num_tokens": 63280644.0,
      "step": 28360
    },
    {
      "entropy": 5.435745334625244,
      "epoch": 3.185657325770764,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040097221019467393,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20811499357223512,
      "num_tokens": 63292328.0,
      "step": 28365
    },
    {
      "entropy": 5.291597557067871,
      "epoch": 3.1862189026787218,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004009389419212262,
      "loss": 4.9218,
      "mean_token_accuracy": 0.21554968655109405,
      "num_tokens": 63302849.0,
      "step": 28370
    },
    {
      "entropy": 5.310023069381714,
      "epoch": 3.1867804795866794,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040090566963783375,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20935720950365067,
      "num_tokens": 63314460.0,
      "step": 28375
    },
    {
      "entropy": 5.418254280090332,
      "epoch": 3.187342056494637,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040087239334555603,
      "loss": 5.0818,
      "mean_token_accuracy": 0.2132833182811737,
      "num_tokens": 63325713.0,
      "step": 28380
    },
    {
      "entropy": 5.384374618530273,
      "epoch": 3.1879036334025943,
      "grad_norm": 1.0,
      "learning_rate": 0.00040083911304545234,
      "loss": 5.0922,
      "mean_token_accuracy": 0.20856288820505142,
      "num_tokens": 63338287.0,
      "step": 28385
    },
    {
      "entropy": 5.4218504428863525,
      "epoch": 3.188465210310552,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040080582873858254,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2171464294195175,
      "num_tokens": 63349592.0,
      "step": 28390
    },
    {
      "entropy": 5.412782287597656,
      "epoch": 3.1890267872185096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004007725404260064,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21879738420248032,
      "num_tokens": 63359913.0,
      "step": 28395
    },
    {
      "entropy": 5.397561120986938,
      "epoch": 3.1895883641264673,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004007392481087837,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20600663125514984,
      "num_tokens": 63370677.0,
      "step": 28400
    },
    {
      "entropy": 5.403140211105347,
      "epoch": 3.1901499410344245,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004007059517879747,
      "loss": 5.0149,
      "mean_token_accuracy": 0.220577572286129,
      "num_tokens": 63381467.0,
      "step": 28405
    },
    {
      "entropy": 5.407635354995728,
      "epoch": 3.190711517942382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004006726514646393,
      "loss": 5.1277,
      "mean_token_accuracy": 0.20360197573900224,
      "num_tokens": 63392808.0,
      "step": 28410
    },
    {
      "entropy": 5.351781177520752,
      "epoch": 3.19127309485034,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040063934713983803,
      "loss": 4.9886,
      "mean_token_accuracy": 0.2114964246749878,
      "num_tokens": 63405206.0,
      "step": 28415
    },
    {
      "entropy": 5.247720527648926,
      "epoch": 3.1918346717582975,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040060603881463127,
      "loss": 4.9779,
      "mean_token_accuracy": 0.22029895782470704,
      "num_tokens": 63416923.0,
      "step": 28420
    },
    {
      "entropy": 5.396749448776245,
      "epoch": 3.1923962486662547,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040057272649007944,
      "loss": 5.0375,
      "mean_token_accuracy": 0.21746097803115844,
      "num_tokens": 63426681.0,
      "step": 28425
    },
    {
      "entropy": 5.474124622344971,
      "epoch": 3.1929578255742124,
      "grad_norm": 1.375,
      "learning_rate": 0.00040053941016724326,
      "loss": 5.0209,
      "mean_token_accuracy": 0.2100441887974739,
      "num_tokens": 63437299.0,
      "step": 28430
    },
    {
      "entropy": 5.4104818344116214,
      "epoch": 3.19351940248217,
      "grad_norm": 1.0,
      "learning_rate": 0.00040050608984718364,
      "loss": 5.1265,
      "mean_token_accuracy": 0.21083631217479706,
      "num_tokens": 63449075.0,
      "step": 28435
    },
    {
      "entropy": 5.346947813034058,
      "epoch": 3.1940809793901277,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004004727655309614,
      "loss": 5.0934,
      "mean_token_accuracy": 0.20612076967954635,
      "num_tokens": 63460592.0,
      "step": 28440
    },
    {
      "entropy": 5.4009459018707275,
      "epoch": 3.194642556298085,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040043943721963766,
      "loss": 5.0881,
      "mean_token_accuracy": 0.21108666509389878,
      "num_tokens": 63471043.0,
      "step": 28445
    },
    {
      "entropy": 5.424647569656372,
      "epoch": 3.1952041332060426,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040040610491427357,
      "loss": 5.1776,
      "mean_token_accuracy": 0.20492968410253526,
      "num_tokens": 63482556.0,
      "step": 28450
    },
    {
      "entropy": 5.546213579177857,
      "epoch": 3.1957657101140002,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004003727686159305,
      "loss": 5.2191,
      "mean_token_accuracy": 0.19550352543592453,
      "num_tokens": 63493973.0,
      "step": 28455
    },
    {
      "entropy": 5.4872173309326175,
      "epoch": 3.1963272870219575,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040033942832566984,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20103065967559813,
      "num_tokens": 63504489.0,
      "step": 28460
    },
    {
      "entropy": 5.45712947845459,
      "epoch": 3.196888863929915,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004003060840445531,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20569583475589753,
      "num_tokens": 63516688.0,
      "step": 28465
    },
    {
      "entropy": 5.497138929367066,
      "epoch": 3.1974504408378728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004002727357736422,
      "loss": 5.186,
      "mean_token_accuracy": 0.19955952763557433,
      "num_tokens": 63528680.0,
      "step": 28470
    },
    {
      "entropy": 5.417816781997681,
      "epoch": 3.1980120177458304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040023938351399865,
      "loss": 5.1244,
      "mean_token_accuracy": 0.20729057341814042,
      "num_tokens": 63540489.0,
      "step": 28475
    },
    {
      "entropy": 5.416037559509277,
      "epoch": 3.1985735946537877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004002060272666846,
      "loss": 5.114,
      "mean_token_accuracy": 0.20621969401836396,
      "num_tokens": 63551974.0,
      "step": 28480
    },
    {
      "entropy": 5.472271680831909,
      "epoch": 3.1991351715617453,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004001726670327621,
      "loss": 5.1608,
      "mean_token_accuracy": 0.20238917022943498,
      "num_tokens": 63564437.0,
      "step": 28485
    },
    {
      "entropy": 5.408360862731934,
      "epoch": 3.199696748469703,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004001393028132933,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2172163262963295,
      "num_tokens": 63575041.0,
      "step": 28490
    },
    {
      "entropy": 5.3929835796356205,
      "epoch": 3.2002583253776606,
      "grad_norm": 1.0,
      "learning_rate": 0.00040010593460934065,
      "loss": 5.0741,
      "mean_token_accuracy": 0.21043936312198638,
      "num_tokens": 63587381.0,
      "step": 28495
    },
    {
      "entropy": 5.441640377044678,
      "epoch": 3.200819902285618,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040007256242196644,
      "loss": 5.141,
      "mean_token_accuracy": 0.2037556990981102,
      "num_tokens": 63599206.0,
      "step": 28500
    },
    {
      "entropy": 5.436296463012695,
      "epoch": 3.2013814791935755,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040003918625223335,
      "loss": 5.0615,
      "mean_token_accuracy": 0.20868833661079406,
      "num_tokens": 63609557.0,
      "step": 28505
    },
    {
      "entropy": 5.4003137111663815,
      "epoch": 3.201943056101533,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040000580610120406,
      "loss": 5.0471,
      "mean_token_accuracy": 0.2103464812040329,
      "num_tokens": 63620423.0,
      "step": 28510
    },
    {
      "entropy": 5.5319455623626705,
      "epoch": 3.202504633009491,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003999724219699415,
      "loss": 5.1965,
      "mean_token_accuracy": 0.2069829598069191,
      "num_tokens": 63632222.0,
      "step": 28515
    },
    {
      "entropy": 5.367495536804199,
      "epoch": 3.203066209917448,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003999390338595085,
      "loss": 5.0692,
      "mean_token_accuracy": 0.20581168979406356,
      "num_tokens": 63643077.0,
      "step": 28520
    },
    {
      "entropy": 5.340481472015381,
      "epoch": 3.2036277868254057,
      "grad_norm": 1.125,
      "learning_rate": 0.00039990564177096817,
      "loss": 4.9724,
      "mean_token_accuracy": 0.22106637954711914,
      "num_tokens": 63653642.0,
      "step": 28525
    },
    {
      "entropy": 5.392149591445923,
      "epoch": 3.2041893637333634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003998722457053838,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20577940344810486,
      "num_tokens": 63665646.0,
      "step": 28530
    },
    {
      "entropy": 5.516965532302857,
      "epoch": 3.204750940641321,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039983884566381866,
      "loss": 5.1894,
      "mean_token_accuracy": 0.20132894068956375,
      "num_tokens": 63677928.0,
      "step": 28535
    },
    {
      "entropy": 5.411691045761108,
      "epoch": 3.2053125175492783,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003998054416473363,
      "loss": 5.0578,
      "mean_token_accuracy": 0.20992235988378524,
      "num_tokens": 63688670.0,
      "step": 28540
    },
    {
      "entropy": 5.469873523712158,
      "epoch": 3.205874094457236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039977203365700026,
      "loss": 5.1626,
      "mean_token_accuracy": 0.20944867581129073,
      "num_tokens": 63699963.0,
      "step": 28545
    },
    {
      "entropy": 5.452686500549317,
      "epoch": 3.2064356713651936,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003997386216938743,
      "loss": 5.1677,
      "mean_token_accuracy": 0.20281137228012086,
      "num_tokens": 63711333.0,
      "step": 28550
    },
    {
      "entropy": 5.389166355133057,
      "epoch": 3.206997248273151,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039970520575902216,
      "loss": 5.0216,
      "mean_token_accuracy": 0.2139328807592392,
      "num_tokens": 63721916.0,
      "step": 28555
    },
    {
      "entropy": 5.33886399269104,
      "epoch": 3.2075588251811085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000399671785853508,
      "loss": 5.0889,
      "mean_token_accuracy": 0.20826449394226074,
      "num_tokens": 63733313.0,
      "step": 28560
    },
    {
      "entropy": 5.450829124450683,
      "epoch": 3.208120402089066,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039963836197839584,
      "loss": 5.1025,
      "mean_token_accuracy": 0.20751923173666,
      "num_tokens": 63743249.0,
      "step": 28565
    },
    {
      "entropy": 5.421046733856201,
      "epoch": 3.208681978997024,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039960493413474977,
      "loss": 5.0859,
      "mean_token_accuracy": 0.2047840029001236,
      "num_tokens": 63754355.0,
      "step": 28570
    },
    {
      "entropy": 5.616373348236084,
      "epoch": 3.209243555904981,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039957150232363433,
      "loss": 5.2407,
      "mean_token_accuracy": 0.2013968363404274,
      "num_tokens": 63766620.0,
      "step": 28575
    },
    {
      "entropy": 5.413175344467163,
      "epoch": 3.2098051328129387,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000399538066546114,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2153948575258255,
      "num_tokens": 63777195.0,
      "step": 28580
    },
    {
      "entropy": 5.251968717575073,
      "epoch": 3.2103667097208963,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039950462680325336,
      "loss": 5.0227,
      "mean_token_accuracy": 0.2120744451880455,
      "num_tokens": 63787653.0,
      "step": 28585
    },
    {
      "entropy": 5.384004306793213,
      "epoch": 3.210928286628854,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000399471183096117,
      "loss": 5.0916,
      "mean_token_accuracy": 0.2100202724337578,
      "num_tokens": 63799371.0,
      "step": 28590
    },
    {
      "entropy": 5.3632035732269285,
      "epoch": 3.211489863536811,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003994377354257698,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22429251968860625,
      "num_tokens": 63809791.0,
      "step": 28595
    },
    {
      "entropy": 5.405872392654419,
      "epoch": 3.212051440444769,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039940428379327703,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20735361874103547,
      "num_tokens": 63819972.0,
      "step": 28600
    },
    {
      "entropy": 5.282958841323852,
      "epoch": 3.2126130173527265,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003993708281997036,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21166979223489762,
      "num_tokens": 63830144.0,
      "step": 28605
    },
    {
      "entropy": 5.4276477813720705,
      "epoch": 3.213174594260684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003993373686461147,
      "loss": 5.1164,
      "mean_token_accuracy": 0.20310243368148803,
      "num_tokens": 63842393.0,
      "step": 28610
    },
    {
      "entropy": 5.4160240650177,
      "epoch": 3.2137361711686414,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003993039051335758,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20646465420722962,
      "num_tokens": 63853392.0,
      "step": 28615
    },
    {
      "entropy": 5.408242559432983,
      "epoch": 3.214297748076599,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003992704376631523,
      "loss": 5.076,
      "mean_token_accuracy": 0.2094378799200058,
      "num_tokens": 63866037.0,
      "step": 28620
    },
    {
      "entropy": 5.4197062969207765,
      "epoch": 3.2148593249845567,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003992369662359099,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21735747903585434,
      "num_tokens": 63876264.0,
      "step": 28625
    },
    {
      "entropy": 5.321604824066162,
      "epoch": 3.2154209018925144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003992034908529143,
      "loss": 4.9593,
      "mean_token_accuracy": 0.2183077111840248,
      "num_tokens": 63887545.0,
      "step": 28630
    },
    {
      "entropy": 5.335014152526855,
      "epoch": 3.2159824788004716,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039917001151523144,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20897796005010605,
      "num_tokens": 63897689.0,
      "step": 28635
    },
    {
      "entropy": 5.446640157699585,
      "epoch": 3.2165440557084293,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003991365282239272,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20989546030759812,
      "num_tokens": 63908884.0,
      "step": 28640
    },
    {
      "entropy": 5.407545757293701,
      "epoch": 3.217105632616387,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039910304098006775,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2112713560461998,
      "num_tokens": 63918515.0,
      "step": 28645
    },
    {
      "entropy": 5.298788070678711,
      "epoch": 3.217667209524344,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003990695497847193,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21219899505376816,
      "num_tokens": 63928802.0,
      "step": 28650
    },
    {
      "entropy": 5.390778064727783,
      "epoch": 3.218228786432302,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003990360546389484,
      "loss": 5.0433,
      "mean_token_accuracy": 0.21562667191028595,
      "num_tokens": 63940770.0,
      "step": 28655
    },
    {
      "entropy": 5.499038791656494,
      "epoch": 3.2187903633402595,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039900255554382126,
      "loss": 5.1033,
      "mean_token_accuracy": 0.21429233849048615,
      "num_tokens": 63951851.0,
      "step": 28660
    },
    {
      "entropy": 5.415889024734497,
      "epoch": 3.219351940248217,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039896905250040473,
      "loss": 5.087,
      "mean_token_accuracy": 0.21313603818416596,
      "num_tokens": 63962768.0,
      "step": 28665
    },
    {
      "entropy": 5.361529588699341,
      "epoch": 3.2199135171561744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039893554550976545,
      "loss": 5.0885,
      "mean_token_accuracy": 0.20271364748477935,
      "num_tokens": 63973855.0,
      "step": 28670
    },
    {
      "entropy": 5.498661947250366,
      "epoch": 3.220475094064132,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003989020345729705,
      "loss": 5.2124,
      "mean_token_accuracy": 0.20209450721740724,
      "num_tokens": 63986342.0,
      "step": 28675
    },
    {
      "entropy": 5.412326860427856,
      "epoch": 3.2210366709720897,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039886851969108653,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21272003501653672,
      "num_tokens": 63997080.0,
      "step": 28680
    },
    {
      "entropy": 5.3629402160644535,
      "epoch": 3.2215982478800473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039883500086518086,
      "loss": 5.0369,
      "mean_token_accuracy": 0.21081904470920562,
      "num_tokens": 64007373.0,
      "step": 28685
    },
    {
      "entropy": 5.398982429504395,
      "epoch": 3.2221598247880046,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039880147809632084,
      "loss": 5.1452,
      "mean_token_accuracy": 0.2108266457915306,
      "num_tokens": 64018429.0,
      "step": 28690
    },
    {
      "entropy": 5.384691953659058,
      "epoch": 3.222721401695962,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039876795138557363,
      "loss": 5.0758,
      "mean_token_accuracy": 0.21195996701717376,
      "num_tokens": 64028633.0,
      "step": 28695
    },
    {
      "entropy": 5.4782429218292235,
      "epoch": 3.22328297860392,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003987344207340069,
      "loss": 5.1024,
      "mean_token_accuracy": 0.21205249130725862,
      "num_tokens": 64040860.0,
      "step": 28700
    },
    {
      "entropy": 5.461325025558471,
      "epoch": 3.2238445555118775,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039870088614268824,
      "loss": 5.1894,
      "mean_token_accuracy": 0.2059697315096855,
      "num_tokens": 64053141.0,
      "step": 28705
    },
    {
      "entropy": 5.312066745758057,
      "epoch": 3.2244061324198348,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039866734761268534,
      "loss": 4.973,
      "mean_token_accuracy": 0.21816044598817824,
      "num_tokens": 64064583.0,
      "step": 28710
    },
    {
      "entropy": 5.492320919036866,
      "epoch": 3.2249677093277924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039863380514506614,
      "loss": 5.1656,
      "mean_token_accuracy": 0.2053701549768448,
      "num_tokens": 64075368.0,
      "step": 28715
    },
    {
      "entropy": 5.4458568572998045,
      "epoch": 3.22552928623575,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039860025874089865,
      "loss": 5.1268,
      "mean_token_accuracy": 0.21108191162347795,
      "num_tokens": 64085815.0,
      "step": 28720
    },
    {
      "entropy": 5.340099620819092,
      "epoch": 3.2260908631437077,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000398566708401251,
      "loss": 5.0216,
      "mean_token_accuracy": 0.21570814549922943,
      "num_tokens": 64096825.0,
      "step": 28725
    },
    {
      "entropy": 5.38615345954895,
      "epoch": 3.226652440051665,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039853315412719144,
      "loss": 5.0676,
      "mean_token_accuracy": 0.21554668098688126,
      "num_tokens": 64108010.0,
      "step": 28730
    },
    {
      "entropy": 5.464564466476441,
      "epoch": 3.2272140169596226,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003984995959197883,
      "loss": 5.1035,
      "mean_token_accuracy": 0.20517665147781372,
      "num_tokens": 64119829.0,
      "step": 28735
    },
    {
      "entropy": 5.363821506500244,
      "epoch": 3.2277755938675803,
      "grad_norm": 1.0,
      "learning_rate": 0.0003984660337801101,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21391381025314332,
      "num_tokens": 64130138.0,
      "step": 28740
    },
    {
      "entropy": 5.395666456222534,
      "epoch": 3.2283371707755375,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039843246770922557,
      "loss": 5.0553,
      "mean_token_accuracy": 0.20505425781011583,
      "num_tokens": 64141651.0,
      "step": 28745
    },
    {
      "entropy": 5.4843505859375,
      "epoch": 3.228898747683495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003983988977082033,
      "loss": 5.1495,
      "mean_token_accuracy": 0.20380108654499055,
      "num_tokens": 64152718.0,
      "step": 28750
    },
    {
      "entropy": 5.403435707092285,
      "epoch": 3.229460324591453,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039836532377811237,
      "loss": 5.1159,
      "mean_token_accuracy": 0.21283408850431443,
      "num_tokens": 64163977.0,
      "step": 28755
    },
    {
      "entropy": 5.350397682189941,
      "epoch": 3.2300219014994105,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003983317459200217,
      "loss": 5.0784,
      "mean_token_accuracy": 0.2126191660761833,
      "num_tokens": 64175298.0,
      "step": 28760
    },
    {
      "entropy": 5.429139471054077,
      "epoch": 3.2305834784073677,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039829816413500035,
      "loss": 5.1069,
      "mean_token_accuracy": 0.2088781028985977,
      "num_tokens": 64186746.0,
      "step": 28765
    },
    {
      "entropy": 5.4399628162384035,
      "epoch": 3.2311450553153254,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003982645784241176,
      "loss": 5.1027,
      "mean_token_accuracy": 0.20521072596311568,
      "num_tokens": 64198682.0,
      "step": 28770
    },
    {
      "entropy": 5.435688638687134,
      "epoch": 3.231706632223283,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003982309887884429,
      "loss": 5.0707,
      "mean_token_accuracy": 0.21290123015642165,
      "num_tokens": 64210425.0,
      "step": 28775
    },
    {
      "entropy": 5.380758810043335,
      "epoch": 3.2322682091312407,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039819739522904576,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21530184000730515,
      "num_tokens": 64222839.0,
      "step": 28780
    },
    {
      "entropy": 5.318582391738891,
      "epoch": 3.232829786039198,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039816379774699573,
      "loss": 5.0559,
      "mean_token_accuracy": 0.21565458178520203,
      "num_tokens": 64234235.0,
      "step": 28785
    },
    {
      "entropy": 5.403061723709106,
      "epoch": 3.2333913629471556,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039813019634336264,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20885741710662842,
      "num_tokens": 64244607.0,
      "step": 28790
    },
    {
      "entropy": 5.417196273803711,
      "epoch": 3.2339529398551132,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003980965910192163,
      "loss": 5.072,
      "mean_token_accuracy": 0.20620408207178115,
      "num_tokens": 64254434.0,
      "step": 28795
    },
    {
      "entropy": 5.323205089569091,
      "epoch": 3.234514516763071,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039806298177562675,
      "loss": 5.0352,
      "mean_token_accuracy": 0.21107323914766313,
      "num_tokens": 64266693.0,
      "step": 28800
    },
    {
      "entropy": 5.386083793640137,
      "epoch": 3.235076093671028,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039802936861366416,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21241853684186934,
      "num_tokens": 64277901.0,
      "step": 28805
    },
    {
      "entropy": 5.4404394149780275,
      "epoch": 3.2356376705789858,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003979957515343988,
      "loss": 5.0954,
      "mean_token_accuracy": 0.21260495036840438,
      "num_tokens": 64288583.0,
      "step": 28810
    },
    {
      "entropy": 5.4045570373535154,
      "epoch": 3.2361992474869434,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000397962130538901,
      "loss": 5.0054,
      "mean_token_accuracy": 0.2141669884324074,
      "num_tokens": 64299688.0,
      "step": 28815
    },
    {
      "entropy": 5.311158037185669,
      "epoch": 3.236760824394901,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039792850562824124,
      "loss": 5.0441,
      "mean_token_accuracy": 0.21075428575277327,
      "num_tokens": 64311150.0,
      "step": 28820
    },
    {
      "entropy": 5.344502210617065,
      "epoch": 3.2373224013028583,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003978948768034902,
      "loss": 5.0567,
      "mean_token_accuracy": 0.21306562423706055,
      "num_tokens": 64322658.0,
      "step": 28825
    },
    {
      "entropy": 5.393837738037109,
      "epoch": 3.237883978210816,
      "grad_norm": 1.0,
      "learning_rate": 0.00039786124406571854,
      "loss": 4.9983,
      "mean_token_accuracy": 0.21311597824096679,
      "num_tokens": 64333350.0,
      "step": 28830
    },
    {
      "entropy": 5.4702075004577635,
      "epoch": 3.2384455551187736,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039782760741599734,
      "loss": 5.1338,
      "mean_token_accuracy": 0.20816671699285508,
      "num_tokens": 64344689.0,
      "step": 28835
    },
    {
      "entropy": 5.328089952468872,
      "epoch": 3.239007132026731,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003977939668553973,
      "loss": 5.0897,
      "mean_token_accuracy": 0.21069820374250411,
      "num_tokens": 64356602.0,
      "step": 28840
    },
    {
      "entropy": 5.378416872024536,
      "epoch": 3.2395687089346885,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039776032238498995,
      "loss": 5.0466,
      "mean_token_accuracy": 0.20884627401828765,
      "num_tokens": 64368496.0,
      "step": 28845
    },
    {
      "entropy": 5.466111707687378,
      "epoch": 3.240130285842646,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003977266740058462,
      "loss": 5.0688,
      "mean_token_accuracy": 0.21359820663928986,
      "num_tokens": 64379374.0,
      "step": 28850
    },
    {
      "entropy": 5.44654483795166,
      "epoch": 3.240691862750604,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003976930217190376,
      "loss": 5.0505,
      "mean_token_accuracy": 0.21293716877698898,
      "num_tokens": 64391006.0,
      "step": 28855
    },
    {
      "entropy": 5.314650249481201,
      "epoch": 3.241253439658561,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003976593655256356,
      "loss": 5.0276,
      "mean_token_accuracy": 0.21335606873035431,
      "num_tokens": 64403100.0,
      "step": 28860
    },
    {
      "entropy": 5.32732048034668,
      "epoch": 3.2418150165665187,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039762570542671176,
      "loss": 4.9796,
      "mean_token_accuracy": 0.21560219228267669,
      "num_tokens": 64413763.0,
      "step": 28865
    },
    {
      "entropy": 5.329283952713013,
      "epoch": 3.2423765934744764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000397592041423338,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21257373541593552,
      "num_tokens": 64425289.0,
      "step": 28870
    },
    {
      "entropy": 5.457923173904419,
      "epoch": 3.242938170382434,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039755837351658605,
      "loss": 5.0291,
      "mean_token_accuracy": 0.21684283316135405,
      "num_tokens": 64436080.0,
      "step": 28875
    },
    {
      "entropy": 5.350285339355469,
      "epoch": 3.2434997472903913,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003975247017075279,
      "loss": 5.0552,
      "mean_token_accuracy": 0.2123725101351738,
      "num_tokens": 64447036.0,
      "step": 28880
    },
    {
      "entropy": 5.468393278121948,
      "epoch": 3.244061324198349,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003974910259972358,
      "loss": 5.2159,
      "mean_token_accuracy": 0.19737545251846314,
      "num_tokens": 64458173.0,
      "step": 28885
    },
    {
      "entropy": 5.464691877365112,
      "epoch": 3.2446229011063066,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003974573463867819,
      "loss": 5.0788,
      "mean_token_accuracy": 0.2061795711517334,
      "num_tokens": 64468985.0,
      "step": 28890
    },
    {
      "entropy": 5.417109537124634,
      "epoch": 3.2451844780142642,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039742366287723855,
      "loss": 5.0808,
      "mean_token_accuracy": 0.21201095283031463,
      "num_tokens": 64479802.0,
      "step": 28895
    },
    {
      "entropy": 5.34335880279541,
      "epoch": 3.2457460549222215,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039738997546967834,
      "loss": 5.0748,
      "mean_token_accuracy": 0.1971924513578415,
      "num_tokens": 64490072.0,
      "step": 28900
    },
    {
      "entropy": 5.380109405517578,
      "epoch": 3.246307631830179,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039735628416517394,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20982010215520858,
      "num_tokens": 64501096.0,
      "step": 28905
    },
    {
      "entropy": 5.418084621429443,
      "epoch": 3.246869208738137,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039732258896479786,
      "loss": 5.0715,
      "mean_token_accuracy": 0.21033720821142196,
      "num_tokens": 64512562.0,
      "step": 28910
    },
    {
      "entropy": 5.2806243896484375,
      "epoch": 3.2474307856460944,
      "grad_norm": 1.0,
      "learning_rate": 0.0003972888898696231,
      "loss": 4.9224,
      "mean_token_accuracy": 0.21237394958734512,
      "num_tokens": 64522955.0,
      "step": 28915
    },
    {
      "entropy": 5.371495962142944,
      "epoch": 3.2479923625540517,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039725518688072276,
      "loss": 5.043,
      "mean_token_accuracy": 0.21253226995468139,
      "num_tokens": 64533805.0,
      "step": 28920
    },
    {
      "entropy": 5.365160512924194,
      "epoch": 3.2485539394620093,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039722147999916976,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20092564225196838,
      "num_tokens": 64545543.0,
      "step": 28925
    },
    {
      "entropy": 5.408982133865356,
      "epoch": 3.249115516369967,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003971877692260375,
      "loss": 5.0964,
      "mean_token_accuracy": 0.20363879203796387,
      "num_tokens": 64557109.0,
      "step": 28930
    },
    {
      "entropy": 5.433042240142822,
      "epoch": 3.249677093277924,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039715405456239924,
      "loss": 5.112,
      "mean_token_accuracy": 0.20656547099351882,
      "num_tokens": 64568880.0,
      "step": 28935
    },
    {
      "entropy": 5.420535469055176,
      "epoch": 3.250238670185882,
      "grad_norm": 1.125,
      "learning_rate": 0.0003971203360093286,
      "loss": 5.0918,
      "mean_token_accuracy": 0.2168443575501442,
      "num_tokens": 64579859.0,
      "step": 28940
    },
    {
      "entropy": 5.3467161655426025,
      "epoch": 3.2508002470938395,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039708661356789896,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21650617867708205,
      "num_tokens": 64589133.0,
      "step": 28945
    },
    {
      "entropy": 5.3822461605072025,
      "epoch": 3.251361824001797,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039705288723918415,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21326380372047424,
      "num_tokens": 64600040.0,
      "step": 28950
    },
    {
      "entropy": 5.293115758895874,
      "epoch": 3.2519234009097544,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003970191570242582,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2207468792796135,
      "num_tokens": 64611756.0,
      "step": 28955
    },
    {
      "entropy": 5.344913291931152,
      "epoch": 3.252484977817712,
      "grad_norm": 1.015625,
      "learning_rate": 0.000396985422924195,
      "loss": 5.126,
      "mean_token_accuracy": 0.2110512912273407,
      "num_tokens": 64622994.0,
      "step": 28960
    },
    {
      "entropy": 5.410715770721436,
      "epoch": 3.2530465547256697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039695168494006857,
      "loss": 4.9516,
      "mean_token_accuracy": 0.21140797734260558,
      "num_tokens": 64633200.0,
      "step": 28965
    },
    {
      "entropy": 5.449553394317627,
      "epoch": 3.2536081316336274,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003969179430729531,
      "loss": 5.156,
      "mean_token_accuracy": 0.20386087745428086,
      "num_tokens": 64644412.0,
      "step": 28970
    },
    {
      "entropy": 5.322927188873291,
      "epoch": 3.2541697085415846,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003968841973239231,
      "loss": 4.9688,
      "mean_token_accuracy": 0.21006208956241607,
      "num_tokens": 64655986.0,
      "step": 28975
    },
    {
      "entropy": 5.206239700317383,
      "epoch": 3.2547312854495423,
      "grad_norm": 1.015625,
      "learning_rate": 0.000396850447694053,
      "loss": 4.9558,
      "mean_token_accuracy": 0.2132605254650116,
      "num_tokens": 64667770.0,
      "step": 28980
    },
    {
      "entropy": 5.496365308761597,
      "epoch": 3.2552928623575,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003968166941844174,
      "loss": 5.1611,
      "mean_token_accuracy": 0.2046256572008133,
      "num_tokens": 64679239.0,
      "step": 28985
    },
    {
      "entropy": 5.399898099899292,
      "epoch": 3.2558544392654576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000396782936796091,
      "loss": 4.9672,
      "mean_token_accuracy": 0.20844251811504363,
      "num_tokens": 64690192.0,
      "step": 28990
    },
    {
      "entropy": 5.345163583755493,
      "epoch": 3.256416016173415,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003967491755301487,
      "loss": 5.028,
      "mean_token_accuracy": 0.21284105479717255,
      "num_tokens": 64700692.0,
      "step": 28995
    },
    {
      "entropy": 5.389095211029053,
      "epoch": 3.2569775930813725,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003967154103876654,
      "loss": 5.1159,
      "mean_token_accuracy": 0.20665353834629058,
      "num_tokens": 64711753.0,
      "step": 29000
    },
    {
      "entropy": 5.479111194610596,
      "epoch": 3.25753916998933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039668164136971615,
      "loss": 5.1864,
      "mean_token_accuracy": 0.1997473120689392,
      "num_tokens": 64722145.0,
      "step": 29005
    },
    {
      "entropy": 5.3459985733032225,
      "epoch": 3.258100746897288,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003966478684773763,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21257347166538237,
      "num_tokens": 64732059.0,
      "step": 29010
    },
    {
      "entropy": 5.412413740158081,
      "epoch": 3.258662323805245,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039661409171172117,
      "loss": 5.0682,
      "mean_token_accuracy": 0.20907711535692214,
      "num_tokens": 64742811.0,
      "step": 29015
    },
    {
      "entropy": 5.31259822845459,
      "epoch": 3.2592239007132027,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039658031107382615,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21967145502567292,
      "num_tokens": 64753287.0,
      "step": 29020
    },
    {
      "entropy": 5.490747308731079,
      "epoch": 3.2597854776211603,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039654652656476695,
      "loss": 5.2655,
      "mean_token_accuracy": 0.19987978488206865,
      "num_tokens": 64767739.0,
      "step": 29025
    },
    {
      "entropy": 5.408587312698364,
      "epoch": 3.2603470545291175,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003965127381856191,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20974887162446976,
      "num_tokens": 64779217.0,
      "step": 29030
    },
    {
      "entropy": 5.423760461807251,
      "epoch": 3.260908631437075,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039647894593745865,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2162157878279686,
      "num_tokens": 64790613.0,
      "step": 29035
    },
    {
      "entropy": 5.289842748641968,
      "epoch": 3.261470208345033,
      "grad_norm": 1.0,
      "learning_rate": 0.00039644514982136136,
      "loss": 4.9675,
      "mean_token_accuracy": 0.2161129593849182,
      "num_tokens": 64801256.0,
      "step": 29040
    },
    {
      "entropy": 5.306471633911133,
      "epoch": 3.2620317852529905,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039641134983840336,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21952452957630159,
      "num_tokens": 64812938.0,
      "step": 29045
    },
    {
      "entropy": 5.417296695709228,
      "epoch": 3.2625933621609478,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000396377545989661,
      "loss": 5.0332,
      "mean_token_accuracy": 0.2108813688158989,
      "num_tokens": 64824041.0,
      "step": 29050
    },
    {
      "entropy": 5.301271629333496,
      "epoch": 3.2631549390689054,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039634373827621047,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21997078657150268,
      "num_tokens": 64833837.0,
      "step": 29055
    },
    {
      "entropy": 5.279598379135132,
      "epoch": 3.263716515976863,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003963099266991282,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21428996473550796,
      "num_tokens": 64843937.0,
      "step": 29060
    },
    {
      "entropy": 5.370728302001953,
      "epoch": 3.2642780928848207,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039627611125949086,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20744584053754805,
      "num_tokens": 64855955.0,
      "step": 29065
    },
    {
      "entropy": 5.482735681533813,
      "epoch": 3.264839669792778,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039624229195837506,
      "loss": 5.1823,
      "mean_token_accuracy": 0.2035892277956009,
      "num_tokens": 64867505.0,
      "step": 29070
    },
    {
      "entropy": 5.344501781463623,
      "epoch": 3.2654012467007356,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039620846879685767,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2207024335861206,
      "num_tokens": 64878746.0,
      "step": 29075
    },
    {
      "entropy": 5.3951225757598875,
      "epoch": 3.2659628236086933,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039617464177601564,
      "loss": 5.1424,
      "mean_token_accuracy": 0.21294564008712769,
      "num_tokens": 64889520.0,
      "step": 29080
    },
    {
      "entropy": 5.42517352104187,
      "epoch": 3.266524400516651,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039614081089692606,
      "loss": 5.1851,
      "mean_token_accuracy": 0.20519082844257355,
      "num_tokens": 64900380.0,
      "step": 29085
    },
    {
      "entropy": 5.424459171295166,
      "epoch": 3.267085977424608,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039610697616066604,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2088787466287613,
      "num_tokens": 64910480.0,
      "step": 29090
    },
    {
      "entropy": 5.335051345825195,
      "epoch": 3.267647554332566,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039607313756831296,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21465830355882645,
      "num_tokens": 64920911.0,
      "step": 29095
    },
    {
      "entropy": 5.342704391479492,
      "epoch": 3.2682091312405235,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039603929512094417,
      "loss": 4.9561,
      "mean_token_accuracy": 0.21397462338209153,
      "num_tokens": 64930687.0,
      "step": 29100
    },
    {
      "entropy": 5.276119613647461,
      "epoch": 3.268770708148481,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003960054488196373,
      "loss": 4.9916,
      "mean_token_accuracy": 0.2167527124285698,
      "num_tokens": 64942143.0,
      "step": 29105
    },
    {
      "entropy": 5.359771251678467,
      "epoch": 3.2693322850564384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039597159866546996,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21299714893102645,
      "num_tokens": 64952440.0,
      "step": 29110
    },
    {
      "entropy": 5.395861577987671,
      "epoch": 3.269893861964396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003959377446595201,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21522184312343598,
      "num_tokens": 64962752.0,
      "step": 29115
    },
    {
      "entropy": 5.3250874996185305,
      "epoch": 3.2704554388723537,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039590388680286543,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2112973153591156,
      "num_tokens": 64975413.0,
      "step": 29120
    },
    {
      "entropy": 5.382338237762451,
      "epoch": 3.271017015780311,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003958700250965842,
      "loss": 5.0318,
      "mean_token_accuracy": 0.2101977914571762,
      "num_tokens": 64987922.0,
      "step": 29125
    },
    {
      "entropy": 5.334604501724243,
      "epoch": 3.2715785926882686,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003958361595417544,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21147289574146272,
      "num_tokens": 64999448.0,
      "step": 29130
    },
    {
      "entropy": 5.43314266204834,
      "epoch": 3.2721401695962262,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003958022901394545,
      "loss": 5.1741,
      "mean_token_accuracy": 0.2018943816423416,
      "num_tokens": 65010771.0,
      "step": 29135
    },
    {
      "entropy": 5.494566822052002,
      "epoch": 3.272701746504184,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003957684168907628,
      "loss": 5.1128,
      "mean_token_accuracy": 0.2082195147871971,
      "num_tokens": 65021368.0,
      "step": 29140
    },
    {
      "entropy": 5.3742814540863035,
      "epoch": 3.273263323412141,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039573453979675793,
      "loss": 4.9721,
      "mean_token_accuracy": 0.22569199353456498,
      "num_tokens": 65032622.0,
      "step": 29145
    },
    {
      "entropy": 5.244518423080445,
      "epoch": 3.2738249003200988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039570065885851844,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2165824219584465,
      "num_tokens": 65043655.0,
      "step": 29150
    },
    {
      "entropy": 5.403975534439087,
      "epoch": 3.2743864772280564,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003956667740771231,
      "loss": 5.1864,
      "mean_token_accuracy": 0.20303287506103515,
      "num_tokens": 65055846.0,
      "step": 29155
    },
    {
      "entropy": 5.52277717590332,
      "epoch": 3.274948054136014,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003956328854536509,
      "loss": 5.1981,
      "mean_token_accuracy": 0.20417805761098862,
      "num_tokens": 65067617.0,
      "step": 29160
    },
    {
      "entropy": 5.5247166633605955,
      "epoch": 3.2755096310439713,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039559899298918094,
      "loss": 5.132,
      "mean_token_accuracy": 0.21998744904994966,
      "num_tokens": 65080111.0,
      "step": 29165
    },
    {
      "entropy": 5.343008804321289,
      "epoch": 3.276071207951929,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003955650966847922,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21785236299037933,
      "num_tokens": 65092284.0,
      "step": 29170
    },
    {
      "entropy": 5.398488855361938,
      "epoch": 3.2766327848598866,
      "grad_norm": 0.96875,
      "learning_rate": 0.000395531196541564,
      "loss": 5.0858,
      "mean_token_accuracy": 0.21134549975395203,
      "num_tokens": 65104133.0,
      "step": 29175
    },
    {
      "entropy": 5.313236665725708,
      "epoch": 3.2771943617678443,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039549729256057583,
      "loss": 4.952,
      "mean_token_accuracy": 0.21240609735250474,
      "num_tokens": 65116065.0,
      "step": 29180
    },
    {
      "entropy": 5.307000541687012,
      "epoch": 3.2777559386758015,
      "grad_norm": 1.046875,
      "learning_rate": 0.000395463384742907,
      "loss": 4.9555,
      "mean_token_accuracy": 0.214469975233078,
      "num_tokens": 65127235.0,
      "step": 29185
    },
    {
      "entropy": 5.434169435501099,
      "epoch": 3.278317515583759,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039542947308963733,
      "loss": 5.1168,
      "mean_token_accuracy": 0.20674622803926468,
      "num_tokens": 65138186.0,
      "step": 29190
    },
    {
      "entropy": 5.363085508346558,
      "epoch": 3.278879092491717,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003953955576018466,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21097055077552795,
      "num_tokens": 65149222.0,
      "step": 29195
    },
    {
      "entropy": 5.361354875564575,
      "epoch": 3.2794406693996745,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003953616382806145,
      "loss": 5.029,
      "mean_token_accuracy": 0.2115889459848404,
      "num_tokens": 65160404.0,
      "step": 29200
    },
    {
      "entropy": 5.3984016418457035,
      "epoch": 3.2800022463076317,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039532771512702124,
      "loss": 5.0335,
      "mean_token_accuracy": 0.20314836055040358,
      "num_tokens": 65171395.0,
      "step": 29205
    },
    {
      "entropy": 5.383121061325073,
      "epoch": 3.2805638232155894,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003952937881421468,
      "loss": 5.0759,
      "mean_token_accuracy": 0.20740336328744888,
      "num_tokens": 65183752.0,
      "step": 29210
    },
    {
      "entropy": 5.315774154663086,
      "epoch": 3.281125400123547,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003952598573270716,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2182427942752838,
      "num_tokens": 65193807.0,
      "step": 29215
    },
    {
      "entropy": 5.362392854690552,
      "epoch": 3.2816869770315042,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003952259226828759,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20512927919626237,
      "num_tokens": 65203482.0,
      "step": 29220
    },
    {
      "entropy": 5.448481464385987,
      "epoch": 3.282248553939462,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039519198421064005,
      "loss": 5.0856,
      "mean_token_accuracy": 0.2069957360625267,
      "num_tokens": 65214965.0,
      "step": 29225
    },
    {
      "entropy": 5.358859729766846,
      "epoch": 3.2828101308474196,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039515804191144496,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2147458016872406,
      "num_tokens": 65227791.0,
      "step": 29230
    },
    {
      "entropy": 5.4664689064025875,
      "epoch": 3.2833717077553772,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039512409578637114,
      "loss": 5.1267,
      "mean_token_accuracy": 0.20669544786214827,
      "num_tokens": 65238937.0,
      "step": 29235
    },
    {
      "entropy": 5.408141374588013,
      "epoch": 3.2839332846633345,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039509014583649947,
      "loss": 5.103,
      "mean_token_accuracy": 0.20498318821191788,
      "num_tokens": 65250912.0,
      "step": 29240
    },
    {
      "entropy": 5.367961788177491,
      "epoch": 3.284494861571292,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039505619206291106,
      "loss": 5.0626,
      "mean_token_accuracy": 0.21673996597528458,
      "num_tokens": 65262094.0,
      "step": 29245
    },
    {
      "entropy": 5.30397744178772,
      "epoch": 3.2850564384792498,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039502223446668687,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21961585581302642,
      "num_tokens": 65271860.0,
      "step": 29250
    },
    {
      "entropy": 5.3818176746368405,
      "epoch": 3.2856180153872074,
      "grad_norm": 1.0,
      "learning_rate": 0.0003949882730489082,
      "loss": 5.1181,
      "mean_token_accuracy": 0.2029322698712349,
      "num_tokens": 65282341.0,
      "step": 29255
    },
    {
      "entropy": 5.424140453338623,
      "epoch": 3.2861795922951647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039495430781065645,
      "loss": 5.1013,
      "mean_token_accuracy": 0.21312126219272615,
      "num_tokens": 65293570.0,
      "step": 29260
    },
    {
      "entropy": 5.351343393325806,
      "epoch": 3.2867411692031223,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000394920338753013,
      "loss": 4.994,
      "mean_token_accuracy": 0.21503273993730546,
      "num_tokens": 65305226.0,
      "step": 29265
    },
    {
      "entropy": 5.298500680923462,
      "epoch": 3.28730274611108,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003948863658770594,
      "loss": 4.9322,
      "mean_token_accuracy": 0.2155624359846115,
      "num_tokens": 65316309.0,
      "step": 29270
    },
    {
      "entropy": 5.355752992630005,
      "epoch": 3.2878643230190376,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003948523891838774,
      "loss": 4.9908,
      "mean_token_accuracy": 0.22021102607250215,
      "num_tokens": 65327491.0,
      "step": 29275
    },
    {
      "entropy": 5.259730052947998,
      "epoch": 3.288425899926995,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039481840867454886,
      "loss": 4.8806,
      "mean_token_accuracy": 0.21837422102689744,
      "num_tokens": 65337908.0,
      "step": 29280
    },
    {
      "entropy": 5.233044910430908,
      "epoch": 3.2889874768349525,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003947844243501558,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22128734290599822,
      "num_tokens": 65348209.0,
      "step": 29285
    },
    {
      "entropy": 5.3591721534729,
      "epoch": 3.28954905374291,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039475043621178005,
      "loss": 5.0248,
      "mean_token_accuracy": 0.21404968798160554,
      "num_tokens": 65358648.0,
      "step": 29290
    },
    {
      "entropy": 5.30058856010437,
      "epoch": 3.290110630650868,
      "grad_norm": 1.0,
      "learning_rate": 0.0003947164442605041,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21017475575208663,
      "num_tokens": 65369132.0,
      "step": 29295
    },
    {
      "entropy": 5.379864454269409,
      "epoch": 3.290672207558825,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003946824484974101,
      "loss": 5.0486,
      "mean_token_accuracy": 0.2145371124148369,
      "num_tokens": 65380334.0,
      "step": 29300
    },
    {
      "entropy": 5.544144964218139,
      "epoch": 3.2912337844667827,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003946484489235805,
      "loss": 5.1954,
      "mean_token_accuracy": 0.19793498069047927,
      "num_tokens": 65394680.0,
      "step": 29305
    },
    {
      "entropy": 5.455538415908814,
      "epoch": 3.2917953613747404,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039461444554009787,
      "loss": 5.0781,
      "mean_token_accuracy": 0.2094290778040886,
      "num_tokens": 65406029.0,
      "step": 29310
    },
    {
      "entropy": 5.338117504119873,
      "epoch": 3.2923569382826976,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000394580438348045,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2139800027012825,
      "num_tokens": 65416626.0,
      "step": 29315
    },
    {
      "entropy": 5.341868162155151,
      "epoch": 3.2929185151906553,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039454642734850447,
      "loss": 5.0424,
      "mean_token_accuracy": 0.21320027261972427,
      "num_tokens": 65427579.0,
      "step": 29320
    },
    {
      "entropy": 5.31810188293457,
      "epoch": 3.293480092098613,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003945124125425594,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20637853890657426,
      "num_tokens": 65439893.0,
      "step": 29325
    },
    {
      "entropy": 5.51945915222168,
      "epoch": 3.2940416690065706,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003944783939312928,
      "loss": 5.1776,
      "mean_token_accuracy": 0.20611874014139175,
      "num_tokens": 65451861.0,
      "step": 29330
    },
    {
      "entropy": 5.3495063304901125,
      "epoch": 3.294603245914528,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003944443715157877,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21031406074762343,
      "num_tokens": 65463764.0,
      "step": 29335
    },
    {
      "entropy": 5.262649583816528,
      "epoch": 3.2951648228224855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003944103452971275,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2028496354818344,
      "num_tokens": 65474367.0,
      "step": 29340
    },
    {
      "entropy": 5.342995119094849,
      "epoch": 3.295726399730443,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039437631527639553,
      "loss": 4.9679,
      "mean_token_accuracy": 0.22034140527248383,
      "num_tokens": 65485347.0,
      "step": 29345
    },
    {
      "entropy": 5.369885730743408,
      "epoch": 3.296287976638401,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039434228145467554,
      "loss": 5.08,
      "mean_token_accuracy": 0.2147718295454979,
      "num_tokens": 65495858.0,
      "step": 29350
    },
    {
      "entropy": 5.312769889831543,
      "epoch": 3.296849553546358,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000394308243833051,
      "loss": 4.8785,
      "mean_token_accuracy": 0.21929091811180115,
      "num_tokens": 65505649.0,
      "step": 29355
    },
    {
      "entropy": 5.380861520767212,
      "epoch": 3.2974111304543157,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00039427420241260557,
      "loss": 5.0317,
      "mean_token_accuracy": 0.21502070277929305,
      "num_tokens": 65516437.0,
      "step": 29360
    },
    {
      "entropy": 5.263468408584595,
      "epoch": 3.2979727073622733,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039424015719442335,
      "loss": 4.976,
      "mean_token_accuracy": 0.21599622070789337,
      "num_tokens": 65528854.0,
      "step": 29365
    },
    {
      "entropy": 5.414577627182007,
      "epoch": 3.298534284270231,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039420610817958836,
      "loss": 5.0887,
      "mean_token_accuracy": 0.2108323946595192,
      "num_tokens": 65539459.0,
      "step": 29370
    },
    {
      "entropy": 5.422315311431885,
      "epoch": 3.299095861178188,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003941720553691846,
      "loss": 5.0693,
      "mean_token_accuracy": 0.2063926413655281,
      "num_tokens": 65550818.0,
      "step": 29375
    },
    {
      "entropy": 5.400266027450561,
      "epoch": 3.299657438086146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003941379987642963,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2134595111012459,
      "num_tokens": 65562572.0,
      "step": 29380
    },
    {
      "entropy": 5.342912673950195,
      "epoch": 3.3002190149941035,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000394103938366008,
      "loss": 5.0678,
      "mean_token_accuracy": 0.20934785455465316,
      "num_tokens": 65573764.0,
      "step": 29385
    },
    {
      "entropy": 5.419212341308594,
      "epoch": 3.300780591902061,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003940698741754041,
      "loss": 5.0737,
      "mean_token_accuracy": 0.21681116968393327,
      "num_tokens": 65585628.0,
      "step": 29390
    },
    {
      "entropy": 5.362552690505981,
      "epoch": 3.3013421688100184,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003940358061935692,
      "loss": 5.029,
      "mean_token_accuracy": 0.2096794381737709,
      "num_tokens": 65596142.0,
      "step": 29395
    },
    {
      "entropy": 5.479443979263306,
      "epoch": 3.301903745717976,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003940017344215881,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19975974410772324,
      "num_tokens": 65607824.0,
      "step": 29400
    },
    {
      "entropy": 5.404713773727417,
      "epoch": 3.3024653226259337,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039396765886054555,
      "loss": 5.0402,
      "mean_token_accuracy": 0.2128315433859825,
      "num_tokens": 65617862.0,
      "step": 29405
    },
    {
      "entropy": 5.368212127685547,
      "epoch": 3.303026899533891,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039393357951152667,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21723357439041138,
      "num_tokens": 65629224.0,
      "step": 29410
    },
    {
      "entropy": 5.2966619491577145,
      "epoch": 3.3035884764418486,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039389949637561646,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2177306666970253,
      "num_tokens": 65640419.0,
      "step": 29415
    },
    {
      "entropy": 5.403443813323975,
      "epoch": 3.3041500533498063,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039386540945390017,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20719340145587922,
      "num_tokens": 65651305.0,
      "step": 29420
    },
    {
      "entropy": 5.4070594787597654,
      "epoch": 3.304711630257764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003938313187474631,
      "loss": 5.1145,
      "mean_token_accuracy": 0.2022452473640442,
      "num_tokens": 65662702.0,
      "step": 29425
    },
    {
      "entropy": 5.375760412216186,
      "epoch": 3.3052732071657216,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003937972242573908,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21411430537700654,
      "num_tokens": 65673981.0,
      "step": 29430
    },
    {
      "entropy": 5.382529067993164,
      "epoch": 3.305834784073679,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003937631259847688,
      "loss": 5.0434,
      "mean_token_accuracy": 0.21629259437322618,
      "num_tokens": 65684633.0,
      "step": 29435
    },
    {
      "entropy": 5.354738521575928,
      "epoch": 3.3063963609816365,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003937290239306828,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20403498858213426,
      "num_tokens": 65696890.0,
      "step": 29440
    },
    {
      "entropy": 5.350843095779419,
      "epoch": 3.306957937889594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039369491809621867,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21572987884283065,
      "num_tokens": 65708620.0,
      "step": 29445
    },
    {
      "entropy": 5.437977123260498,
      "epoch": 3.3075195147975514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003936608084824623,
      "loss": 5.1058,
      "mean_token_accuracy": 0.21223676949739456,
      "num_tokens": 65720182.0,
      "step": 29450
    },
    {
      "entropy": 5.4377936840057375,
      "epoch": 3.308081091705509,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039362669509049975,
      "loss": 5.134,
      "mean_token_accuracy": 0.20627740621566773,
      "num_tokens": 65731376.0,
      "step": 29455
    },
    {
      "entropy": 5.432421445846558,
      "epoch": 3.3086426686134667,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039359257792141717,
      "loss": 5.0492,
      "mean_token_accuracy": 0.2125987723469734,
      "num_tokens": 65743303.0,
      "step": 29460
    },
    {
      "entropy": 5.497948026657104,
      "epoch": 3.3092042455214243,
      "grad_norm": 0.953125,
      "learning_rate": 0.000393558456976301,
      "loss": 5.2016,
      "mean_token_accuracy": 0.19800219386816026,
      "num_tokens": 65755486.0,
      "step": 29465
    },
    {
      "entropy": 5.337993860244751,
      "epoch": 3.3097658224293816,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003935243322562375,
      "loss": 5.009,
      "mean_token_accuracy": 0.20839072167873382,
      "num_tokens": 65766855.0,
      "step": 29470
    },
    {
      "entropy": 5.334459447860718,
      "epoch": 3.310327399337339,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003934902037623133,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21309670209884643,
      "num_tokens": 65778077.0,
      "step": 29475
    },
    {
      "entropy": 5.336214780807495,
      "epoch": 3.310888976245297,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00039345607149561513,
      "loss": 5.0593,
      "mean_token_accuracy": 0.209893162548542,
      "num_tokens": 65791051.0,
      "step": 29480
    },
    {
      "entropy": 5.476957082748413,
      "epoch": 3.3114505531532545,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039342193545722966,
      "loss": 5.154,
      "mean_token_accuracy": 0.2069030299782753,
      "num_tokens": 65802769.0,
      "step": 29485
    },
    {
      "entropy": 5.410486555099487,
      "epoch": 3.3120121300612118,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003933877956482438,
      "loss": 5.0328,
      "mean_token_accuracy": 0.21095822751522064,
      "num_tokens": 65813602.0,
      "step": 29490
    },
    {
      "entropy": 5.38528790473938,
      "epoch": 3.3125737069691694,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039335365206974465,
      "loss": 5.1227,
      "mean_token_accuracy": 0.20879632979631424,
      "num_tokens": 65826675.0,
      "step": 29495
    },
    {
      "entropy": 5.34553108215332,
      "epoch": 3.313135283877127,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003933195047228193,
      "loss": 4.9816,
      "mean_token_accuracy": 0.21879787892103195,
      "num_tokens": 65837643.0,
      "step": 29500
    },
    {
      "entropy": 5.445144462585449,
      "epoch": 3.3136968607850843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039328535360855503,
      "loss": 5.0611,
      "mean_token_accuracy": 0.21754615902900695,
      "num_tokens": 65850184.0,
      "step": 29505
    },
    {
      "entropy": 5.392917537689209,
      "epoch": 3.314258437693042,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039325119872803924,
      "loss": 5.0553,
      "mean_token_accuracy": 0.2146216571331024,
      "num_tokens": 65861178.0,
      "step": 29510
    },
    {
      "entropy": 5.317704820632935,
      "epoch": 3.3148200146009996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039321704008235943,
      "loss": 4.962,
      "mean_token_accuracy": 0.22073178440332414,
      "num_tokens": 65872458.0,
      "step": 29515
    },
    {
      "entropy": 5.304272556304932,
      "epoch": 3.3153815915089573,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039318287767260316,
      "loss": 4.976,
      "mean_token_accuracy": 0.2166664183139801,
      "num_tokens": 65883152.0,
      "step": 29520
    },
    {
      "entropy": 5.337661695480347,
      "epoch": 3.315943168416915,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039314871149985834,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21779426336288452,
      "num_tokens": 65894220.0,
      "step": 29525
    },
    {
      "entropy": 5.359192514419556,
      "epoch": 3.316504745324872,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003931145415652126,
      "loss": 5.0975,
      "mean_token_accuracy": 0.20933461785316468,
      "num_tokens": 65904159.0,
      "step": 29530
    },
    {
      "entropy": 5.480965900421142,
      "epoch": 3.31706632223283,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039308036786975416,
      "loss": 5.1313,
      "mean_token_accuracy": 0.20319838374853133,
      "num_tokens": 65915599.0,
      "step": 29535
    },
    {
      "entropy": 5.376408815383911,
      "epoch": 3.3176278991407875,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003930461904145709,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21272777020931244,
      "num_tokens": 65926518.0,
      "step": 29540
    },
    {
      "entropy": 5.368619918823242,
      "epoch": 3.3181894760487447,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003930120092007512,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20876848846673965,
      "num_tokens": 65938901.0,
      "step": 29545
    },
    {
      "entropy": 5.3764056205749515,
      "epoch": 3.3187510529567024,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003929778242293834,
      "loss": 5.0258,
      "mean_token_accuracy": 0.21177875399589538,
      "num_tokens": 65949798.0,
      "step": 29550
    },
    {
      "entropy": 5.356381464004516,
      "epoch": 3.31931262986466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003929436355015558,
      "loss": 4.9705,
      "mean_token_accuracy": 0.21998427659273148,
      "num_tokens": 65961222.0,
      "step": 29555
    },
    {
      "entropy": 5.429873466491699,
      "epoch": 3.3198742067726177,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039290944301835723,
      "loss": 5.1286,
      "mean_token_accuracy": 0.2014615163207054,
      "num_tokens": 65972373.0,
      "step": 29560
    },
    {
      "entropy": 5.339236211776734,
      "epoch": 3.320435783680575,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039287524678087617,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21630117744207383,
      "num_tokens": 65983716.0,
      "step": 29565
    },
    {
      "entropy": 5.4719651222229,
      "epoch": 3.3209973605885326,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039284104679020154,
      "loss": 5.1383,
      "mean_token_accuracy": 0.21011116355657578,
      "num_tokens": 65994827.0,
      "step": 29570
    },
    {
      "entropy": 5.389466142654419,
      "epoch": 3.3215589374964902,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003928068430474224,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21971674561500548,
      "num_tokens": 66006249.0,
      "step": 29575
    },
    {
      "entropy": 5.35979061126709,
      "epoch": 3.322120514404448,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003927726355536275,
      "loss": 5.0348,
      "mean_token_accuracy": 0.2091402754187584,
      "num_tokens": 66017175.0,
      "step": 29580
    },
    {
      "entropy": 5.30323600769043,
      "epoch": 3.322682091312405,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00039273842430990634,
      "loss": 4.9593,
      "mean_token_accuracy": 0.2178204596042633,
      "num_tokens": 66028518.0,
      "step": 29585
    },
    {
      "entropy": 5.383296775817871,
      "epoch": 3.3232436682203628,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003927042093173481,
      "loss": 5.0304,
      "mean_token_accuracy": 0.2076626256108284,
      "num_tokens": 66038927.0,
      "step": 29590
    },
    {
      "entropy": 5.431569337844849,
      "epoch": 3.3238052451283204,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003926699905770421,
      "loss": 5.0261,
      "mean_token_accuracy": 0.21320104449987412,
      "num_tokens": 66050649.0,
      "step": 29595
    },
    {
      "entropy": 5.394930124282837,
      "epoch": 3.3243668220362776,
      "grad_norm": 1.015625,
      "learning_rate": 0.000392635768090078,
      "loss": 5.0638,
      "mean_token_accuracy": 0.21024627536535262,
      "num_tokens": 66061585.0,
      "step": 29600
    },
    {
      "entropy": 5.349187803268433,
      "epoch": 3.3249283989442353,
      "grad_norm": 1.0,
      "learning_rate": 0.00039260154185754545,
      "loss": 5.0118,
      "mean_token_accuracy": 0.2145455449819565,
      "num_tokens": 66071312.0,
      "step": 29605
    },
    {
      "entropy": 5.3404584407806395,
      "epoch": 3.325489975852193,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003925673118805341,
      "loss": 5.0582,
      "mean_token_accuracy": 0.21381541341543198,
      "num_tokens": 66082643.0,
      "step": 29610
    },
    {
      "entropy": 5.325112581253052,
      "epoch": 3.3260515527601506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000392533078160134,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21535566151142121,
      "num_tokens": 66094094.0,
      "step": 29615
    },
    {
      "entropy": 5.364333391189575,
      "epoch": 3.3266131296681083,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039249884069743514,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21410958617925643,
      "num_tokens": 66105365.0,
      "step": 29620
    },
    {
      "entropy": 5.405356931686401,
      "epoch": 3.3271747065760655,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003924645994935276,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20570087432861328,
      "num_tokens": 66115440.0,
      "step": 29625
    },
    {
      "entropy": 5.368456506729126,
      "epoch": 3.327736283484023,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003924303545495017,
      "loss": 5.0341,
      "mean_token_accuracy": 0.20177303850650788,
      "num_tokens": 66127213.0,
      "step": 29630
    },
    {
      "entropy": 5.370482540130615,
      "epoch": 3.328297860391981,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039239610586644765,
      "loss": 4.942,
      "mean_token_accuracy": 0.2156612604856491,
      "num_tokens": 66138016.0,
      "step": 29635
    },
    {
      "entropy": 5.267062330245972,
      "epoch": 3.328859437299938,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003923618534454561,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2184280574321747,
      "num_tokens": 66149981.0,
      "step": 29640
    },
    {
      "entropy": 5.349187707901001,
      "epoch": 3.3294210142078957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003923275972876176,
      "loss": 5.052,
      "mean_token_accuracy": 0.21020550578832625,
      "num_tokens": 66161609.0,
      "step": 29645
    },
    {
      "entropy": 5.301196956634522,
      "epoch": 3.3299825911158534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039229333739402303,
      "loss": 4.946,
      "mean_token_accuracy": 0.21668857485055923,
      "num_tokens": 66171665.0,
      "step": 29650
    },
    {
      "entropy": 5.288922214508057,
      "epoch": 3.330544168023811,
      "grad_norm": 1.0,
      "learning_rate": 0.000392259073765763,
      "loss": 4.9519,
      "mean_token_accuracy": 0.21599392741918563,
      "num_tokens": 66181773.0,
      "step": 29655
    },
    {
      "entropy": 5.393890237808227,
      "epoch": 3.3311057449317683,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039222480640392854,
      "loss": 5.0798,
      "mean_token_accuracy": 0.2148784413933754,
      "num_tokens": 66193578.0,
      "step": 29660
    },
    {
      "entropy": 5.425887632369995,
      "epoch": 3.331667321839726,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039219053530961087,
      "loss": 5.0878,
      "mean_token_accuracy": 0.20491078943014146,
      "num_tokens": 66205185.0,
      "step": 29665
    },
    {
      "entropy": 5.278323173522949,
      "epoch": 3.3322288987476836,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039215626048390105,
      "loss": 4.9796,
      "mean_token_accuracy": 0.21064774692058563,
      "num_tokens": 66217143.0,
      "step": 29670
    },
    {
      "entropy": 5.363730764389038,
      "epoch": 3.3327904756556412,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003921219819278905,
      "loss": 5.0252,
      "mean_token_accuracy": 0.22076896131038665,
      "num_tokens": 66229022.0,
      "step": 29675
    },
    {
      "entropy": 5.437602996826172,
      "epoch": 3.3333520525635985,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003920876996426706,
      "loss": 5.0982,
      "mean_token_accuracy": 0.20251217484474182,
      "num_tokens": 66240028.0,
      "step": 29680
    },
    {
      "entropy": 5.388151597976685,
      "epoch": 3.333913629471556,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039205341362933294,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2157108649611473,
      "num_tokens": 66251656.0,
      "step": 29685
    },
    {
      "entropy": 5.310275268554688,
      "epoch": 3.334475206379514,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003920191238889691,
      "loss": 4.9984,
      "mean_token_accuracy": 0.22030124217271804,
      "num_tokens": 66262644.0,
      "step": 29690
    },
    {
      "entropy": 5.254477262496948,
      "epoch": 3.335036783287471,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003919848304226711,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21373129040002822,
      "num_tokens": 66273374.0,
      "step": 29695
    },
    {
      "entropy": 5.359227228164673,
      "epoch": 3.3355983601954287,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003919505332315307,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21394117325544357,
      "num_tokens": 66284490.0,
      "step": 29700
    },
    {
      "entropy": 5.431166744232177,
      "epoch": 3.3361599371033863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039191623231663985,
      "loss": 5.0211,
      "mean_token_accuracy": 0.2123098075389862,
      "num_tokens": 66295555.0,
      "step": 29705
    },
    {
      "entropy": 5.356332015991211,
      "epoch": 3.336721514011344,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000391881927679091,
      "loss": 5.0039,
      "mean_token_accuracy": 0.2127162903547287,
      "num_tokens": 66307234.0,
      "step": 29710
    },
    {
      "entropy": 5.436452198028564,
      "epoch": 3.3372830909193016,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039184761931997607,
      "loss": 5.1458,
      "mean_token_accuracy": 0.20040260553359984,
      "num_tokens": 66319239.0,
      "step": 29715
    },
    {
      "entropy": 5.483844757080078,
      "epoch": 3.337844667827259,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003918133072403878,
      "loss": 5.1017,
      "mean_token_accuracy": 0.2131364345550537,
      "num_tokens": 66332260.0,
      "step": 29720
    },
    {
      "entropy": 5.3978784561157225,
      "epoch": 3.3384062447352165,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039177899144141835,
      "loss": 5.0841,
      "mean_token_accuracy": 0.20385850220918655,
      "num_tokens": 66343817.0,
      "step": 29725
    },
    {
      "entropy": 5.420223617553711,
      "epoch": 3.338967821643174,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003917446719241606,
      "loss": 5.0981,
      "mean_token_accuracy": 0.20456845611333846,
      "num_tokens": 66355724.0,
      "step": 29730
    },
    {
      "entropy": 5.356211853027344,
      "epoch": 3.3395293985511314,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039171034868970724,
      "loss": 4.974,
      "mean_token_accuracy": 0.2145402744412422,
      "num_tokens": 66365994.0,
      "step": 29735
    },
    {
      "entropy": 5.382691478729248,
      "epoch": 3.340090975459089,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039167602173915104,
      "loss": 5.1694,
      "mean_token_accuracy": 0.20327653586864472,
      "num_tokens": 66377929.0,
      "step": 29740
    },
    {
      "entropy": 5.348143863677978,
      "epoch": 3.3406525523670467,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003916416910735851,
      "loss": 4.9936,
      "mean_token_accuracy": 0.21764630526304246,
      "num_tokens": 66389292.0,
      "step": 29745
    },
    {
      "entropy": 5.369740390777588,
      "epoch": 3.3412141292750044,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039160735669410246,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21389829814434053,
      "num_tokens": 66399266.0,
      "step": 29750
    },
    {
      "entropy": 5.329080724716187,
      "epoch": 3.3417757061829616,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003915730186017964,
      "loss": 5.0575,
      "mean_token_accuracy": 0.2133892998099327,
      "num_tokens": 66410718.0,
      "step": 29755
    },
    {
      "entropy": 5.360497522354126,
      "epoch": 3.3423372830909193,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039153867679776023,
      "loss": 5.0791,
      "mean_token_accuracy": 0.21272919178009034,
      "num_tokens": 66420982.0,
      "step": 29760
    },
    {
      "entropy": 5.40855131149292,
      "epoch": 3.342898859998877,
      "grad_norm": 1.0,
      "learning_rate": 0.0003915043312830873,
      "loss": 5.0388,
      "mean_token_accuracy": 0.21751929670572281,
      "num_tokens": 66431378.0,
      "step": 29765
    },
    {
      "entropy": 5.395191240310669,
      "epoch": 3.3434604369068346,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003914699820588714,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21691346019506455,
      "num_tokens": 66442033.0,
      "step": 29770
    },
    {
      "entropy": 5.3135298728942875,
      "epoch": 3.344022013814792,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039143562912620603,
      "loss": 4.9747,
      "mean_token_accuracy": 0.2102789208292961,
      "num_tokens": 66453486.0,
      "step": 29775
    },
    {
      "entropy": 5.429760837554932,
      "epoch": 3.3445835907227495,
      "grad_norm": 1.046875,
      "learning_rate": 0.000391401272486185,
      "loss": 5.1359,
      "mean_token_accuracy": 0.20395235419273378,
      "num_tokens": 66464931.0,
      "step": 29780
    },
    {
      "entropy": 5.588534021377564,
      "epoch": 3.345145167630707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003913669121399024,
      "loss": 5.2292,
      "mean_token_accuracy": 0.202242611348629,
      "num_tokens": 66477187.0,
      "step": 29785
    },
    {
      "entropy": 5.388653993606567,
      "epoch": 3.3457067445386643,
      "grad_norm": 1.0,
      "learning_rate": 0.00039133254808845216,
      "loss": 5.0879,
      "mean_token_accuracy": 0.2126736968755722,
      "num_tokens": 66488777.0,
      "step": 29790
    },
    {
      "entropy": 5.340397834777832,
      "epoch": 3.346268321446622,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00039129818033292843,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2150785133242607,
      "num_tokens": 66499576.0,
      "step": 29795
    },
    {
      "entropy": 5.285248422622681,
      "epoch": 3.3468298983545797,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003912638088744256,
      "loss": 4.8554,
      "mean_token_accuracy": 0.23262681514024736,
      "num_tokens": 66510824.0,
      "step": 29800
    },
    {
      "entropy": 5.396010255813598,
      "epoch": 3.3473914752625373,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000391229433714038,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20793729424476623,
      "num_tokens": 66523151.0,
      "step": 29805
    },
    {
      "entropy": 5.309005928039551,
      "epoch": 3.347953052170495,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003911950548528601,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2154133915901184,
      "num_tokens": 66533921.0,
      "step": 29810
    },
    {
      "entropy": 5.365468692779541,
      "epoch": 3.348514629078452,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003911606722919866,
      "loss": 4.9943,
      "mean_token_accuracy": 0.2160593017935753,
      "num_tokens": 66545867.0,
      "step": 29815
    },
    {
      "entropy": 5.344428300857544,
      "epoch": 3.34907620598641,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003911262860325123,
      "loss": 5.013,
      "mean_token_accuracy": 0.2179664358496666,
      "num_tokens": 66557522.0,
      "step": 29820
    },
    {
      "entropy": 5.318701076507568,
      "epoch": 3.3496377828943675,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003910918960755319,
      "loss": 5.0202,
      "mean_token_accuracy": 0.2201538309454918,
      "num_tokens": 66568792.0,
      "step": 29825
    },
    {
      "entropy": 5.409763526916504,
      "epoch": 3.3501993598023248,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003910575024221406,
      "loss": 5.1162,
      "mean_token_accuracy": 0.21802058219909667,
      "num_tokens": 66580231.0,
      "step": 29830
    },
    {
      "entropy": 5.298758172988892,
      "epoch": 3.3507609367102824,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039102310507343335,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21818821728229523,
      "num_tokens": 66591171.0,
      "step": 29835
    },
    {
      "entropy": 5.358959007263183,
      "epoch": 3.35132251361824,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003909887040305055,
      "loss": 5.069,
      "mean_token_accuracy": 0.2039777919650078,
      "num_tokens": 66602676.0,
      "step": 29840
    },
    {
      "entropy": 5.380969619750976,
      "epoch": 3.3518840905261977,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003909542992944523,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21503109037876128,
      "num_tokens": 66613526.0,
      "step": 29845
    },
    {
      "entropy": 5.320434284210205,
      "epoch": 3.352445667434155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003909198908663692,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21434350311756134,
      "num_tokens": 66624279.0,
      "step": 29850
    },
    {
      "entropy": 5.393058919906617,
      "epoch": 3.3530072443421126,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039088547874735187,
      "loss": 4.9823,
      "mean_token_accuracy": 0.21619857102632523,
      "num_tokens": 66636908.0,
      "step": 29855
    },
    {
      "entropy": 5.496581268310547,
      "epoch": 3.3535688212500703,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039085106293849594,
      "loss": 5.1794,
      "mean_token_accuracy": 0.2016322582960129,
      "num_tokens": 66648545.0,
      "step": 29860
    },
    {
      "entropy": 5.360687732696533,
      "epoch": 3.354130398158028,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039081664344089724,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21058827489614487,
      "num_tokens": 66659245.0,
      "step": 29865
    },
    {
      "entropy": 5.267288637161255,
      "epoch": 3.354691975065985,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003907822202556516,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21682801991701126,
      "num_tokens": 66670866.0,
      "step": 29870
    },
    {
      "entropy": 5.393119239807129,
      "epoch": 3.355253551973943,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003907477933838553,
      "loss": 5.0848,
      "mean_token_accuracy": 0.20990979224443435,
      "num_tokens": 66681970.0,
      "step": 29875
    },
    {
      "entropy": 5.51213264465332,
      "epoch": 3.3558151288819005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039071336282660437,
      "loss": 5.1171,
      "mean_token_accuracy": 0.21335382014513016,
      "num_tokens": 66693252.0,
      "step": 29880
    },
    {
      "entropy": 5.309964609146118,
      "epoch": 3.3563767057898577,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039067892858499496,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21142490804195405,
      "num_tokens": 66703821.0,
      "step": 29885
    },
    {
      "entropy": 5.274274635314941,
      "epoch": 3.3569382826978154,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003906444906601237,
      "loss": 4.9702,
      "mean_token_accuracy": 0.21837663650512695,
      "num_tokens": 66715945.0,
      "step": 29890
    },
    {
      "entropy": 5.4097565650939945,
      "epoch": 3.357499859605773,
      "grad_norm": 0.90625,
      "learning_rate": 0.000390610049053087,
      "loss": 5.0672,
      "mean_token_accuracy": 0.2133797064423561,
      "num_tokens": 66727551.0,
      "step": 29895
    },
    {
      "entropy": 5.409413003921509,
      "epoch": 3.3580614365137307,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039057560376498147,
      "loss": 5.0402,
      "mean_token_accuracy": 0.21121651977300643,
      "num_tokens": 66738257.0,
      "step": 29900
    },
    {
      "entropy": 5.470018672943115,
      "epoch": 3.3586230134216883,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003905411547969038,
      "loss": 5.1566,
      "mean_token_accuracy": 0.19749877005815505,
      "num_tokens": 66750305.0,
      "step": 29905
    },
    {
      "entropy": 5.420934534072876,
      "epoch": 3.3591845903296456,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003905067021499511,
      "loss": 5.0431,
      "mean_token_accuracy": 0.2064980983734131,
      "num_tokens": 66761659.0,
      "step": 29910
    },
    {
      "entropy": 5.350599098205566,
      "epoch": 3.359746167237603,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039047224582522015,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22108001857995987,
      "num_tokens": 66771981.0,
      "step": 29915
    },
    {
      "entropy": 5.375402212142944,
      "epoch": 3.360307744145561,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003904377858238081,
      "loss": 5.1508,
      "mean_token_accuracy": 0.203959684073925,
      "num_tokens": 66783915.0,
      "step": 29920
    },
    {
      "entropy": 5.365793514251709,
      "epoch": 3.360869321053518,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003904033221468122,
      "loss": 5.0221,
      "mean_token_accuracy": 0.21170813590288162,
      "num_tokens": 66795691.0,
      "step": 29925
    },
    {
      "entropy": 5.390090227127075,
      "epoch": 3.3614308979614758,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003903688547953297,
      "loss": 5.0842,
      "mean_token_accuracy": 0.20903043448925018,
      "num_tokens": 66805792.0,
      "step": 29930
    },
    {
      "entropy": 5.339382314682007,
      "epoch": 3.3619924748694334,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003903343837704582,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21827088743448259,
      "num_tokens": 66817657.0,
      "step": 29935
    },
    {
      "entropy": 5.319599246978759,
      "epoch": 3.362554051777391,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003902999090732951,
      "loss": 4.9062,
      "mean_token_accuracy": 0.22163426131010056,
      "num_tokens": 66828593.0,
      "step": 29940
    },
    {
      "entropy": 5.329289340972901,
      "epoch": 3.3631156286853483,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039026543070493823,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21971505433320998,
      "num_tokens": 66839405.0,
      "step": 29945
    },
    {
      "entropy": 5.2940223693847654,
      "epoch": 3.363677205593306,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003902309486664854,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2110673114657402,
      "num_tokens": 66849938.0,
      "step": 29950
    },
    {
      "entropy": 5.369170475006103,
      "epoch": 3.3642387825012636,
      "grad_norm": 1.0,
      "learning_rate": 0.0003901964629590344,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21327344924211503,
      "num_tokens": 66861105.0,
      "step": 29955
    },
    {
      "entropy": 5.342259407043457,
      "epoch": 3.3648003594092213,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003901619735836833,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21162937879562377,
      "num_tokens": 66873065.0,
      "step": 29960
    },
    {
      "entropy": 5.504744815826416,
      "epoch": 3.3653619363171785,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039012748054153044,
      "loss": 5.1999,
      "mean_token_accuracy": 0.20130303353071213,
      "num_tokens": 66885639.0,
      "step": 29965
    },
    {
      "entropy": 5.372947788238525,
      "epoch": 3.365923513225136,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003900929838336738,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20559097379446029,
      "num_tokens": 66896576.0,
      "step": 29970
    },
    {
      "entropy": 5.3655290603637695,
      "epoch": 3.366485090133094,
      "grad_norm": 1.0,
      "learning_rate": 0.0003900584834612119,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21405138969421386,
      "num_tokens": 66907595.0,
      "step": 29975
    },
    {
      "entropy": 5.430545473098755,
      "epoch": 3.367046667041051,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003900239794252434,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20915476530790328,
      "num_tokens": 66918231.0,
      "step": 29980
    },
    {
      "entropy": 5.434838676452637,
      "epoch": 3.3676082439490087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003899894717268667,
      "loss": 5.2198,
      "mean_token_accuracy": 0.19938337355852126,
      "num_tokens": 66930929.0,
      "step": 29985
    },
    {
      "entropy": 5.395609903335571,
      "epoch": 3.3681698208569664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003899549603671807,
      "loss": 5.0639,
      "mean_token_accuracy": 0.21238719224929808,
      "num_tokens": 66941714.0,
      "step": 29990
    },
    {
      "entropy": 5.415493154525757,
      "epoch": 3.368731397764924,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038992044534728404,
      "loss": 5.0918,
      "mean_token_accuracy": 0.2100331738591194,
      "num_tokens": 66951732.0,
      "step": 29995
    },
    {
      "entropy": 5.395217990875244,
      "epoch": 3.3692929746728817,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038988592666827594,
      "loss": 5.056,
      "mean_token_accuracy": 0.21412342935800552,
      "num_tokens": 66962941.0,
      "step": 30000
    },
    {
      "epoch": 3.3692929746728817,
      "eval_entropy": 5.177071620300624,
      "eval_loss": 5.1879377365112305,
      "eval_mean_token_accuracy": 0.21260422172622026,
      "eval_num_tokens": 66962941.0,
      "eval_runtime": 23.8708,
      "eval_samples_per_second": 1299.201,
      "eval_steps_per_second": 162.416,
      "step": 30000
    },
    {
      "entropy": 5.4534954071044925,
      "epoch": 3.369854551580839,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003898514043312553,
      "loss": 5.1638,
      "mean_token_accuracy": 0.20689600855112075,
      "num_tokens": 66975096.0,
      "step": 30005
    },
    {
      "entropy": 5.451320314407349,
      "epoch": 3.3704161284887966,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038981687833732146,
      "loss": 5.0678,
      "mean_token_accuracy": 0.21220825910568236,
      "num_tokens": 66987621.0,
      "step": 30010
    },
    {
      "entropy": 5.403622674942016,
      "epoch": 3.3709777053967542,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003897823486875737,
      "loss": 5.1251,
      "mean_token_accuracy": 0.20674302726984023,
      "num_tokens": 66998801.0,
      "step": 30015
    },
    {
      "entropy": 5.4244894027709964,
      "epoch": 3.3715392823047114,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003897478153831114,
      "loss": 5.1139,
      "mean_token_accuracy": 0.20223870128393173,
      "num_tokens": 67010698.0,
      "step": 30020
    },
    {
      "entropy": 5.380762338638306,
      "epoch": 3.372100859212669,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038971327842503414,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21794713884592057,
      "num_tokens": 67022487.0,
      "step": 30025
    },
    {
      "entropy": 5.343285703659058,
      "epoch": 3.3726624361206268,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038967873781444167,
      "loss": 5.0215,
      "mean_token_accuracy": 0.21425683200359344,
      "num_tokens": 67032517.0,
      "step": 30030
    },
    {
      "entropy": 5.345498132705688,
      "epoch": 3.3732240130285844,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038964419355243365,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2158302441239357,
      "num_tokens": 67042466.0,
      "step": 30035
    },
    {
      "entropy": 5.413795280456543,
      "epoch": 3.3737855899365417,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003896096456401101,
      "loss": 5.09,
      "mean_token_accuracy": 0.21146479547023772,
      "num_tokens": 67053090.0,
      "step": 30040
    },
    {
      "entropy": 5.362494850158692,
      "epoch": 3.3743471668444993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038957509407857097,
      "loss": 4.9402,
      "mean_token_accuracy": 0.21361034661531447,
      "num_tokens": 67064332.0,
      "step": 30045
    },
    {
      "entropy": 5.31009840965271,
      "epoch": 3.374908743752457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038954053886891635,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21242313981056213,
      "num_tokens": 67074788.0,
      "step": 30050
    },
    {
      "entropy": 5.320407247543335,
      "epoch": 3.3754703206604146,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003895059800122466,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21271974295377732,
      "num_tokens": 67086510.0,
      "step": 30055
    },
    {
      "entropy": 5.385377216339111,
      "epoch": 3.376031897568372,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000389471417509662,
      "loss": 5.0317,
      "mean_token_accuracy": 0.2123967170715332,
      "num_tokens": 67096561.0,
      "step": 30060
    },
    {
      "entropy": 5.332107591629028,
      "epoch": 3.3765934744763295,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000389436851362263,
      "loss": 4.9908,
      "mean_token_accuracy": 0.21586178988218307,
      "num_tokens": 67107897.0,
      "step": 30065
    },
    {
      "entropy": 5.335859155654907,
      "epoch": 3.377155051384287,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038940228157115036,
      "loss": 5.1193,
      "mean_token_accuracy": 0.20771542936563492,
      "num_tokens": 67119686.0,
      "step": 30070
    },
    {
      "entropy": 5.419267702102661,
      "epoch": 3.3777166282922444,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038936770813742473,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21207644492387773,
      "num_tokens": 67131041.0,
      "step": 30075
    },
    {
      "entropy": 5.513225650787353,
      "epoch": 3.378278205200202,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038933313106218694,
      "loss": 5.1416,
      "mean_token_accuracy": 0.21388377994298935,
      "num_tokens": 67143065.0,
      "step": 30080
    },
    {
      "entropy": 5.361063814163208,
      "epoch": 3.3788397821081597,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038929855034653776,
      "loss": 4.9715,
      "mean_token_accuracy": 0.21915498226881028,
      "num_tokens": 67154173.0,
      "step": 30085
    },
    {
      "entropy": 5.320414590835571,
      "epoch": 3.3794013590161174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038926396599157855,
      "loss": 5.0297,
      "mean_token_accuracy": 0.2054807499051094,
      "num_tokens": 67165909.0,
      "step": 30090
    },
    {
      "entropy": 5.347616338729859,
      "epoch": 3.379962935924075,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003892293779984103,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21979046165943145,
      "num_tokens": 67177474.0,
      "step": 30095
    },
    {
      "entropy": 5.428892421722412,
      "epoch": 3.3805245128320323,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003891947863681344,
      "loss": 5.099,
      "mean_token_accuracy": 0.207650788128376,
      "num_tokens": 67187590.0,
      "step": 30100
    },
    {
      "entropy": 5.308780527114868,
      "epoch": 3.38108608973999,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003891601911018521,
      "loss": 4.9759,
      "mean_token_accuracy": 0.21467913538217545,
      "num_tokens": 67198579.0,
      "step": 30105
    },
    {
      "entropy": 5.350521564483643,
      "epoch": 3.3816476666479476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003891255922006652,
      "loss": 4.9994,
      "mean_token_accuracy": 0.22020458281040192,
      "num_tokens": 67208503.0,
      "step": 30110
    },
    {
      "entropy": 5.295143270492554,
      "epoch": 3.382209243555905,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000389090989665675,
      "loss": 4.9705,
      "mean_token_accuracy": 0.21536457538604736,
      "num_tokens": 67219363.0,
      "step": 30115
    },
    {
      "entropy": 5.292697238922119,
      "epoch": 3.3827708204638625,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003890563834979835,
      "loss": 5.063,
      "mean_token_accuracy": 0.20983703434467316,
      "num_tokens": 67231030.0,
      "step": 30120
    },
    {
      "entropy": 5.444101715087891,
      "epoch": 3.38333239737182,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003890217736986926,
      "loss": 5.0372,
      "mean_token_accuracy": 0.20746132582426072,
      "num_tokens": 67242962.0,
      "step": 30125
    },
    {
      "entropy": 5.420209312438965,
      "epoch": 3.383893974279778,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038898716026890414,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20374323576688766,
      "num_tokens": 67253871.0,
      "step": 30130
    },
    {
      "entropy": 5.366298198699951,
      "epoch": 3.384455551187735,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003889525432097203,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21110064387321473,
      "num_tokens": 67265875.0,
      "step": 30135
    },
    {
      "entropy": 5.3291843891143795,
      "epoch": 3.3850171280956927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003889179225222432,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21602510958909987,
      "num_tokens": 67274920.0,
      "step": 30140
    },
    {
      "entropy": 5.306484079360962,
      "epoch": 3.3855787050036503,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038888329820757536,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2148293748497963,
      "num_tokens": 67286440.0,
      "step": 30145
    },
    {
      "entropy": 5.345148372650146,
      "epoch": 3.386140281911608,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038884867026681906,
      "loss": 4.972,
      "mean_token_accuracy": 0.21785303056240082,
      "num_tokens": 67297803.0,
      "step": 30150
    },
    {
      "entropy": 5.242723798751831,
      "epoch": 3.386701858819565,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003888140387010771,
      "loss": 4.8958,
      "mean_token_accuracy": 0.2230183884501457,
      "num_tokens": 67309153.0,
      "step": 30155
    },
    {
      "entropy": 5.235397720336914,
      "epoch": 3.387263435727523,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038877940351145186,
      "loss": 4.9864,
      "mean_token_accuracy": 0.21358411461114885,
      "num_tokens": 67319695.0,
      "step": 30160
    },
    {
      "entropy": 5.335029792785645,
      "epoch": 3.3878250126354805,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038874476469904633,
      "loss": 4.9917,
      "mean_token_accuracy": 0.2187188521027565,
      "num_tokens": 67331615.0,
      "step": 30165
    },
    {
      "entropy": 5.322209358215332,
      "epoch": 3.3883865895434377,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003887101222649634,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22161663323640823,
      "num_tokens": 67341898.0,
      "step": 30170
    },
    {
      "entropy": 5.386251640319824,
      "epoch": 3.3889481664513954,
      "grad_norm": 1.015625,
      "learning_rate": 0.000388675476210306,
      "loss": 5.0832,
      "mean_token_accuracy": 0.21715029180049897,
      "num_tokens": 67352677.0,
      "step": 30175
    },
    {
      "entropy": 5.333758211135864,
      "epoch": 3.389509743359353,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038864082653617744,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21207329779863357,
      "num_tokens": 67363215.0,
      "step": 30180
    },
    {
      "entropy": 5.319951772689819,
      "epoch": 3.3900713202673107,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038860617324368085,
      "loss": 5.0187,
      "mean_token_accuracy": 0.21429772973060607,
      "num_tokens": 67374489.0,
      "step": 30185
    },
    {
      "entropy": 5.395950794219971,
      "epoch": 3.3906328971752684,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003885715163339196,
      "loss": 5.0354,
      "mean_token_accuracy": 0.21528981775045394,
      "num_tokens": 67384890.0,
      "step": 30190
    },
    {
      "entropy": 5.3208846092224125,
      "epoch": 3.3911944740832256,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038853685580799726,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21424264311790467,
      "num_tokens": 67395811.0,
      "step": 30195
    },
    {
      "entropy": 5.317122507095337,
      "epoch": 3.3917560509911833,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003885021916670174,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22014525085687636,
      "num_tokens": 67407056.0,
      "step": 30200
    },
    {
      "entropy": 5.374187326431274,
      "epoch": 3.392317627899141,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003884675239120838,
      "loss": 5.0392,
      "mean_token_accuracy": 0.21116345077753068,
      "num_tokens": 67418371.0,
      "step": 30205
    },
    {
      "entropy": 5.367884683609009,
      "epoch": 3.392879204807098,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038843285254430005,
      "loss": 5.0301,
      "mean_token_accuracy": 0.21094926595687866,
      "num_tokens": 67429165.0,
      "step": 30210
    },
    {
      "entropy": 5.5298666000366214,
      "epoch": 3.393440781715056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003883981775647705,
      "loss": 5.2464,
      "mean_token_accuracy": 0.20436334908008574,
      "num_tokens": 67442317.0,
      "step": 30215
    },
    {
      "entropy": 5.509282922744751,
      "epoch": 3.3940023586230135,
      "grad_norm": 1.0,
      "learning_rate": 0.0003883634989745988,
      "loss": 5.1428,
      "mean_token_accuracy": 0.21388594359159468,
      "num_tokens": 67454086.0,
      "step": 30220
    },
    {
      "entropy": 5.419427347183228,
      "epoch": 3.394563935530971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038832881677488944,
      "loss": 5.1027,
      "mean_token_accuracy": 0.202974334359169,
      "num_tokens": 67465066.0,
      "step": 30225
    },
    {
      "entropy": 5.291957569122315,
      "epoch": 3.3951255124389284,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038829413096674653,
      "loss": 4.954,
      "mean_token_accuracy": 0.21786050200462342,
      "num_tokens": 67475456.0,
      "step": 30230
    },
    {
      "entropy": 5.373084306716919,
      "epoch": 3.395687089346886,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003882594415512746,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21428725570440293,
      "num_tokens": 67486315.0,
      "step": 30235
    },
    {
      "entropy": 5.318059396743775,
      "epoch": 3.3962486662548437,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038822474852957817,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21651425808668137,
      "num_tokens": 67497431.0,
      "step": 30240
    },
    {
      "entropy": 5.329329109191894,
      "epoch": 3.3968102431628013,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003881900519027618,
      "loss": 5.0248,
      "mean_token_accuracy": 0.21906078308820726,
      "num_tokens": 67508054.0,
      "step": 30245
    },
    {
      "entropy": 5.414961338043213,
      "epoch": 3.3973718200707586,
      "grad_norm": 1.0,
      "learning_rate": 0.0003881553516719303,
      "loss": 5.0811,
      "mean_token_accuracy": 0.20679268538951873,
      "num_tokens": 67519083.0,
      "step": 30250
    },
    {
      "entropy": 5.40166482925415,
      "epoch": 3.397933396978716,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038812064783818847,
      "loss": 5.0616,
      "mean_token_accuracy": 0.21086174249649048,
      "num_tokens": 67528909.0,
      "step": 30255
    },
    {
      "entropy": 5.251445341110229,
      "epoch": 3.398494973886674,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038808594040264127,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22917675822973252,
      "num_tokens": 67540069.0,
      "step": 30260
    },
    {
      "entropy": 5.292755794525147,
      "epoch": 3.399056550794631,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000388051229366394,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21921857297420502,
      "num_tokens": 67550733.0,
      "step": 30265
    },
    {
      "entropy": 5.461176204681396,
      "epoch": 3.3996181277025888,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038801651473055167,
      "loss": 5.1395,
      "mean_token_accuracy": 0.20966355353593827,
      "num_tokens": 67562118.0,
      "step": 30270
    },
    {
      "entropy": 5.345594835281372,
      "epoch": 3.4001797046105464,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003879817964962197,
      "loss": 4.9524,
      "mean_token_accuracy": 0.2192150041460991,
      "num_tokens": 67572408.0,
      "step": 30275
    },
    {
      "entropy": 5.408023834228516,
      "epoch": 3.400741281518504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003879470746645035,
      "loss": 5.149,
      "mean_token_accuracy": 0.20989259332418442,
      "num_tokens": 67584125.0,
      "step": 30280
    },
    {
      "entropy": 5.376199388504029,
      "epoch": 3.4013028584264617,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038791234923650866,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21159509867429732,
      "num_tokens": 67595093.0,
      "step": 30285
    },
    {
      "entropy": 5.414803075790405,
      "epoch": 3.401864435334419,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038787762021334077,
      "loss": 5.0885,
      "mean_token_accuracy": 0.20873144119977952,
      "num_tokens": 67606794.0,
      "step": 30290
    },
    {
      "entropy": 5.291760015487671,
      "epoch": 3.4024260122423766,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003878428875961057,
      "loss": 4.8014,
      "mean_token_accuracy": 0.2295370355248451,
      "num_tokens": 67617207.0,
      "step": 30295
    },
    {
      "entropy": 5.212384223937988,
      "epoch": 3.4029875891503343,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038780815138590935,
      "loss": 4.9454,
      "mean_token_accuracy": 0.21602405607700348,
      "num_tokens": 67628301.0,
      "step": 30300
    },
    {
      "entropy": 5.365674209594727,
      "epoch": 3.4035491660582915,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003877734115838577,
      "loss": 5.0714,
      "mean_token_accuracy": 0.2103058859705925,
      "num_tokens": 67640363.0,
      "step": 30305
    },
    {
      "entropy": 5.384312009811401,
      "epoch": 3.404110742966249,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038773866819105684,
      "loss": 5.0575,
      "mean_token_accuracy": 0.20850632786750795,
      "num_tokens": 67651515.0,
      "step": 30310
    },
    {
      "entropy": 5.3321919441223145,
      "epoch": 3.404672319874207,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003877039212086131,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21417766958475112,
      "num_tokens": 67661669.0,
      "step": 30315
    },
    {
      "entropy": 5.336136722564698,
      "epoch": 3.4052338967821645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038766917063763287,
      "loss": 4.943,
      "mean_token_accuracy": 0.21848686337471007,
      "num_tokens": 67671945.0,
      "step": 30320
    },
    {
      "entropy": 5.3223127841949465,
      "epoch": 3.4057954736901217,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038763441647922246,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20769304186105728,
      "num_tokens": 67683090.0,
      "step": 30325
    },
    {
      "entropy": 5.353286457061768,
      "epoch": 3.4063570505980794,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003875996587344885,
      "loss": 5.0283,
      "mean_token_accuracy": 0.218906469643116,
      "num_tokens": 67694036.0,
      "step": 30330
    },
    {
      "entropy": 5.429412937164306,
      "epoch": 3.406918627506037,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038756489740453785,
      "loss": 5.0101,
      "mean_token_accuracy": 0.21323282718658448,
      "num_tokens": 67704677.0,
      "step": 30335
    },
    {
      "entropy": 5.3625938415527346,
      "epoch": 3.4074802044139947,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038753013249047706,
      "loss": 5.0701,
      "mean_token_accuracy": 0.21416995972394942,
      "num_tokens": 67715165.0,
      "step": 30340
    },
    {
      "entropy": 5.326519298553467,
      "epoch": 3.408041781321952,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003874953639934134,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2228758916258812,
      "num_tokens": 67726226.0,
      "step": 30345
    },
    {
      "entropy": 5.3486326217651365,
      "epoch": 3.4086033582299096,
      "grad_norm": 1.0,
      "learning_rate": 0.0003874605919144536,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2193448141217232,
      "num_tokens": 67739564.0,
      "step": 30350
    },
    {
      "entropy": 5.3447465896606445,
      "epoch": 3.4091649351378672,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003874258162547049,
      "loss": 5.115,
      "mean_token_accuracy": 0.2124485731124878,
      "num_tokens": 67751683.0,
      "step": 30355
    },
    {
      "entropy": 5.377430725097656,
      "epoch": 3.4097265120458244,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003873910370152747,
      "loss": 5.0535,
      "mean_token_accuracy": 0.2136751651763916,
      "num_tokens": 67762782.0,
      "step": 30360
    },
    {
      "entropy": 5.406807851791382,
      "epoch": 3.410288088953782,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003873562541972702,
      "loss": 5.023,
      "mean_token_accuracy": 0.21507148295640946,
      "num_tokens": 67773583.0,
      "step": 30365
    },
    {
      "entropy": 5.333729934692383,
      "epoch": 3.4108496658617398,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003873214678017991,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21802588552236557,
      "num_tokens": 67785428.0,
      "step": 30370
    },
    {
      "entropy": 5.439136505126953,
      "epoch": 3.4114112427696974,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003872866778299688,
      "loss": 5.1234,
      "mean_token_accuracy": 0.2071713387966156,
      "num_tokens": 67797678.0,
      "step": 30375
    },
    {
      "entropy": 5.4954229354858395,
      "epoch": 3.411972819677655,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003872518842828872,
      "loss": 5.0535,
      "mean_token_accuracy": 0.21821678578853607,
      "num_tokens": 67809314.0,
      "step": 30380
    },
    {
      "entropy": 5.319619178771973,
      "epoch": 3.4125343965856123,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038721708716166203,
      "loss": 5.0368,
      "mean_token_accuracy": 0.2156158372759819,
      "num_tokens": 67820591.0,
      "step": 30385
    },
    {
      "entropy": 5.278154897689819,
      "epoch": 3.41309597349357,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003871822864674013,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21803202629089355,
      "num_tokens": 67831436.0,
      "step": 30390
    },
    {
      "entropy": 5.3759815216064455,
      "epoch": 3.4136575504015276,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038714748220121305,
      "loss": 5.0463,
      "mean_token_accuracy": 0.21730043441057206,
      "num_tokens": 67842813.0,
      "step": 30395
    },
    {
      "entropy": 5.417090892791748,
      "epoch": 3.414219127309485,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003871126743642055,
      "loss": 5.0569,
      "mean_token_accuracy": 0.2078885480761528,
      "num_tokens": 67853899.0,
      "step": 30400
    },
    {
      "entropy": 5.396058845520019,
      "epoch": 3.4147807042174425,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038707786295748693,
      "loss": 4.9993,
      "mean_token_accuracy": 0.215189366042614,
      "num_tokens": 67864237.0,
      "step": 30405
    },
    {
      "entropy": 5.3132318496704105,
      "epoch": 3.4153422811254,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038704304798216566,
      "loss": 4.98,
      "mean_token_accuracy": 0.20995001643896102,
      "num_tokens": 67876109.0,
      "step": 30410
    },
    {
      "entropy": 5.27262544631958,
      "epoch": 3.415903858033358,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003870082294393503,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2108785793185234,
      "num_tokens": 67888461.0,
      "step": 30415
    },
    {
      "entropy": 5.358737277984619,
      "epoch": 3.416465434941315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003869734073301495,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21752513498067855,
      "num_tokens": 67899543.0,
      "step": 30420
    },
    {
      "entropy": 5.4097377300262455,
      "epoch": 3.4170270118492727,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038693858165567195,
      "loss": 5.11,
      "mean_token_accuracy": 0.21224105805158616,
      "num_tokens": 67910606.0,
      "step": 30425
    },
    {
      "entropy": 5.358897256851196,
      "epoch": 3.4175885887572304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003869037524170266,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21804407835006714,
      "num_tokens": 67921240.0,
      "step": 30430
    },
    {
      "entropy": 5.433197975158691,
      "epoch": 3.418150165665188,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038686891961532227,
      "loss": 5.2408,
      "mean_token_accuracy": 0.20136552304029465,
      "num_tokens": 67935178.0,
      "step": 30435
    },
    {
      "entropy": 5.277958011627197,
      "epoch": 3.4187117425731453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003868340832516683,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22259962558746338,
      "num_tokens": 67947260.0,
      "step": 30440
    },
    {
      "entropy": 5.339145851135254,
      "epoch": 3.419273319481103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038679924332717363,
      "loss": 4.9466,
      "mean_token_accuracy": 0.217559514939785,
      "num_tokens": 67956665.0,
      "step": 30445
    },
    {
      "entropy": 5.351485729217529,
      "epoch": 3.4198348963890606,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038676439984294776,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21113453209400176,
      "num_tokens": 67967927.0,
      "step": 30450
    },
    {
      "entropy": 5.343424606323242,
      "epoch": 3.420396473297018,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038672955280009996,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21755767315626146,
      "num_tokens": 67979574.0,
      "step": 30455
    },
    {
      "entropy": 5.347293281555176,
      "epoch": 3.4209580502049755,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038669470219973996,
      "loss": 5.0761,
      "mean_token_accuracy": 0.21034935116767883,
      "num_tokens": 67991671.0,
      "step": 30460
    },
    {
      "entropy": 5.449505567550659,
      "epoch": 3.421519627112933,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038665984804297734,
      "loss": 5.0768,
      "mean_token_accuracy": 0.20592614263296127,
      "num_tokens": 68002443.0,
      "step": 30465
    },
    {
      "entropy": 5.380095624923706,
      "epoch": 3.4220812040208908,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003866249903309218,
      "loss": 4.9997,
      "mean_token_accuracy": 0.20927030444145203,
      "num_tokens": 68013681.0,
      "step": 30470
    },
    {
      "entropy": 5.396969032287598,
      "epoch": 3.4226427809288484,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003865901290646833,
      "loss": 5.0187,
      "mean_token_accuracy": 0.21917644441127776,
      "num_tokens": 68023757.0,
      "step": 30475
    },
    {
      "entropy": 5.278738975524902,
      "epoch": 3.4232043578368057,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038655526424537175,
      "loss": 5.0019,
      "mean_token_accuracy": 0.2140437752008438,
      "num_tokens": 68035491.0,
      "step": 30480
    },
    {
      "entropy": 5.286426401138305,
      "epoch": 3.4237659347447633,
      "grad_norm": 1.0,
      "learning_rate": 0.00038652039587409734,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21297864317893983,
      "num_tokens": 68047581.0,
      "step": 30485
    },
    {
      "entropy": 5.390673398971558,
      "epoch": 3.424327511652721,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003864855239519703,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21332051157951354,
      "num_tokens": 68058232.0,
      "step": 30490
    },
    {
      "entropy": 5.536237001419067,
      "epoch": 3.424889088560678,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00038645064848010093,
      "loss": 5.211,
      "mean_token_accuracy": 0.2048870176076889,
      "num_tokens": 68071602.0,
      "step": 30495
    },
    {
      "entropy": 5.381994676589966,
      "epoch": 3.425450665468636,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038641576945959964,
      "loss": 5.0093,
      "mean_token_accuracy": 0.2121684581041336,
      "num_tokens": 68083169.0,
      "step": 30500
    },
    {
      "entropy": 5.379086303710937,
      "epoch": 3.4260122423765935,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038638088689157713,
      "loss": 5.0406,
      "mean_token_accuracy": 0.20722143352031708,
      "num_tokens": 68093951.0,
      "step": 30505
    },
    {
      "entropy": 5.437827205657959,
      "epoch": 3.426573819284551,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038634600077714395,
      "loss": 5.1285,
      "mean_token_accuracy": 0.20868780761957167,
      "num_tokens": 68105398.0,
      "step": 30510
    },
    {
      "entropy": 5.380607986450196,
      "epoch": 3.4271353961925084,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003863111111174109,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21454898864030839,
      "num_tokens": 68116544.0,
      "step": 30515
    },
    {
      "entropy": 5.380779790878296,
      "epoch": 3.427696973100466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038627621791348886,
      "loss": 4.9695,
      "mean_token_accuracy": 0.22695867866277694,
      "num_tokens": 68126977.0,
      "step": 30520
    },
    {
      "entropy": 5.451254510879517,
      "epoch": 3.4282585500084237,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038624132116648897,
      "loss": 5.1336,
      "mean_token_accuracy": 0.20618606954813004,
      "num_tokens": 68137884.0,
      "step": 30525
    },
    {
      "entropy": 5.423811721801758,
      "epoch": 3.4288201269163814,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003862064208775223,
      "loss": 5.1144,
      "mean_token_accuracy": 0.21106421649456025,
      "num_tokens": 68149026.0,
      "step": 30530
    },
    {
      "entropy": 5.360766315460205,
      "epoch": 3.4293817038243386,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038617151704769994,
      "loss": 5.0432,
      "mean_token_accuracy": 0.2135568991303444,
      "num_tokens": 68159962.0,
      "step": 30535
    },
    {
      "entropy": 5.39732985496521,
      "epoch": 3.4299432807322963,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003861366096781334,
      "loss": 5.0341,
      "mean_token_accuracy": 0.21574080139398574,
      "num_tokens": 68170849.0,
      "step": 30540
    },
    {
      "entropy": 5.326971769332886,
      "epoch": 3.430504857640254,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003861016987699342,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21906704306602479,
      "num_tokens": 68181623.0,
      "step": 30545
    },
    {
      "entropy": 5.2901129722595215,
      "epoch": 3.431066434548211,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038606678432421366,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22573311179876326,
      "num_tokens": 68191943.0,
      "step": 30550
    },
    {
      "entropy": 5.297085523605347,
      "epoch": 3.431628011456169,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003860318663420837,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21072250008583068,
      "num_tokens": 68203833.0,
      "step": 30555
    },
    {
      "entropy": 5.310635280609131,
      "epoch": 3.4321895883641265,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038599694482465606,
      "loss": 5.0036,
      "mean_token_accuracy": 0.21280494332313538,
      "num_tokens": 68213992.0,
      "step": 30560
    },
    {
      "entropy": 5.35171365737915,
      "epoch": 3.432751165272084,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038596201977304263,
      "loss": 5.0004,
      "mean_token_accuracy": 0.21705715358257294,
      "num_tokens": 68224575.0,
      "step": 30565
    },
    {
      "entropy": 5.373250818252563,
      "epoch": 3.433312742180042,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038592709118835547,
      "loss": 5.0483,
      "mean_token_accuracy": 0.2093303844332695,
      "num_tokens": 68235513.0,
      "step": 30570
    },
    {
      "entropy": 5.451681804656983,
      "epoch": 3.433874319087999,
      "grad_norm": 1.0,
      "learning_rate": 0.00038589215907170664,
      "loss": 5.0681,
      "mean_token_accuracy": 0.2130262777209282,
      "num_tokens": 68246560.0,
      "step": 30575
    },
    {
      "entropy": 5.268670272827149,
      "epoch": 3.4344358959959567,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003858572234242086,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22357730567455292,
      "num_tokens": 68257760.0,
      "step": 30580
    },
    {
      "entropy": 5.2810564041137695,
      "epoch": 3.4349974729039143,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038582228424697347,
      "loss": 4.9204,
      "mean_token_accuracy": 0.23064145147800447,
      "num_tokens": 68269631.0,
      "step": 30585
    },
    {
      "entropy": 5.387024307250977,
      "epoch": 3.4355590498118715,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003857873415411138,
      "loss": 5.0393,
      "mean_token_accuracy": 0.21885640323162078,
      "num_tokens": 68280198.0,
      "step": 30590
    },
    {
      "entropy": 5.35499587059021,
      "epoch": 3.436120626719829,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003857523953077423,
      "loss": 5.0839,
      "mean_token_accuracy": 0.21083260029554368,
      "num_tokens": 68291812.0,
      "step": 30595
    },
    {
      "entropy": 5.390143966674804,
      "epoch": 3.436682203627787,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038571744554797156,
      "loss": 5.0812,
      "mean_token_accuracy": 0.21129658669233323,
      "num_tokens": 68303202.0,
      "step": 30600
    },
    {
      "entropy": 5.3266161441802975,
      "epoch": 3.4372437805357445,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038568249226291433,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2145319402217865,
      "num_tokens": 68313768.0,
      "step": 30605
    },
    {
      "entropy": 5.368358755111695,
      "epoch": 3.4378053574437017,
      "grad_norm": 1.0,
      "learning_rate": 0.00038564753545368375,
      "loss": 5.0219,
      "mean_token_accuracy": 0.21536147892475127,
      "num_tokens": 68325325.0,
      "step": 30610
    },
    {
      "entropy": 5.40307469367981,
      "epoch": 3.4383669343516594,
      "grad_norm": 1.0,
      "learning_rate": 0.00038561257512139264,
      "loss": 5.1066,
      "mean_token_accuracy": 0.21310515850782394,
      "num_tokens": 68337166.0,
      "step": 30615
    },
    {
      "entropy": 5.388151359558106,
      "epoch": 3.438928511259617,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038557761126715426,
      "loss": 5.0132,
      "mean_token_accuracy": 0.22017728090286254,
      "num_tokens": 68348898.0,
      "step": 30620
    },
    {
      "entropy": 5.381800746917724,
      "epoch": 3.4394900881675747,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003855426438920818,
      "loss": 5.0779,
      "mean_token_accuracy": 0.21035739481449128,
      "num_tokens": 68360589.0,
      "step": 30625
    },
    {
      "entropy": 5.239636707305908,
      "epoch": 3.440051665075532,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038550767299728875,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22481377720832824,
      "num_tokens": 68371813.0,
      "step": 30630
    },
    {
      "entropy": 5.369158220291138,
      "epoch": 3.4406132419834896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038547269858388853,
      "loss": 5.0985,
      "mean_token_accuracy": 0.21178159415721892,
      "num_tokens": 68382060.0,
      "step": 30635
    },
    {
      "entropy": 5.422509670257568,
      "epoch": 3.4411748188914473,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003854377206529947,
      "loss": 5.0996,
      "mean_token_accuracy": 0.21597285717725753,
      "num_tokens": 68395603.0,
      "step": 30640
    },
    {
      "entropy": 5.356976175308228,
      "epoch": 3.4417363957994045,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000385402739205721,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2228722557425499,
      "num_tokens": 68406725.0,
      "step": 30645
    },
    {
      "entropy": 5.376346254348755,
      "epoch": 3.442297972707362,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003853677542431813,
      "loss": 5.1309,
      "mean_token_accuracy": 0.20524288713932037,
      "num_tokens": 68418577.0,
      "step": 30650
    },
    {
      "entropy": 5.326123189926148,
      "epoch": 3.44285954961532,
      "grad_norm": 1.0,
      "learning_rate": 0.0003853327657664895,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21629838645458221,
      "num_tokens": 68429708.0,
      "step": 30655
    },
    {
      "entropy": 5.38738784790039,
      "epoch": 3.4434211265232775,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003852977737767596,
      "loss": 5.1312,
      "mean_token_accuracy": 0.20819007903337478,
      "num_tokens": 68442115.0,
      "step": 30660
    },
    {
      "entropy": 5.361652088165283,
      "epoch": 3.443982703431235,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003852627782751059,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2177476093173027,
      "num_tokens": 68453555.0,
      "step": 30665
    },
    {
      "entropy": 5.321253395080566,
      "epoch": 3.4445442803391924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003852277792626425,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22145827412605285,
      "num_tokens": 68466859.0,
      "step": 30670
    },
    {
      "entropy": 5.29543628692627,
      "epoch": 3.44510585724715,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003851927767404839,
      "loss": 5.0043,
      "mean_token_accuracy": 0.21579211056232453,
      "num_tokens": 68477628.0,
      "step": 30675
    },
    {
      "entropy": 5.3807878494262695,
      "epoch": 3.4456674341551077,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003851577707097445,
      "loss": 5.1038,
      "mean_token_accuracy": 0.2164286732673645,
      "num_tokens": 68488764.0,
      "step": 30680
    },
    {
      "entropy": 5.460570240020752,
      "epoch": 3.446229011063065,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000385122761171539,
      "loss": 5.1394,
      "mean_token_accuracy": 0.21168999075889589,
      "num_tokens": 68499433.0,
      "step": 30685
    },
    {
      "entropy": 5.391172170639038,
      "epoch": 3.4467905879710226,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038508774812698207,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21339063793420793,
      "num_tokens": 68510832.0,
      "step": 30690
    },
    {
      "entropy": 5.2627920627594,
      "epoch": 3.44735216487898,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003850527315771885,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21751293987035752,
      "num_tokens": 68520946.0,
      "step": 30695
    },
    {
      "entropy": 5.374011707305908,
      "epoch": 3.447913741786938,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038501771152327333,
      "loss": 5.0956,
      "mean_token_accuracy": 0.21109937727451325,
      "num_tokens": 68531769.0,
      "step": 30700
    },
    {
      "entropy": 5.445743131637573,
      "epoch": 3.448475318694895,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038498268796635155,
      "loss": 5.112,
      "mean_token_accuracy": 0.2046330213546753,
      "num_tokens": 68543172.0,
      "step": 30705
    },
    {
      "entropy": 5.276662063598633,
      "epoch": 3.4490368956028528,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038494766090753824,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21804124861955643,
      "num_tokens": 68554078.0,
      "step": 30710
    },
    {
      "entropy": 5.315173625946045,
      "epoch": 3.4495984725108104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038491263034794886,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21862834542989731,
      "num_tokens": 68564705.0,
      "step": 30715
    },
    {
      "entropy": 5.439404582977295,
      "epoch": 3.450160049418768,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003848775962886987,
      "loss": 5.2222,
      "mean_token_accuracy": 0.2029855564236641,
      "num_tokens": 68577010.0,
      "step": 30720
    },
    {
      "entropy": 5.423171758651733,
      "epoch": 3.4507216263267253,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003848425587309032,
      "loss": 5.0551,
      "mean_token_accuracy": 0.21266096830368042,
      "num_tokens": 68588465.0,
      "step": 30725
    },
    {
      "entropy": 5.385078620910645,
      "epoch": 3.451283203234683,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038480751767567807,
      "loss": 5.0429,
      "mean_token_accuracy": 0.21389450877904892,
      "num_tokens": 68599666.0,
      "step": 30730
    },
    {
      "entropy": 5.31045823097229,
      "epoch": 3.4518447801426406,
      "grad_norm": 0.921875,
      "learning_rate": 0.000384772473124139,
      "loss": 4.9971,
      "mean_token_accuracy": 0.21728697121143342,
      "num_tokens": 68610530.0,
      "step": 30735
    },
    {
      "entropy": 5.269164085388184,
      "epoch": 3.452406357050598,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003847374250774018,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21999448239803315,
      "num_tokens": 68621360.0,
      "step": 30740
    },
    {
      "entropy": 5.44544677734375,
      "epoch": 3.4529679339585555,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038470237353658237,
      "loss": 5.0525,
      "mean_token_accuracy": 0.21059583872556686,
      "num_tokens": 68632874.0,
      "step": 30745
    },
    {
      "entropy": 5.3436933040618895,
      "epoch": 3.453529510866513,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038466731850279683,
      "loss": 5.0281,
      "mean_token_accuracy": 0.21019955426454545,
      "num_tokens": 68643715.0,
      "step": 30750
    },
    {
      "entropy": 5.301168394088745,
      "epoch": 3.454091087774471,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038463225997716137,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20859742164611816,
      "num_tokens": 68655433.0,
      "step": 30755
    },
    {
      "entropy": 5.411481952667236,
      "epoch": 3.4546526646824285,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038459719796079227,
      "loss": 5.0589,
      "mean_token_accuracy": 0.21599926054477692,
      "num_tokens": 68667321.0,
      "step": 30760
    },
    {
      "entropy": 5.408344554901123,
      "epoch": 3.4552142415903857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003845621324548059,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2113675743341446,
      "num_tokens": 68677314.0,
      "step": 30765
    },
    {
      "entropy": 5.360374975204468,
      "epoch": 3.4557758184983434,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003845270634603186,
      "loss": 5.1087,
      "mean_token_accuracy": 0.212170971930027,
      "num_tokens": 68688871.0,
      "step": 30770
    },
    {
      "entropy": 5.364226484298706,
      "epoch": 3.456337395406301,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038449199097844724,
      "loss": 5.1064,
      "mean_token_accuracy": 0.21647169589996337,
      "num_tokens": 68699760.0,
      "step": 30775
    },
    {
      "entropy": 5.504874753952026,
      "epoch": 3.4568989723142582,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003844569150103084,
      "loss": 5.1566,
      "mean_token_accuracy": 0.20138760954141616,
      "num_tokens": 68711113.0,
      "step": 30780
    },
    {
      "entropy": 5.428476285934448,
      "epoch": 3.457460549222216,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003844218355570189,
      "loss": 5.0119,
      "mean_token_accuracy": 0.21746293902397157,
      "num_tokens": 68722904.0,
      "step": 30785
    },
    {
      "entropy": 5.306151008605957,
      "epoch": 3.4580221261301736,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003843867526196957,
      "loss": 4.9522,
      "mean_token_accuracy": 0.2152404099702835,
      "num_tokens": 68733753.0,
      "step": 30790
    },
    {
      "entropy": 5.3150684356689455,
      "epoch": 3.4585837030381312,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003843516661994559,
      "loss": 5.0386,
      "mean_token_accuracy": 0.2081532746553421,
      "num_tokens": 68744477.0,
      "step": 30795
    },
    {
      "entropy": 5.448066377639771,
      "epoch": 3.4591452799460884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003843165762974167,
      "loss": 5.107,
      "mean_token_accuracy": 0.21200543642044067,
      "num_tokens": 68755607.0,
      "step": 30800
    },
    {
      "entropy": 5.4901978969573975,
      "epoch": 3.459706856854046,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038428148291469515,
      "loss": 5.077,
      "mean_token_accuracy": 0.21149835139513015,
      "num_tokens": 68766477.0,
      "step": 30805
    },
    {
      "entropy": 5.32675199508667,
      "epoch": 3.4602684337620038,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000384246386052409,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2280939221382141,
      "num_tokens": 68776806.0,
      "step": 30810
    },
    {
      "entropy": 5.270185565948486,
      "epoch": 3.4608300106699614,
      "grad_norm": 1.0,
      "learning_rate": 0.0003842112857116754,
      "loss": 5.0015,
      "mean_token_accuracy": 0.21584051996469497,
      "num_tokens": 68788011.0,
      "step": 30815
    },
    {
      "entropy": 5.319165182113648,
      "epoch": 3.4613915875779186,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003841761818936122,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21433429569005966,
      "num_tokens": 68800370.0,
      "step": 30820
    },
    {
      "entropy": 5.4621867656707765,
      "epoch": 3.4619531644858763,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000384141074599337,
      "loss": 5.1132,
      "mean_token_accuracy": 0.20136052817106248,
      "num_tokens": 68811644.0,
      "step": 30825
    },
    {
      "entropy": 5.357368898391724,
      "epoch": 3.462514741393834,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038410596382996764,
      "loss": 4.973,
      "mean_token_accuracy": 0.20293720811605453,
      "num_tokens": 68823672.0,
      "step": 30830
    },
    {
      "entropy": 5.340794944763184,
      "epoch": 3.463076318301791,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003840708495866221,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2077446848154068,
      "num_tokens": 68835340.0,
      "step": 30835
    },
    {
      "entropy": 5.365924119949341,
      "epoch": 3.463637895209749,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003840357318704183,
      "loss": 4.9321,
      "mean_token_accuracy": 0.2202227607369423,
      "num_tokens": 68845903.0,
      "step": 30840
    },
    {
      "entropy": 5.326478052139282,
      "epoch": 3.4641994721177065,
      "grad_norm": 1.0,
      "learning_rate": 0.0003840006106824746,
      "loss": 5.0294,
      "mean_token_accuracy": 0.20978325009346008,
      "num_tokens": 68857287.0,
      "step": 30845
    },
    {
      "entropy": 5.386662912368775,
      "epoch": 3.464761049025664,
      "grad_norm": 1.125,
      "learning_rate": 0.0003839654860239091,
      "loss": 5.1201,
      "mean_token_accuracy": 0.20770289450883866,
      "num_tokens": 68868961.0,
      "step": 30850
    },
    {
      "entropy": 5.458471488952637,
      "epoch": 3.465322625933622,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003839303578958404,
      "loss": 5.0196,
      "mean_token_accuracy": 0.21429990977048874,
      "num_tokens": 68880275.0,
      "step": 30855
    },
    {
      "entropy": 5.416710424423218,
      "epoch": 3.465884202841579,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038389522629938676,
      "loss": 4.9875,
      "mean_token_accuracy": 0.22419403195381166,
      "num_tokens": 68891445.0,
      "step": 30860
    },
    {
      "entropy": 5.293928384780884,
      "epoch": 3.4664457797495367,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003838600912356669,
      "loss": 4.9709,
      "mean_token_accuracy": 0.21458418816328048,
      "num_tokens": 68903591.0,
      "step": 30865
    },
    {
      "entropy": 5.3166327476501465,
      "epoch": 3.4670073566574944,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003838249527057995,
      "loss": 5.013,
      "mean_token_accuracy": 0.21558609157800673,
      "num_tokens": 68914094.0,
      "step": 30870
    },
    {
      "entropy": 5.420800638198853,
      "epoch": 3.4675689335654516,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038378981071090346,
      "loss": 5.0751,
      "mean_token_accuracy": 0.21663724929094313,
      "num_tokens": 68925929.0,
      "step": 30875
    },
    {
      "entropy": 5.348216390609741,
      "epoch": 3.4681305104734093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038375466525209764,
      "loss": 4.967,
      "mean_token_accuracy": 0.2129657134413719,
      "num_tokens": 68936087.0,
      "step": 30880
    },
    {
      "entropy": 5.289228773117065,
      "epoch": 3.468692087381367,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038371951633050106,
      "loss": 4.9984,
      "mean_token_accuracy": 0.22223833948373795,
      "num_tokens": 68946878.0,
      "step": 30885
    },
    {
      "entropy": 5.411066198348999,
      "epoch": 3.4692536642893246,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038368436394723294,
      "loss": 5.0329,
      "mean_token_accuracy": 0.20940038859844207,
      "num_tokens": 68957896.0,
      "step": 30890
    },
    {
      "entropy": 5.3887049674987795,
      "epoch": 3.469815241197282,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038364920810341255,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21351155191659926,
      "num_tokens": 68969173.0,
      "step": 30895
    },
    {
      "entropy": 5.4095831394195555,
      "epoch": 3.4703768181052395,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003836140488001592,
      "loss": 5.0905,
      "mean_token_accuracy": 0.21897031515836715,
      "num_tokens": 68980537.0,
      "step": 30900
    },
    {
      "entropy": 5.362191581726075,
      "epoch": 3.470938395013197,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038357888603859236,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2169913485646248,
      "num_tokens": 68991736.0,
      "step": 30905
    },
    {
      "entropy": 5.349088573455811,
      "epoch": 3.471499971921155,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038354371981983175,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21907387673854828,
      "num_tokens": 69002316.0,
      "step": 30910
    },
    {
      "entropy": 5.26616415977478,
      "epoch": 3.472061548829112,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003835085501449969,
      "loss": 5.0079,
      "mean_token_accuracy": 0.21120173782110213,
      "num_tokens": 69014204.0,
      "step": 30915
    },
    {
      "entropy": 5.363573169708252,
      "epoch": 3.4726231257370697,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038347337701520785,
      "loss": 5.0541,
      "mean_token_accuracy": 0.20957633405923842,
      "num_tokens": 69026482.0,
      "step": 30920
    },
    {
      "entropy": 5.465552902221679,
      "epoch": 3.4731847026450273,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038343820043158436,
      "loss": 5.0244,
      "mean_token_accuracy": 0.213612861931324,
      "num_tokens": 69036527.0,
      "step": 30925
    },
    {
      "entropy": 5.422449016571045,
      "epoch": 3.4737462795529845,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003834030203952465,
      "loss": 5.038,
      "mean_token_accuracy": 0.21499565839767457,
      "num_tokens": 69047684.0,
      "step": 30930
    },
    {
      "entropy": 5.341845321655273,
      "epoch": 3.474307856460942,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003833678369073144,
      "loss": 5.0412,
      "mean_token_accuracy": 0.21146660894155503,
      "num_tokens": 69059236.0,
      "step": 30935
    },
    {
      "entropy": 5.282328987121582,
      "epoch": 3.4748694333689,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003833326499689084,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2204239174723625,
      "num_tokens": 69069713.0,
      "step": 30940
    },
    {
      "entropy": 5.339412117004395,
      "epoch": 3.4754310102768575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003832974595811488,
      "loss": 5.0618,
      "mean_token_accuracy": 0.20588571578264236,
      "num_tokens": 69079046.0,
      "step": 30945
    },
    {
      "entropy": 5.43354229927063,
      "epoch": 3.475992587184815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000383262265745156,
      "loss": 5.0681,
      "mean_token_accuracy": 0.21365218609571457,
      "num_tokens": 69090080.0,
      "step": 30950
    },
    {
      "entropy": 5.443773317337036,
      "epoch": 3.4765541640927724,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038322706846205073,
      "loss": 5.1112,
      "mean_token_accuracy": 0.2056988850235939,
      "num_tokens": 69102374.0,
      "step": 30955
    },
    {
      "entropy": 5.377245950698852,
      "epoch": 3.47711574100073,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038319186773295364,
      "loss": 5.0459,
      "mean_token_accuracy": 0.21622643768787383,
      "num_tokens": 69112688.0,
      "step": 30960
    },
    {
      "entropy": 5.395568370819092,
      "epoch": 3.4776773179086877,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003831566635589856,
      "loss": 5.0399,
      "mean_token_accuracy": 0.2145000770688057,
      "num_tokens": 69124861.0,
      "step": 30965
    },
    {
      "entropy": 5.398483419418335,
      "epoch": 3.478238894816645,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003831214559412674,
      "loss": 5.0931,
      "mean_token_accuracy": 0.2097925901412964,
      "num_tokens": 69135898.0,
      "step": 30970
    },
    {
      "entropy": 5.461172676086425,
      "epoch": 3.4788004717246026,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038308624488092005,
      "loss": 5.1255,
      "mean_token_accuracy": 0.20694369971752166,
      "num_tokens": 69146692.0,
      "step": 30975
    },
    {
      "entropy": 5.357588052749634,
      "epoch": 3.4793620486325603,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003830510303790647,
      "loss": 5.0044,
      "mean_token_accuracy": 0.21286374032497407,
      "num_tokens": 69157908.0,
      "step": 30980
    },
    {
      "entropy": 5.365853214263916,
      "epoch": 3.479923625540518,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038301581243682263,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21555214375257492,
      "num_tokens": 69168596.0,
      "step": 30985
    },
    {
      "entropy": 5.283097982406616,
      "epoch": 3.480485202448475,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003829805910553153,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21930192559957504,
      "num_tokens": 69179261.0,
      "step": 30990
    },
    {
      "entropy": 5.3252404689788815,
      "epoch": 3.481046779356433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038294536623566407,
      "loss": 4.9674,
      "mean_token_accuracy": 0.21390282213687897,
      "num_tokens": 69190410.0,
      "step": 30995
    },
    {
      "entropy": 5.393063974380493,
      "epoch": 3.4816083562643905,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003829101379789905,
      "loss": 5.1218,
      "mean_token_accuracy": 0.21277871429920198,
      "num_tokens": 69201114.0,
      "step": 31000
    },
    {
      "entropy": 5.39140510559082,
      "epoch": 3.482169933172348,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003828749062864162,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21919576078653336,
      "num_tokens": 69212240.0,
      "step": 31005
    },
    {
      "entropy": 5.353231573104859,
      "epoch": 3.4827315100803053,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003828396711590631,
      "loss": 5.0702,
      "mean_token_accuracy": 0.2093835160136223,
      "num_tokens": 69223714.0,
      "step": 31010
    },
    {
      "entropy": 5.387914657592773,
      "epoch": 3.483293086988263,
      "grad_norm": 1.0,
      "learning_rate": 0.00038280443259805307,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21373282223939896,
      "num_tokens": 69234364.0,
      "step": 31015
    },
    {
      "entropy": 5.307966423034668,
      "epoch": 3.4838546638962207,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003827691906045081,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2189030811190605,
      "num_tokens": 69245584.0,
      "step": 31020
    },
    {
      "entropy": 5.355924367904663,
      "epoch": 3.484416240804178,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003827339451795503,
      "loss": 4.9703,
      "mean_token_accuracy": 0.21119096279144287,
      "num_tokens": 69256891.0,
      "step": 31025
    },
    {
      "entropy": 5.413004207611084,
      "epoch": 3.4849778177121356,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003826986963243018,
      "loss": 4.9664,
      "mean_token_accuracy": 0.21722579896450042,
      "num_tokens": 69266972.0,
      "step": 31030
    },
    {
      "entropy": 5.391071033477783,
      "epoch": 3.485539394620093,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003826634440398852,
      "loss": 5.0282,
      "mean_token_accuracy": 0.22355865836143493,
      "num_tokens": 69277993.0,
      "step": 31035
    },
    {
      "entropy": 5.2843985080719,
      "epoch": 3.486100971528051,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038262818832742276,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21453755348920822,
      "num_tokens": 69289720.0,
      "step": 31040
    },
    {
      "entropy": 5.406031894683838,
      "epoch": 3.4866625484360085,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038259292918803696,
      "loss": 5.1205,
      "mean_token_accuracy": 0.20881295055150986,
      "num_tokens": 69301169.0,
      "step": 31045
    },
    {
      "entropy": 5.354175233840943,
      "epoch": 3.4872241253439658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003825576666228506,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2143484726548195,
      "num_tokens": 69311443.0,
      "step": 31050
    },
    {
      "entropy": 5.401350593566894,
      "epoch": 3.4877857022519234,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003825224006329864,
      "loss": 5.0395,
      "mean_token_accuracy": 0.2131688892841339,
      "num_tokens": 69322035.0,
      "step": 31055
    },
    {
      "entropy": 5.290612077713012,
      "epoch": 3.488347279159881,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003824871312195673,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2300107479095459,
      "num_tokens": 69331955.0,
      "step": 31060
    },
    {
      "entropy": 5.3263914585113525,
      "epoch": 3.4889088560678383,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003824518583837163,
      "loss": 5.0346,
      "mean_token_accuracy": 0.21355758756399154,
      "num_tokens": 69343897.0,
      "step": 31065
    },
    {
      "entropy": 5.281171274185181,
      "epoch": 3.489470432975796,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003824165821265564,
      "loss": 4.9047,
      "mean_token_accuracy": 0.22356727123260497,
      "num_tokens": 69353886.0,
      "step": 31070
    },
    {
      "entropy": 5.245556783676148,
      "epoch": 3.4900320098837536,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003823813024492108,
      "loss": 4.9654,
      "mean_token_accuracy": 0.22017951756715776,
      "num_tokens": 69365328.0,
      "step": 31075
    },
    {
      "entropy": 5.390545606613159,
      "epoch": 3.4905935867917113,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000382346019352803,
      "loss": 5.1059,
      "mean_token_accuracy": 0.20603012293577194,
      "num_tokens": 69376007.0,
      "step": 31080
    },
    {
      "entropy": 5.469788026809693,
      "epoch": 3.4911551636996685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003823107328384562,
      "loss": 5.0947,
      "mean_token_accuracy": 0.20948586463928223,
      "num_tokens": 69386668.0,
      "step": 31085
    },
    {
      "entropy": 5.3803362369537355,
      "epoch": 3.491716740607626,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003822754429072942,
      "loss": 4.8778,
      "mean_token_accuracy": 0.2333522379398346,
      "num_tokens": 69397827.0,
      "step": 31090
    },
    {
      "entropy": 5.314143896102905,
      "epoch": 3.492278317515584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003822401495604403,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21945687830448152,
      "num_tokens": 69409268.0,
      "step": 31095
    },
    {
      "entropy": 5.335787773132324,
      "epoch": 3.4928398944235415,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038220485279901855,
      "loss": 5.0006,
      "mean_token_accuracy": 0.22070402055978774,
      "num_tokens": 69420995.0,
      "step": 31100
    },
    {
      "entropy": 5.400365257263184,
      "epoch": 3.4934014713314987,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038216955262415273,
      "loss": 4.9981,
      "mean_token_accuracy": 0.2161305159330368,
      "num_tokens": 69432919.0,
      "step": 31105
    },
    {
      "entropy": 5.367403268814087,
      "epoch": 3.4939630482394564,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038213424903696673,
      "loss": 5.0453,
      "mean_token_accuracy": 0.214018876850605,
      "num_tokens": 69443321.0,
      "step": 31110
    },
    {
      "entropy": 5.357817459106445,
      "epoch": 3.494524625147414,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038209894203858477,
      "loss": 4.9952,
      "mean_token_accuracy": 0.21759818643331527,
      "num_tokens": 69454695.0,
      "step": 31115
    },
    {
      "entropy": 5.381619882583618,
      "epoch": 3.4950862020553712,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038206363163013093,
      "loss": 4.94,
      "mean_token_accuracy": 0.21969179511070253,
      "num_tokens": 69465617.0,
      "step": 31120
    },
    {
      "entropy": 5.286784315109253,
      "epoch": 3.495647778963329,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038202831781272944,
      "loss": 5.0114,
      "mean_token_accuracy": 0.21283484995365143,
      "num_tokens": 69476730.0,
      "step": 31125
    },
    {
      "entropy": 5.299883842468262,
      "epoch": 3.4962093558712866,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038199300058750494,
      "loss": 4.9772,
      "mean_token_accuracy": 0.21471781134605408,
      "num_tokens": 69488488.0,
      "step": 31130
    },
    {
      "entropy": 5.43042221069336,
      "epoch": 3.4967709327792442,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038195767995558174,
      "loss": 5.072,
      "mean_token_accuracy": 0.20745187401771545,
      "num_tokens": 69498718.0,
      "step": 31135
    },
    {
      "entropy": 5.365119218826294,
      "epoch": 3.497332509687202,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003819223559180845,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21384021043777465,
      "num_tokens": 69509567.0,
      "step": 31140
    },
    {
      "entropy": 5.332442760467529,
      "epoch": 3.497894086595159,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000381887028476138,
      "loss": 4.9911,
      "mean_token_accuracy": 0.214080348610878,
      "num_tokens": 69519392.0,
      "step": 31145
    },
    {
      "entropy": 5.458573579788208,
      "epoch": 3.4984556635031168,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00038185169763086703,
      "loss": 5.1246,
      "mean_token_accuracy": 0.20100812315940858,
      "num_tokens": 69531228.0,
      "step": 31150
    },
    {
      "entropy": 5.468828392028809,
      "epoch": 3.4990172404110744,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003818163633833966,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21894100308418274,
      "num_tokens": 69542888.0,
      "step": 31155
    },
    {
      "entropy": 5.287666082382202,
      "epoch": 3.4995788173190316,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038178102573485173,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21915661692619323,
      "num_tokens": 69553481.0,
      "step": 31160
    },
    {
      "entropy": 5.327647924423218,
      "epoch": 3.5001403942269893,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003817456846863575,
      "loss": 5.0158,
      "mean_token_accuracy": 0.21329982131719588,
      "num_tokens": 69565314.0,
      "step": 31165
    },
    {
      "entropy": 5.347225761413574,
      "epoch": 3.500701971134947,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038171034023903937,
      "loss": 4.9859,
      "mean_token_accuracy": 0.20576197206974028,
      "num_tokens": 69575807.0,
      "step": 31170
    },
    {
      "entropy": 5.406551647186279,
      "epoch": 3.5012635480429046,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003816749923940226,
      "loss": 5.1235,
      "mean_token_accuracy": 0.21208667755126953,
      "num_tokens": 69587873.0,
      "step": 31175
    },
    {
      "entropy": 5.397897100448608,
      "epoch": 3.5018251249508623,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038163964115243276,
      "loss": 5.0683,
      "mean_token_accuracy": 0.2172987312078476,
      "num_tokens": 69598015.0,
      "step": 31180
    },
    {
      "entropy": 5.374255084991455,
      "epoch": 3.5023867018588195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038160428651539535,
      "loss": 5.0238,
      "mean_token_accuracy": 0.21130315214395523,
      "num_tokens": 69609345.0,
      "step": 31185
    },
    {
      "entropy": 5.367508268356323,
      "epoch": 3.502948278766777,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038156892848403606,
      "loss": 5.018,
      "mean_token_accuracy": 0.20179952979087828,
      "num_tokens": 69620075.0,
      "step": 31190
    },
    {
      "entropy": 5.377548456192017,
      "epoch": 3.503509855674735,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003815335670594808,
      "loss": 5.0179,
      "mean_token_accuracy": 0.21915940791368485,
      "num_tokens": 69630825.0,
      "step": 31195
    },
    {
      "entropy": 5.455663013458252,
      "epoch": 3.504071432582692,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038149820224285546,
      "loss": 5.1111,
      "mean_token_accuracy": 0.21145035326480865,
      "num_tokens": 69642415.0,
      "step": 31200
    },
    {
      "entropy": 5.253020429611206,
      "epoch": 3.5046330094906497,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000381462834035286,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22187747657299042,
      "num_tokens": 69652735.0,
      "step": 31205
    },
    {
      "entropy": 5.280878353118896,
      "epoch": 3.5051945863986074,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003814274624378986,
      "loss": 4.9987,
      "mean_token_accuracy": 0.22689505964517592,
      "num_tokens": 69664312.0,
      "step": 31210
    },
    {
      "entropy": 5.328192043304443,
      "epoch": 3.5057561633065646,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003813920874518196,
      "loss": 4.9702,
      "mean_token_accuracy": 0.21368768215179443,
      "num_tokens": 69674570.0,
      "step": 31215
    },
    {
      "entropy": 5.338263988494873,
      "epoch": 3.5063177402145222,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003813567090781752,
      "loss": 5.0171,
      "mean_token_accuracy": 0.21209612339735032,
      "num_tokens": 69686591.0,
      "step": 31220
    },
    {
      "entropy": 5.41046986579895,
      "epoch": 3.50687931712248,
      "grad_norm": 1.046875,
      "learning_rate": 0.000381321327318092,
      "loss": 5.0454,
      "mean_token_accuracy": 0.2165257900953293,
      "num_tokens": 69697104.0,
      "step": 31225
    },
    {
      "entropy": 5.413427734375,
      "epoch": 3.5074408940304376,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003812859421726964,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21876952648162842,
      "num_tokens": 69708827.0,
      "step": 31230
    },
    {
      "entropy": 5.336478424072266,
      "epoch": 3.5080024709383952,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038125055364311526,
      "loss": 5.0428,
      "mean_token_accuracy": 0.22015126943588256,
      "num_tokens": 69720426.0,
      "step": 31235
    },
    {
      "entropy": 5.490678405761718,
      "epoch": 3.5085640478463525,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003812151617304753,
      "loss": 5.238,
      "mean_token_accuracy": 0.20407590866088868,
      "num_tokens": 69733199.0,
      "step": 31240
    },
    {
      "entropy": 5.306839466094971,
      "epoch": 3.50912562475431,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003811797664359033,
      "loss": 4.8677,
      "mean_token_accuracy": 0.22371850162744522,
      "num_tokens": 69744322.0,
      "step": 31245
    },
    {
      "entropy": 5.2996506690979,
      "epoch": 3.5096872016622678,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038114436776052644,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21332952827215196,
      "num_tokens": 69755662.0,
      "step": 31250
    },
    {
      "entropy": 5.262163591384888,
      "epoch": 3.510248778570225,
      "grad_norm": 1.0,
      "learning_rate": 0.0003811089657054717,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22302572429180145,
      "num_tokens": 69765663.0,
      "step": 31255
    },
    {
      "entropy": 5.371672630310059,
      "epoch": 3.5108103554781827,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003810735602718664,
      "loss": 5.1795,
      "mean_token_accuracy": 0.20814565420150757,
      "num_tokens": 69777483.0,
      "step": 31260
    },
    {
      "entropy": 5.44297604560852,
      "epoch": 3.5113719323861403,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003810381514608378,
      "loss": 5.0628,
      "mean_token_accuracy": 0.20481681525707246,
      "num_tokens": 69789034.0,
      "step": 31265
    },
    {
      "entropy": 5.413743019104004,
      "epoch": 3.511933509294098,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003810027392735133,
      "loss": 5.0267,
      "mean_token_accuracy": 0.21109580099582673,
      "num_tokens": 69801288.0,
      "step": 31270
    },
    {
      "entropy": 5.316257953643799,
      "epoch": 3.5124950862020556,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003809673237110204,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21678188145160676,
      "num_tokens": 69811063.0,
      "step": 31275
    },
    {
      "entropy": 5.31076078414917,
      "epoch": 3.513056663110013,
      "grad_norm": 1.0,
      "learning_rate": 0.0003809319047744869,
      "loss": 4.9917,
      "mean_token_accuracy": 0.21564977169036864,
      "num_tokens": 69822572.0,
      "step": 31280
    },
    {
      "entropy": 5.259940481185913,
      "epoch": 3.5136182400179705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038089648246504045,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21923152059316636,
      "num_tokens": 69834082.0,
      "step": 31285
    },
    {
      "entropy": 5.354159259796143,
      "epoch": 3.514179816925928,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038086105678380896,
      "loss": 5.006,
      "mean_token_accuracy": 0.21684324592351914,
      "num_tokens": 69844560.0,
      "step": 31290
    },
    {
      "entropy": 5.391580200195312,
      "epoch": 3.5147413938338854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038082562773192026,
      "loss": 4.9917,
      "mean_token_accuracy": 0.21378025263547898,
      "num_tokens": 69854993.0,
      "step": 31295
    },
    {
      "entropy": 5.347278213500976,
      "epoch": 3.515302970741843,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038079019531050257,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22233842611312865,
      "num_tokens": 69866074.0,
      "step": 31300
    },
    {
      "entropy": 5.367187547683716,
      "epoch": 3.5158645476498007,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038075475952068405,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21382921934127808,
      "num_tokens": 69877101.0,
      "step": 31305
    },
    {
      "entropy": 5.298641586303711,
      "epoch": 3.516426124557758,
      "grad_norm": 1.015625,
      "learning_rate": 0.000380719320363593,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2252766966819763,
      "num_tokens": 69887388.0,
      "step": 31310
    },
    {
      "entropy": 5.388827180862426,
      "epoch": 3.5169877014657156,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038068387784035774,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21979601979255675,
      "num_tokens": 69898242.0,
      "step": 31315
    },
    {
      "entropy": 5.3530683517456055,
      "epoch": 3.5175492783736733,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003806484319521069,
      "loss": 5.0355,
      "mean_token_accuracy": 0.2166160076856613,
      "num_tokens": 69909645.0,
      "step": 31320
    },
    {
      "entropy": 5.388315057754516,
      "epoch": 3.518110855281631,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038061298269996893,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21466407626867295,
      "num_tokens": 69920862.0,
      "step": 31325
    },
    {
      "entropy": 5.364881181716919,
      "epoch": 3.5186724321895886,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003805775300850727,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2209607869386673,
      "num_tokens": 69932219.0,
      "step": 31330
    },
    {
      "entropy": 5.39084644317627,
      "epoch": 3.519234009097546,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003805420741085469,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2135359063744545,
      "num_tokens": 69942642.0,
      "step": 31335
    },
    {
      "entropy": 5.405378293991089,
      "epoch": 3.5197955860055035,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003805066147715205,
      "loss": 5.1613,
      "mean_token_accuracy": 0.19321514368057252,
      "num_tokens": 69954336.0,
      "step": 31340
    },
    {
      "entropy": 5.3600507259368895,
      "epoch": 3.520357162913461,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003804711520751226,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21729453057050704,
      "num_tokens": 69965716.0,
      "step": 31345
    },
    {
      "entropy": 5.357483959197998,
      "epoch": 3.5209187398214183,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038043568602048227,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21355781108140945,
      "num_tokens": 69976545.0,
      "step": 31350
    },
    {
      "entropy": 5.382663345336914,
      "epoch": 3.521480316729376,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003804002166087289,
      "loss": 5.0422,
      "mean_token_accuracy": 0.22126413136720657,
      "num_tokens": 69987846.0,
      "step": 31355
    },
    {
      "entropy": 5.333971261978149,
      "epoch": 3.5220418936373337,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038036474384099174,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2196497306227684,
      "num_tokens": 69998089.0,
      "step": 31360
    },
    {
      "entropy": 5.3745533466339115,
      "epoch": 3.5226034705452913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003803292677184002,
      "loss": 5.0873,
      "mean_token_accuracy": 0.21071239411830903,
      "num_tokens": 70008900.0,
      "step": 31365
    },
    {
      "entropy": 5.427185201644898,
      "epoch": 3.523165047453249,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038029378824208394,
      "loss": 5.0595,
      "mean_token_accuracy": 0.21183212101459503,
      "num_tokens": 70019675.0,
      "step": 31370
    },
    {
      "entropy": 5.459985446929932,
      "epoch": 3.523726624361206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003802583054131727,
      "loss": 5.1331,
      "mean_token_accuracy": 0.2070818990468979,
      "num_tokens": 70031077.0,
      "step": 31375
    },
    {
      "entropy": 5.411845541000366,
      "epoch": 3.524288201269164,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003802228192327961,
      "loss": 5.138,
      "mean_token_accuracy": 0.1989188551902771,
      "num_tokens": 70044569.0,
      "step": 31380
    },
    {
      "entropy": 5.3514093399047855,
      "epoch": 3.5248497781771215,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038018732970208414,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2166432872414589,
      "num_tokens": 70056396.0,
      "step": 31385
    },
    {
      "entropy": 5.488380432128906,
      "epoch": 3.5254113550850787,
      "grad_norm": 1.0,
      "learning_rate": 0.0003801518368221668,
      "loss": 5.2045,
      "mean_token_accuracy": 0.2093486249446869,
      "num_tokens": 70068341.0,
      "step": 31390
    },
    {
      "entropy": 5.379701948165893,
      "epoch": 3.5259729319930364,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003801163405941743,
      "loss": 4.9166,
      "mean_token_accuracy": 0.21751874983310698,
      "num_tokens": 70079069.0,
      "step": 31395
    },
    {
      "entropy": 5.396098852157593,
      "epoch": 3.526534508900994,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038008084101923665,
      "loss": 5.0926,
      "mean_token_accuracy": 0.21362614035606384,
      "num_tokens": 70090274.0,
      "step": 31400
    },
    {
      "entropy": 5.311895418167114,
      "epoch": 3.5270960858089513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038004533809848435,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21332114189863205,
      "num_tokens": 70101144.0,
      "step": 31405
    },
    {
      "entropy": 5.371632194519043,
      "epoch": 3.527657662716909,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038000983183304765,
      "loss": 5.014,
      "mean_token_accuracy": 0.21538894027471542,
      "num_tokens": 70112210.0,
      "step": 31410
    },
    {
      "entropy": 5.445079135894775,
      "epoch": 3.5282192396248666,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003799743222240572,
      "loss": 5.1255,
      "mean_token_accuracy": 0.21043878197669982,
      "num_tokens": 70124302.0,
      "step": 31415
    },
    {
      "entropy": 5.451379823684692,
      "epoch": 3.5287808165328243,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003799388092726436,
      "loss": 5.1045,
      "mean_token_accuracy": 0.21609489172697066,
      "num_tokens": 70136026.0,
      "step": 31420
    },
    {
      "entropy": 5.41684889793396,
      "epoch": 3.529342393440782,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037990329297993763,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21301512718200682,
      "num_tokens": 70147439.0,
      "step": 31425
    },
    {
      "entropy": 5.282562732696533,
      "epoch": 3.529903970348739,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003798677733470701,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22608915865421295,
      "num_tokens": 70158367.0,
      "step": 31430
    },
    {
      "entropy": 5.312399196624756,
      "epoch": 3.530465547256697,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037983225037517207,
      "loss": 5.1046,
      "mean_token_accuracy": 0.2103608563542366,
      "num_tokens": 70169226.0,
      "step": 31435
    },
    {
      "entropy": 5.42043285369873,
      "epoch": 3.5310271241646545,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037979672406537445,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21405208855867386,
      "num_tokens": 70179440.0,
      "step": 31440
    },
    {
      "entropy": 5.431687736511231,
      "epoch": 3.5315887010726117,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003797611944188086,
      "loss": 5.0631,
      "mean_token_accuracy": 0.21415450870990754,
      "num_tokens": 70191160.0,
      "step": 31445
    },
    {
      "entropy": 5.383655405044555,
      "epoch": 3.5321502779805694,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037972566143660566,
      "loss": 5.084,
      "mean_token_accuracy": 0.21398237198591233,
      "num_tokens": 70202143.0,
      "step": 31450
    },
    {
      "entropy": 5.378739500045777,
      "epoch": 3.532711854888527,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003796901251198969,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21327701658010484,
      "num_tokens": 70213957.0,
      "step": 31455
    },
    {
      "entropy": 5.39954137802124,
      "epoch": 3.5332734317964847,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000379654585469814,
      "loss": 5.012,
      "mean_token_accuracy": 0.21656121164560319,
      "num_tokens": 70224153.0,
      "step": 31460
    },
    {
      "entropy": 5.345164346694946,
      "epoch": 3.5338350087044423,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003796190424874886,
      "loss": 5.0385,
      "mean_token_accuracy": 0.21122560948133468,
      "num_tokens": 70236093.0,
      "step": 31465
    },
    {
      "entropy": 5.340242719650268,
      "epoch": 3.5343965856123996,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003795834961740523,
      "loss": 5.0377,
      "mean_token_accuracy": 0.21832483261823654,
      "num_tokens": 70246865.0,
      "step": 31470
    },
    {
      "entropy": 5.391931009292603,
      "epoch": 3.534958162520357,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037954794653063695,
      "loss": 5.0874,
      "mean_token_accuracy": 0.20533659011125566,
      "num_tokens": 70258601.0,
      "step": 31475
    },
    {
      "entropy": 5.382684898376465,
      "epoch": 3.535519739428315,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003795123935583743,
      "loss": 5.0078,
      "mean_token_accuracy": 0.213176029920578,
      "num_tokens": 70269020.0,
      "step": 31480
    },
    {
      "entropy": 5.366488027572632,
      "epoch": 3.536081316336272,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003794768372583966,
      "loss": 5.0475,
      "mean_token_accuracy": 0.2061367526650429,
      "num_tokens": 70280582.0,
      "step": 31485
    },
    {
      "entropy": 5.484293794631958,
      "epoch": 3.5366428932442298,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037944127763183587,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21495116055011748,
      "num_tokens": 70292848.0,
      "step": 31490
    },
    {
      "entropy": 5.4457344055175785,
      "epoch": 3.5372044701521874,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037940571467982436,
      "loss": 5.0931,
      "mean_token_accuracy": 0.2139238476753235,
      "num_tokens": 70304892.0,
      "step": 31495
    },
    {
      "entropy": 5.281602478027343,
      "epoch": 3.5377660470601446,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003793701484034944,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22130076885223388,
      "num_tokens": 70316244.0,
      "step": 31500
    },
    {
      "entropy": 5.390929555892944,
      "epoch": 3.5383276239681023,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037933457880397843,
      "loss": 5.0354,
      "mean_token_accuracy": 0.2093072548508644,
      "num_tokens": 70327037.0,
      "step": 31505
    },
    {
      "entropy": 5.369195508956909,
      "epoch": 3.53888920087606,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037929900588240897,
      "loss": 4.924,
      "mean_token_accuracy": 0.22150439471006395,
      "num_tokens": 70338587.0,
      "step": 31510
    },
    {
      "entropy": 5.24082350730896,
      "epoch": 3.5394507777840176,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003792634296399187,
      "loss": 5.0395,
      "mean_token_accuracy": 0.21533430814743043,
      "num_tokens": 70349930.0,
      "step": 31515
    },
    {
      "entropy": 5.245293283462525,
      "epoch": 3.5400123546919753,
      "grad_norm": 1.0,
      "learning_rate": 0.00037922785007764043,
      "loss": 4.8719,
      "mean_token_accuracy": 0.22105963975191117,
      "num_tokens": 70360880.0,
      "step": 31520
    },
    {
      "entropy": 5.480115270614624,
      "epoch": 3.5405739315999325,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000379192267196707,
      "loss": 5.1454,
      "mean_token_accuracy": 0.20767911821603774,
      "num_tokens": 70372508.0,
      "step": 31525
    },
    {
      "entropy": 5.430534839630127,
      "epoch": 3.54113550850789,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037915668099825135,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21408879905939102,
      "num_tokens": 70383960.0,
      "step": 31530
    },
    {
      "entropy": 5.370037269592285,
      "epoch": 3.541697085415848,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037912109148340663,
      "loss": 5.0359,
      "mean_token_accuracy": 0.2127966210246086,
      "num_tokens": 70395881.0,
      "step": 31535
    },
    {
      "entropy": 5.343982410430908,
      "epoch": 3.542258662323805,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037908549865330596,
      "loss": 5.04,
      "mean_token_accuracy": 0.2134324550628662,
      "num_tokens": 70407590.0,
      "step": 31540
    },
    {
      "entropy": 5.37364935874939,
      "epoch": 3.5428202392317627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037904990250908256,
      "loss": 4.9984,
      "mean_token_accuracy": 0.22460844218730927,
      "num_tokens": 70418078.0,
      "step": 31545
    },
    {
      "entropy": 5.354344701766967,
      "epoch": 3.5433818161397204,
      "grad_norm": 1.0,
      "learning_rate": 0.00037901430305187,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21535616368055344,
      "num_tokens": 70428292.0,
      "step": 31550
    },
    {
      "entropy": 5.2816938877105715,
      "epoch": 3.543943393047678,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037897870028280167,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21989400535821915,
      "num_tokens": 70438481.0,
      "step": 31555
    },
    {
      "entropy": 5.343485403060913,
      "epoch": 3.5445049699556357,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003789430942030112,
      "loss": 5.0762,
      "mean_token_accuracy": 0.2078214168548584,
      "num_tokens": 70449949.0,
      "step": 31560
    },
    {
      "entropy": 5.461041975021362,
      "epoch": 3.545066546863593,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037890748481363216,
      "loss": 5.1169,
      "mean_token_accuracy": 0.21154679358005524,
      "num_tokens": 70461159.0,
      "step": 31565
    },
    {
      "entropy": 5.359609174728393,
      "epoch": 3.5456281237715506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003788718721157987,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21476232558488845,
      "num_tokens": 70471212.0,
      "step": 31570
    },
    {
      "entropy": 5.342015075683594,
      "epoch": 3.5461897006795082,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003788362561106445,
      "loss": 5.0157,
      "mean_token_accuracy": 0.2186439484357834,
      "num_tokens": 70481587.0,
      "step": 31575
    },
    {
      "entropy": 5.449720811843872,
      "epoch": 3.5467512775874654,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037880063679930355,
      "loss": 5.0867,
      "mean_token_accuracy": 0.21345228403806688,
      "num_tokens": 70491952.0,
      "step": 31580
    },
    {
      "entropy": 5.388182783126831,
      "epoch": 3.547312854495423,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037876501418291015,
      "loss": 4.9998,
      "mean_token_accuracy": 0.20984189808368683,
      "num_tokens": 70503279.0,
      "step": 31585
    },
    {
      "entropy": 5.328646326065064,
      "epoch": 3.5478744314033808,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003787293882625984,
      "loss": 5.0714,
      "mean_token_accuracy": 0.21108720749616622,
      "num_tokens": 70514779.0,
      "step": 31590
    },
    {
      "entropy": 5.319090270996094,
      "epoch": 3.548436008311338,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003786937590395027,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2135763645172119,
      "num_tokens": 70526478.0,
      "step": 31595
    },
    {
      "entropy": 5.385490369796753,
      "epoch": 3.5489975852192956,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003786581265147575,
      "loss": 5.0774,
      "mean_token_accuracy": 0.2109088957309723,
      "num_tokens": 70537293.0,
      "step": 31600
    },
    {
      "entropy": 5.440050983428955,
      "epoch": 3.5495591621272533,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003786224906894974,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21795266419649123,
      "num_tokens": 70547357.0,
      "step": 31605
    },
    {
      "entropy": 5.368253898620606,
      "epoch": 3.550120739035211,
      "grad_norm": 0.96875,
      "learning_rate": 0.000378586851564857,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2081439256668091,
      "num_tokens": 70558102.0,
      "step": 31610
    },
    {
      "entropy": 5.408520889282227,
      "epoch": 3.5506823159431686,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000378551209141971,
      "loss": 5.076,
      "mean_token_accuracy": 0.2183656558394432,
      "num_tokens": 70569454.0,
      "step": 31615
    },
    {
      "entropy": 5.394662284851075,
      "epoch": 3.551243892851126,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003785155634219744,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21212248802185057,
      "num_tokens": 70580590.0,
      "step": 31620
    },
    {
      "entropy": 5.317335700988769,
      "epoch": 3.5518054697590835,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003784799144060021,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2160568043589592,
      "num_tokens": 70591233.0,
      "step": 31625
    },
    {
      "entropy": 5.36469292640686,
      "epoch": 3.552367046667041,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037844426209518916,
      "loss": 5.0048,
      "mean_token_accuracy": 0.22044412642717362,
      "num_tokens": 70603029.0,
      "step": 31630
    },
    {
      "entropy": 5.48450984954834,
      "epoch": 3.5529286235749984,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037840860649067084,
      "loss": 5.1572,
      "mean_token_accuracy": 0.20457223206758499,
      "num_tokens": 70616021.0,
      "step": 31635
    },
    {
      "entropy": 5.3849289894104,
      "epoch": 3.553490200482956,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037837294759358235,
      "loss": 5.0275,
      "mean_token_accuracy": 0.22033473700284958,
      "num_tokens": 70627044.0,
      "step": 31640
    },
    {
      "entropy": 5.30291109085083,
      "epoch": 3.5540517773909137,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037833728540505914,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21616286933422088,
      "num_tokens": 70637435.0,
      "step": 31645
    },
    {
      "entropy": 5.373443365097046,
      "epoch": 3.5546133542988714,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003783016199262367,
      "loss": 5.0091,
      "mean_token_accuracy": 0.21726883798837662,
      "num_tokens": 70648106.0,
      "step": 31650
    },
    {
      "entropy": 5.319651985168457,
      "epoch": 3.555174931206829,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003782659511582505,
      "loss": 5.009,
      "mean_token_accuracy": 0.21413856446743013,
      "num_tokens": 70661242.0,
      "step": 31655
    },
    {
      "entropy": 5.318558692932129,
      "epoch": 3.5557365081147863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003782302791022365,
      "loss": 5.0428,
      "mean_token_accuracy": 0.2149222031235695,
      "num_tokens": 70672758.0,
      "step": 31660
    },
    {
      "entropy": 5.445973300933838,
      "epoch": 3.556298085022744,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037819460375933034,
      "loss": 5.0466,
      "mean_token_accuracy": 0.20979889929294587,
      "num_tokens": 70683365.0,
      "step": 31665
    },
    {
      "entropy": 5.318071460723877,
      "epoch": 3.5568596619307016,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003781589251306679,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22454005777835845,
      "num_tokens": 70693984.0,
      "step": 31670
    },
    {
      "entropy": 5.36376724243164,
      "epoch": 3.557421238838659,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003781232432173853,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2143866941332817,
      "num_tokens": 70704748.0,
      "step": 31675
    },
    {
      "entropy": 5.298790550231933,
      "epoch": 3.5579828157466165,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037808755802061874,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2203442022204399,
      "num_tokens": 70715415.0,
      "step": 31680
    },
    {
      "entropy": 5.323739862442016,
      "epoch": 3.558544392654574,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003780518695415043,
      "loss": 4.9995,
      "mean_token_accuracy": 0.21451320499181747,
      "num_tokens": 70726646.0,
      "step": 31685
    },
    {
      "entropy": 5.487664127349854,
      "epoch": 3.5591059695625313,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037801617778117835,
      "loss": 5.1887,
      "mean_token_accuracy": 0.1991393744945526,
      "num_tokens": 70737515.0,
      "step": 31690
    },
    {
      "entropy": 5.435110139846802,
      "epoch": 3.559667546470489,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037798048274077735,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2153919294476509,
      "num_tokens": 70747751.0,
      "step": 31695
    },
    {
      "entropy": 5.388730573654175,
      "epoch": 3.5602291233784467,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037794478442143786,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21912371814250947,
      "num_tokens": 70757951.0,
      "step": 31700
    },
    {
      "entropy": 5.400559806823731,
      "epoch": 3.5607907002864043,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00037790908282429647,
      "loss": 5.123,
      "mean_token_accuracy": 0.2003370389342308,
      "num_tokens": 70769515.0,
      "step": 31705
    },
    {
      "entropy": 5.400623083114624,
      "epoch": 3.561352277194362,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037787337795049,
      "loss": 5.041,
      "mean_token_accuracy": 0.21440519094467164,
      "num_tokens": 70780567.0,
      "step": 31710
    },
    {
      "entropy": 5.411202096939087,
      "epoch": 3.561913854102319,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037783766980115525,
      "loss": 5.129,
      "mean_token_accuracy": 0.2066635936498642,
      "num_tokens": 70793250.0,
      "step": 31715
    },
    {
      "entropy": 5.487799739837646,
      "epoch": 3.562475431010277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037780195837742927,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20818829536437988,
      "num_tokens": 70804579.0,
      "step": 31720
    },
    {
      "entropy": 5.377258586883545,
      "epoch": 3.5630370079182345,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037776624368044895,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21938524544239044,
      "num_tokens": 70815931.0,
      "step": 31725
    },
    {
      "entropy": 5.411744594573975,
      "epoch": 3.5635985848261917,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00037773052571135157,
      "loss": 5.1086,
      "mean_token_accuracy": 0.20783221423625947,
      "num_tokens": 70828058.0,
      "step": 31730
    },
    {
      "entropy": 5.434455156326294,
      "epoch": 3.5641601617341494,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003776948044712745,
      "loss": 5.0485,
      "mean_token_accuracy": 0.2146427810192108,
      "num_tokens": 70838177.0,
      "step": 31735
    },
    {
      "entropy": 5.364833116531372,
      "epoch": 3.564721738642107,
      "grad_norm": 0.921875,
      "learning_rate": 0.000377659079961355,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21907975822687148,
      "num_tokens": 70849888.0,
      "step": 31740
    },
    {
      "entropy": 5.353806734085083,
      "epoch": 3.5652833155500647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037762335218273057,
      "loss": 5.0161,
      "mean_token_accuracy": 0.21832160502672196,
      "num_tokens": 70859754.0,
      "step": 31745
    },
    {
      "entropy": 5.456119108200073,
      "epoch": 3.5658448924580224,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003775876211365388,
      "loss": 5.1045,
      "mean_token_accuracy": 0.21108708828687667,
      "num_tokens": 70870166.0,
      "step": 31750
    },
    {
      "entropy": 5.412425804138183,
      "epoch": 3.5664064693659796,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003775518868239174,
      "loss": 5.007,
      "mean_token_accuracy": 0.21662432253360747,
      "num_tokens": 70882204.0,
      "step": 31755
    },
    {
      "entropy": 5.271044063568115,
      "epoch": 3.5669680462739373,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037751614924600417,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21672936528921127,
      "num_tokens": 70894317.0,
      "step": 31760
    },
    {
      "entropy": 5.400760221481323,
      "epoch": 3.567529623181895,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000377480408403937,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21448483914136887,
      "num_tokens": 70906224.0,
      "step": 31765
    },
    {
      "entropy": 5.489929676055908,
      "epoch": 3.568091200089852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003774446642988538,
      "loss": 5.1315,
      "mean_token_accuracy": 0.20607823878526688,
      "num_tokens": 70918134.0,
      "step": 31770
    },
    {
      "entropy": 5.324291610717774,
      "epoch": 3.56865277699781,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037740891693189286,
      "loss": 4.9876,
      "mean_token_accuracy": 0.21535690873861313,
      "num_tokens": 70929695.0,
      "step": 31775
    },
    {
      "entropy": 5.413259744644165,
      "epoch": 3.5692143539057675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003773731663041922,
      "loss": 5.1038,
      "mean_token_accuracy": 0.2061096742749214,
      "num_tokens": 70941204.0,
      "step": 31780
    },
    {
      "entropy": 5.476252555847168,
      "epoch": 3.5697759308137247,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037733741241689024,
      "loss": 5.1031,
      "mean_token_accuracy": 0.21800947189331055,
      "num_tokens": 70953221.0,
      "step": 31785
    },
    {
      "entropy": 5.424005794525146,
      "epoch": 3.5703375077216823,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003773016552711254,
      "loss": 5.0881,
      "mean_token_accuracy": 0.21298297643661498,
      "num_tokens": 70967081.0,
      "step": 31790
    },
    {
      "entropy": 5.335922384262085,
      "epoch": 3.57089908462964,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037726589486803617,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2177308440208435,
      "num_tokens": 70977911.0,
      "step": 31795
    },
    {
      "entropy": 5.402113437652588,
      "epoch": 3.5714606615375977,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037723013120876113,
      "loss": 5.0116,
      "mean_token_accuracy": 0.2193724900484085,
      "num_tokens": 70989001.0,
      "step": 31800
    },
    {
      "entropy": 5.383483839035034,
      "epoch": 3.5720222384455553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037719436429443915,
      "loss": 5.1161,
      "mean_token_accuracy": 0.20519346594810486,
      "num_tokens": 70999698.0,
      "step": 31805
    },
    {
      "entropy": 5.287365484237671,
      "epoch": 3.5725838153535125,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00037715859412620895,
      "loss": 4.8789,
      "mean_token_accuracy": 0.21750407814979553,
      "num_tokens": 71010490.0,
      "step": 31810
    },
    {
      "entropy": 5.395344305038452,
      "epoch": 3.57314539226147,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003771228207052095,
      "loss": 5.0513,
      "mean_token_accuracy": 0.2190518468618393,
      "num_tokens": 71021581.0,
      "step": 31815
    },
    {
      "entropy": 5.408976125717163,
      "epoch": 3.573706969169428,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037708704403257974,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21244900971651076,
      "num_tokens": 71033215.0,
      "step": 31820
    },
    {
      "entropy": 5.339655113220215,
      "epoch": 3.574268546077385,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000377051264109459,
      "loss": 4.8809,
      "mean_token_accuracy": 0.2267154186964035,
      "num_tokens": 71043484.0,
      "step": 31825
    },
    {
      "entropy": 5.339943504333496,
      "epoch": 3.5748301229853428,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003770154809369864,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2158287450671196,
      "num_tokens": 71054101.0,
      "step": 31830
    },
    {
      "entropy": 5.3316810607910154,
      "epoch": 3.5753916998933004,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037697969451630135,
      "loss": 5.0715,
      "mean_token_accuracy": 0.20758387595415115,
      "num_tokens": 71065543.0,
      "step": 31835
    },
    {
      "entropy": 5.3025710105896,
      "epoch": 3.575953276801258,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037694390484854323,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21647845953702927,
      "num_tokens": 71076450.0,
      "step": 31840
    },
    {
      "entropy": 5.2632972240448,
      "epoch": 3.5765148537092157,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003769081119348517,
      "loss": 4.8407,
      "mean_token_accuracy": 0.22231144607067108,
      "num_tokens": 71087625.0,
      "step": 31845
    },
    {
      "entropy": 5.348230361938477,
      "epoch": 3.577076430617173,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037687231577636626,
      "loss": 5.0347,
      "mean_token_accuracy": 0.21219077706336975,
      "num_tokens": 71098205.0,
      "step": 31850
    },
    {
      "entropy": 5.397412395477295,
      "epoch": 3.5776380075251306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003768365163742269,
      "loss": 5.1032,
      "mean_token_accuracy": 0.21260980069637297,
      "num_tokens": 71109207.0,
      "step": 31855
    },
    {
      "entropy": 5.414638376235962,
      "epoch": 3.5781995844330883,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037680071372957327,
      "loss": 5.0553,
      "mean_token_accuracy": 0.2132648453116417,
      "num_tokens": 71120911.0,
      "step": 31860
    },
    {
      "entropy": 5.46744818687439,
      "epoch": 3.5787611613410455,
      "grad_norm": 1.0,
      "learning_rate": 0.00037676490784354543,
      "loss": 5.1103,
      "mean_token_accuracy": 0.21269345730543138,
      "num_tokens": 71131398.0,
      "step": 31865
    },
    {
      "entropy": 5.396149587631226,
      "epoch": 3.579322738249003,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003767290987172835,
      "loss": 5.0532,
      "mean_token_accuracy": 0.2156805843114853,
      "num_tokens": 71142574.0,
      "step": 31870
    },
    {
      "entropy": 5.455266237258911,
      "epoch": 3.579884315156961,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037669328635192757,
      "loss": 5.0678,
      "mean_token_accuracy": 0.20854142755270005,
      "num_tokens": 71154624.0,
      "step": 31875
    },
    {
      "entropy": 5.321757888793945,
      "epoch": 3.580445892064918,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000376657470748618,
      "loss": 4.9136,
      "mean_token_accuracy": 0.21632030457258225,
      "num_tokens": 71165375.0,
      "step": 31880
    },
    {
      "entropy": 5.331001853942871,
      "epoch": 3.5810074689728757,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003766216519084952,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22347745001316072,
      "num_tokens": 71175835.0,
      "step": 31885
    },
    {
      "entropy": 5.389562797546387,
      "epoch": 3.5815690458808334,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003765858298326994,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21424651741981507,
      "num_tokens": 71187173.0,
      "step": 31890
    },
    {
      "entropy": 5.386519098281861,
      "epoch": 3.582130622788791,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037655000452237154,
      "loss": 5.0535,
      "mean_token_accuracy": 0.21431889683008193,
      "num_tokens": 71198424.0,
      "step": 31895
    },
    {
      "entropy": 5.39782133102417,
      "epoch": 3.5826921996967487,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003765141759786521,
      "loss": 5.0628,
      "mean_token_accuracy": 0.20817395746707917,
      "num_tokens": 71210198.0,
      "step": 31900
    },
    {
      "entropy": 5.452140140533447,
      "epoch": 3.583253776604706,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00037647834420268195,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20744257420301437,
      "num_tokens": 71221367.0,
      "step": 31905
    },
    {
      "entropy": 5.3906584739685055,
      "epoch": 3.5838153535126636,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003764425091956019,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21769795268774034,
      "num_tokens": 71232413.0,
      "step": 31910
    },
    {
      "entropy": 5.3784496784210205,
      "epoch": 3.584376930420621,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037640667095855307,
      "loss": 5.0631,
      "mean_token_accuracy": 0.21246958523988724,
      "num_tokens": 71244219.0,
      "step": 31915
    },
    {
      "entropy": 5.301324319839478,
      "epoch": 3.5849385073285784,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037637082949267654,
      "loss": 4.9278,
      "mean_token_accuracy": 0.21962211579084395,
      "num_tokens": 71254452.0,
      "step": 31920
    },
    {
      "entropy": 5.406702661514283,
      "epoch": 3.585500084236536,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003763349847991134,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21405214369297026,
      "num_tokens": 71265267.0,
      "step": 31925
    },
    {
      "entropy": 5.420621585845947,
      "epoch": 3.5860616611444938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000376299136879005,
      "loss": 5.0996,
      "mean_token_accuracy": 0.20791285783052443,
      "num_tokens": 71277334.0,
      "step": 31930
    },
    {
      "entropy": 5.363897228240967,
      "epoch": 3.5866232380524514,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037626328573349296,
      "loss": 5.0161,
      "mean_token_accuracy": 0.2176870211958885,
      "num_tokens": 71289159.0,
      "step": 31935
    },
    {
      "entropy": 5.348265790939331,
      "epoch": 3.587184814960409,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003762274313637185,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2126743108034134,
      "num_tokens": 71300715.0,
      "step": 31940
    },
    {
      "entropy": 5.404525232315064,
      "epoch": 3.5877463918683663,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003761915737708234,
      "loss": 5.012,
      "mean_token_accuracy": 0.21684534698724747,
      "num_tokens": 71310948.0,
      "step": 31945
    },
    {
      "entropy": 5.370616340637207,
      "epoch": 3.588307968776324,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037615571295594936,
      "loss": 5.0927,
      "mean_token_accuracy": 0.2076032429933548,
      "num_tokens": 71321958.0,
      "step": 31950
    },
    {
      "entropy": 5.404864406585693,
      "epoch": 3.5888695456842816,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003761198489202382,
      "loss": 5.0388,
      "mean_token_accuracy": 0.21507318764925004,
      "num_tokens": 71332522.0,
      "step": 31955
    },
    {
      "entropy": 5.433495569229126,
      "epoch": 3.589431122592239,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003760839816648317,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21686243414878845,
      "num_tokens": 71344198.0,
      "step": 31960
    },
    {
      "entropy": 5.368850946426392,
      "epoch": 3.5899926995001965,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037604811119087213,
      "loss": 5.0327,
      "mean_token_accuracy": 0.22034877240657808,
      "num_tokens": 71355094.0,
      "step": 31965
    },
    {
      "entropy": 5.413002634048462,
      "epoch": 3.590554276408154,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003760122374995015,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20936728566884993,
      "num_tokens": 71366596.0,
      "step": 31970
    },
    {
      "entropy": 5.3648255348205565,
      "epoch": 3.5911158533161114,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003759763605918621,
      "loss": 4.9383,
      "mean_token_accuracy": 0.2165632963180542,
      "num_tokens": 71377607.0,
      "step": 31975
    },
    {
      "entropy": 5.309058332443238,
      "epoch": 3.591677430224069,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037594048046909616,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21648721545934677,
      "num_tokens": 71388815.0,
      "step": 31980
    },
    {
      "entropy": 5.301339054107666,
      "epoch": 3.5922390071320267,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003759045971323462,
      "loss": 5.0365,
      "mean_token_accuracy": 0.2136469841003418,
      "num_tokens": 71399600.0,
      "step": 31985
    },
    {
      "entropy": 5.416711950302124,
      "epoch": 3.5928005840399844,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037586871058275476,
      "loss": 4.9961,
      "mean_token_accuracy": 0.21850170493125914,
      "num_tokens": 71410652.0,
      "step": 31990
    },
    {
      "entropy": 5.389948606491089,
      "epoch": 3.593362160947942,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037583282082146434,
      "loss": 5.0914,
      "mean_token_accuracy": 0.21341708898544312,
      "num_tokens": 71422656.0,
      "step": 31995
    },
    {
      "entropy": 5.316637468338013,
      "epoch": 3.5939237378558992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037579692784961785,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2134132817387581,
      "num_tokens": 71433677.0,
      "step": 32000
    },
    {
      "entropy": 5.325966882705688,
      "epoch": 3.594485314763857,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003757610316683581,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21023479998111724,
      "num_tokens": 71445922.0,
      "step": 32005
    },
    {
      "entropy": 5.353648614883423,
      "epoch": 3.5950468916718146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037572513227882807,
      "loss": 4.9654,
      "mean_token_accuracy": 0.21842485219240187,
      "num_tokens": 71457317.0,
      "step": 32010
    },
    {
      "entropy": 5.281710958480835,
      "epoch": 3.595608468579772,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037568922968217065,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21936111450195311,
      "num_tokens": 71468622.0,
      "step": 32015
    },
    {
      "entropy": 5.356270408630371,
      "epoch": 3.5961700454877294,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037565332387952914,
      "loss": 5.0757,
      "mean_token_accuracy": 0.21983907967805863,
      "num_tokens": 71479254.0,
      "step": 32020
    },
    {
      "entropy": 5.3508298873901365,
      "epoch": 3.596731622395687,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037561741487204684,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2191203013062477,
      "num_tokens": 71489694.0,
      "step": 32025
    },
    {
      "entropy": 5.466102600097656,
      "epoch": 3.5972931993036448,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003755815026608669,
      "loss": 5.085,
      "mean_token_accuracy": 0.21422487497329712,
      "num_tokens": 71499983.0,
      "step": 32030
    },
    {
      "entropy": 5.350766706466675,
      "epoch": 3.5978547762116024,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037554558724713305,
      "loss": 4.9674,
      "mean_token_accuracy": 0.22262002825736998,
      "num_tokens": 71511628.0,
      "step": 32035
    },
    {
      "entropy": 5.500546979904175,
      "epoch": 3.5984163531195597,
      "grad_norm": 0.875,
      "learning_rate": 0.0003755096686319886,
      "loss": 5.1541,
      "mean_token_accuracy": 0.20351076126098633,
      "num_tokens": 71524432.0,
      "step": 32040
    },
    {
      "entropy": 5.385019063949585,
      "epoch": 3.5989779300275173,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037547374681657736,
      "loss": 4.9972,
      "mean_token_accuracy": 0.22204924374818802,
      "num_tokens": 71534858.0,
      "step": 32045
    },
    {
      "entropy": 5.357357501983643,
      "epoch": 3.599539506935475,
      "grad_norm": 1.0,
      "learning_rate": 0.00037543782180204303,
      "loss": 5.0828,
      "mean_token_accuracy": 0.21655477881431578,
      "num_tokens": 71547028.0,
      "step": 32050
    },
    {
      "entropy": 5.439584064483642,
      "epoch": 3.600101083843432,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037540189358952955,
      "loss": 5.0229,
      "mean_token_accuracy": 0.21138044446706772,
      "num_tokens": 71557477.0,
      "step": 32055
    },
    {
      "entropy": 5.367985677719116,
      "epoch": 3.60066266075139,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003753659621801808,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21776287108659745,
      "num_tokens": 71568568.0,
      "step": 32060
    },
    {
      "entropy": 5.364889335632324,
      "epoch": 3.6012242376593475,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037533002757514095,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21991496682167053,
      "num_tokens": 71579143.0,
      "step": 32065
    },
    {
      "entropy": 5.337669849395752,
      "epoch": 3.6017858145673047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000375294089775554,
      "loss": 5.0023,
      "mean_token_accuracy": 0.22397076785564424,
      "num_tokens": 71590674.0,
      "step": 32070
    },
    {
      "entropy": 5.326374769210815,
      "epoch": 3.6023473914752624,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003752581487825644,
      "loss": 5.0182,
      "mean_token_accuracy": 0.22071768194437028,
      "num_tokens": 71602133.0,
      "step": 32075
    },
    {
      "entropy": 5.335265398025513,
      "epoch": 3.60290896838322,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037522220459731644,
      "loss": 4.9775,
      "mean_token_accuracy": 0.2181639403104782,
      "num_tokens": 71613735.0,
      "step": 32080
    },
    {
      "entropy": 5.403387641906738,
      "epoch": 3.6034705452911777,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003751862572209546,
      "loss": 4.996,
      "mean_token_accuracy": 0.2175822600722313,
      "num_tokens": 71625553.0,
      "step": 32085
    },
    {
      "entropy": 5.218759298324585,
      "epoch": 3.6040321221991354,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00037515030665462345,
      "loss": 4.8527,
      "mean_token_accuracy": 0.22533819824457169,
      "num_tokens": 71636677.0,
      "step": 32090
    },
    {
      "entropy": 5.356887865066528,
      "epoch": 3.6045936991070926,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037511435289946775,
      "loss": 5.0501,
      "mean_token_accuracy": 0.21088921576738356,
      "num_tokens": 71648040.0,
      "step": 32095
    },
    {
      "entropy": 5.3907393455505375,
      "epoch": 3.6051552760150503,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003750783959566322,
      "loss": 5.002,
      "mean_token_accuracy": 0.214664027094841,
      "num_tokens": 71659189.0,
      "step": 32100
    },
    {
      "entropy": 5.315635824203492,
      "epoch": 3.605716852923008,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003750424358272617,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22381283938884736,
      "num_tokens": 71670262.0,
      "step": 32105
    },
    {
      "entropy": 5.2546977519989015,
      "epoch": 3.606278429830965,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003750064725125013,
      "loss": 4.9863,
      "mean_token_accuracy": 0.21378156393766404,
      "num_tokens": 71680271.0,
      "step": 32110
    },
    {
      "entropy": 5.365630722045898,
      "epoch": 3.606840006738923,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037497050601349605,
      "loss": 5.0385,
      "mean_token_accuracy": 0.21399200111627578,
      "num_tokens": 71691299.0,
      "step": 32115
    },
    {
      "entropy": 5.353423070907593,
      "epoch": 3.6074015836468805,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037493453633139103,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21679147183895112,
      "num_tokens": 71702517.0,
      "step": 32120
    },
    {
      "entropy": 5.29959774017334,
      "epoch": 3.607963160554838,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037489856346733163,
      "loss": 4.9967,
      "mean_token_accuracy": 0.21737123429775237,
      "num_tokens": 71713386.0,
      "step": 32125
    },
    {
      "entropy": 5.375289106369019,
      "epoch": 3.608524737462796,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003748625874224632,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21758001297712326,
      "num_tokens": 71724361.0,
      "step": 32130
    },
    {
      "entropy": 5.421258163452149,
      "epoch": 3.609086314370753,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037482660819793134,
      "loss": 5.1219,
      "mean_token_accuracy": 0.20568763762712478,
      "num_tokens": 71735558.0,
      "step": 32135
    },
    {
      "entropy": 5.424272537231445,
      "epoch": 3.6096478912787107,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003747906257948814,
      "loss": 5.0831,
      "mean_token_accuracy": 0.214164038002491,
      "num_tokens": 71747253.0,
      "step": 32140
    },
    {
      "entropy": 5.450344896316528,
      "epoch": 3.6102094681866683,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003747546402144594,
      "loss": 5.0888,
      "mean_token_accuracy": 0.21168033480644227,
      "num_tokens": 71760223.0,
      "step": 32145
    },
    {
      "entropy": 5.390631008148193,
      "epoch": 3.6107710450946255,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037471865145781094,
      "loss": 4.9477,
      "mean_token_accuracy": 0.20871066749095918,
      "num_tokens": 71771976.0,
      "step": 32150
    },
    {
      "entropy": 5.2973274230957035,
      "epoch": 3.611332622002583,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037468265952608196,
      "loss": 5.0355,
      "mean_token_accuracy": 0.21283090859651566,
      "num_tokens": 71784364.0,
      "step": 32155
    },
    {
      "entropy": 5.399561500549316,
      "epoch": 3.611894198910541,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003746466644204183,
      "loss": 5.0981,
      "mean_token_accuracy": 0.21213567852973939,
      "num_tokens": 71795495.0,
      "step": 32160
    },
    {
      "entropy": 5.421964550018311,
      "epoch": 3.612455775818498,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003746106661419663,
      "loss": 5.0506,
      "mean_token_accuracy": 0.21864643841981887,
      "num_tokens": 71806475.0,
      "step": 32165
    },
    {
      "entropy": 5.389732694625854,
      "epoch": 3.6130173527264557,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000374574664691872,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21502620428800584,
      "num_tokens": 71816701.0,
      "step": 32170
    },
    {
      "entropy": 5.322675895690918,
      "epoch": 3.6135789296344134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037453866007128176,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21448271125555038,
      "num_tokens": 71827860.0,
      "step": 32175
    },
    {
      "entropy": 5.337767648696899,
      "epoch": 3.614140506542371,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037450265228134203,
      "loss": 5.0009,
      "mean_token_accuracy": 0.22033149898052215,
      "num_tokens": 71839631.0,
      "step": 32180
    },
    {
      "entropy": 5.373101043701172,
      "epoch": 3.6147020834503287,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037446664132319925,
      "loss": 4.9986,
      "mean_token_accuracy": 0.2173807606101036,
      "num_tokens": 71849926.0,
      "step": 32185
    },
    {
      "entropy": 5.299531364440918,
      "epoch": 3.615263660358286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000374430627198,
      "loss": 4.967,
      "mean_token_accuracy": 0.21949126273393632,
      "num_tokens": 71860815.0,
      "step": 32190
    },
    {
      "entropy": 5.381127500534058,
      "epoch": 3.6158252372662436,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037439460990689106,
      "loss": 4.9654,
      "mean_token_accuracy": 0.22284289449453354,
      "num_tokens": 71871814.0,
      "step": 32195
    },
    {
      "entropy": 5.376840877532959,
      "epoch": 3.6163868141742013,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003743585894510192,
      "loss": 5.004,
      "mean_token_accuracy": 0.21376866698265076,
      "num_tokens": 71881606.0,
      "step": 32200
    },
    {
      "entropy": 5.319563579559326,
      "epoch": 3.6169483910821585,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037432256583153127,
      "loss": 5.0208,
      "mean_token_accuracy": 0.2139994904398918,
      "num_tokens": 71893371.0,
      "step": 32205
    },
    {
      "entropy": 5.360753917694092,
      "epoch": 3.617509967990116,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037428653904957433,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21744778007268906,
      "num_tokens": 71906068.0,
      "step": 32210
    },
    {
      "entropy": 5.365038251876831,
      "epoch": 3.618071544898074,
      "grad_norm": 1.0,
      "learning_rate": 0.0003742505091062955,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20726863741874696,
      "num_tokens": 71917450.0,
      "step": 32215
    },
    {
      "entropy": 5.41720724105835,
      "epoch": 3.6186331218060315,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003742144760028419,
      "loss": 4.9985,
      "mean_token_accuracy": 0.20926965177059173,
      "num_tokens": 71928333.0,
      "step": 32220
    },
    {
      "entropy": 5.336137485504151,
      "epoch": 3.619194698713989,
      "grad_norm": 1.03125,
      "learning_rate": 0.000374178439740361,
      "loss": 5.0274,
      "mean_token_accuracy": 0.21682146340608596,
      "num_tokens": 71939836.0,
      "step": 32225
    },
    {
      "entropy": 5.433050870895386,
      "epoch": 3.6197562756219464,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003741424003200001,
      "loss": 5.0905,
      "mean_token_accuracy": 0.21390762329101562,
      "num_tokens": 71951120.0,
      "step": 32230
    },
    {
      "entropy": 5.421815299987793,
      "epoch": 3.620317852529904,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037410635774290667,
      "loss": 5.0156,
      "mean_token_accuracy": 0.2207029178738594,
      "num_tokens": 71961547.0,
      "step": 32235
    },
    {
      "entropy": 5.365297508239746,
      "epoch": 3.6208794294378617,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003740703120102284,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2178094670176506,
      "num_tokens": 71973287.0,
      "step": 32240
    },
    {
      "entropy": 5.369452047348022,
      "epoch": 3.621441006345819,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000374034263123113,
      "loss": 5.0005,
      "mean_token_accuracy": 0.22316586822271348,
      "num_tokens": 71984205.0,
      "step": 32245
    },
    {
      "entropy": 5.270245838165283,
      "epoch": 3.6220025832537766,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037399821108270823,
      "loss": 4.9998,
      "mean_token_accuracy": 0.2136416882276535,
      "num_tokens": 71995090.0,
      "step": 32250
    },
    {
      "entropy": 5.453562068939209,
      "epoch": 3.622564160161734,
      "grad_norm": 0.953125,
      "learning_rate": 0.000373962155890162,
      "loss": 5.1508,
      "mean_token_accuracy": 0.21236258745193481,
      "num_tokens": 72005504.0,
      "step": 32255
    },
    {
      "entropy": 5.3908017635345455,
      "epoch": 3.6231257370696914,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003739260975466223,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21978026628494263,
      "num_tokens": 72017250.0,
      "step": 32260
    },
    {
      "entropy": 5.384848499298096,
      "epoch": 3.623687313977649,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037389003605323737,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2180093452334404,
      "num_tokens": 72028089.0,
      "step": 32265
    },
    {
      "entropy": 5.350279808044434,
      "epoch": 3.6242488908856068,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037385397141115535,
      "loss": 4.9959,
      "mean_token_accuracy": 0.2145090490579605,
      "num_tokens": 72039140.0,
      "step": 32270
    },
    {
      "entropy": 5.321670246124268,
      "epoch": 3.6248104677935644,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037381790362152453,
      "loss": 5.0244,
      "mean_token_accuracy": 0.2118471547961235,
      "num_tokens": 72049944.0,
      "step": 32275
    },
    {
      "entropy": 5.28717074394226,
      "epoch": 3.625372044701522,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037378183268549326,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2167445972561836,
      "num_tokens": 72061799.0,
      "step": 32280
    },
    {
      "entropy": 5.367202663421631,
      "epoch": 3.6259336216094793,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003737457586042102,
      "loss": 4.9803,
      "mean_token_accuracy": 0.22183501422405244,
      "num_tokens": 72073539.0,
      "step": 32285
    },
    {
      "entropy": 5.340781021118164,
      "epoch": 3.626495198517437,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003737096813788238,
      "loss": 5.0046,
      "mean_token_accuracy": 0.21146558225154877,
      "num_tokens": 72085268.0,
      "step": 32290
    },
    {
      "entropy": 5.4372461318969725,
      "epoch": 3.6270567754253946,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003736736010104829,
      "loss": 5.1977,
      "mean_token_accuracy": 0.20042970925569534,
      "num_tokens": 72096269.0,
      "step": 32295
    },
    {
      "entropy": 5.347432708740234,
      "epoch": 3.627618352333352,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003736375175003363,
      "loss": 4.9874,
      "mean_token_accuracy": 0.22317456156015397,
      "num_tokens": 72108934.0,
      "step": 32300
    },
    {
      "entropy": 5.339038515090943,
      "epoch": 3.6281799292413095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037360143084953285,
      "loss": 4.9902,
      "mean_token_accuracy": 0.2159942716360092,
      "num_tokens": 72118371.0,
      "step": 32305
    },
    {
      "entropy": 5.277317857742309,
      "epoch": 3.628741506149267,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003735653410592217,
      "loss": 4.9035,
      "mean_token_accuracy": 0.22319502234458924,
      "num_tokens": 72129187.0,
      "step": 32310
    },
    {
      "entropy": 5.34218316078186,
      "epoch": 3.629303083057225,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037352924813055185,
      "loss": 4.9792,
      "mean_token_accuracy": 0.22550194710493088,
      "num_tokens": 72141317.0,
      "step": 32315
    },
    {
      "entropy": 5.358471441268921,
      "epoch": 3.6298646599651825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003734931520646724,
      "loss": 4.9919,
      "mean_token_accuracy": 0.22387682646512985,
      "num_tokens": 72152754.0,
      "step": 32320
    },
    {
      "entropy": 5.257072401046753,
      "epoch": 3.6304262368731397,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037345705286273285,
      "loss": 4.8639,
      "mean_token_accuracy": 0.23085146099328996,
      "num_tokens": 72162115.0,
      "step": 32325
    },
    {
      "entropy": 5.284404659271241,
      "epoch": 3.6309878137810974,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003734209505258825,
      "loss": 4.941,
      "mean_token_accuracy": 0.22192294299602508,
      "num_tokens": 72173633.0,
      "step": 32330
    },
    {
      "entropy": 5.379550123214722,
      "epoch": 3.631549390689055,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037338484505527103,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20829136222600936,
      "num_tokens": 72185274.0,
      "step": 32335
    },
    {
      "entropy": 5.43889012336731,
      "epoch": 3.6321109675970122,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037334873645204785,
      "loss": 5.0815,
      "mean_token_accuracy": 0.21240959763526918,
      "num_tokens": 72195991.0,
      "step": 32340
    },
    {
      "entropy": 5.315511989593506,
      "epoch": 3.63267254450497,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037331262471736285,
      "loss": 5.0095,
      "mean_token_accuracy": 0.21945120841264726,
      "num_tokens": 72207390.0,
      "step": 32345
    },
    {
      "entropy": 5.311455297470093,
      "epoch": 3.6332341214129276,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037327650985236564,
      "loss": 5.036,
      "mean_token_accuracy": 0.21587470173835754,
      "num_tokens": 72218747.0,
      "step": 32350
    },
    {
      "entropy": 5.373234987258911,
      "epoch": 3.633795698320885,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003732403918582063,
      "loss": 5.0196,
      "mean_token_accuracy": 0.21758537590503693,
      "num_tokens": 72230623.0,
      "step": 32355
    },
    {
      "entropy": 5.365130376815796,
      "epoch": 3.6343572752288424,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00037320427073603475,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21783121973276137,
      "num_tokens": 72242510.0,
      "step": 32360
    },
    {
      "entropy": 5.292915153503418,
      "epoch": 3.6349188521368,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037316814648700113,
      "loss": 4.9498,
      "mean_token_accuracy": 0.22010420262813568,
      "num_tokens": 72253498.0,
      "step": 32365
    },
    {
      "entropy": 5.334832334518433,
      "epoch": 3.6354804290447578,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003731320191122557,
      "loss": 5.0701,
      "mean_token_accuracy": 0.21585175693035125,
      "num_tokens": 72265698.0,
      "step": 32370
    },
    {
      "entropy": 5.325212478637695,
      "epoch": 3.6360420059527154,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003730958886129486,
      "loss": 4.9015,
      "mean_token_accuracy": 0.23041606545448304,
      "num_tokens": 72276563.0,
      "step": 32375
    },
    {
      "entropy": 5.331287288665772,
      "epoch": 3.6366035828606726,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037305975499023046,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21958916932344436,
      "num_tokens": 72286826.0,
      "step": 32380
    },
    {
      "entropy": 5.429649877548218,
      "epoch": 3.6371651597686303,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003730236182452517,
      "loss": 5.129,
      "mean_token_accuracy": 0.21019827872514724,
      "num_tokens": 72299446.0,
      "step": 32385
    },
    {
      "entropy": 5.417607831954956,
      "epoch": 3.637726736676588,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003729874783791629,
      "loss": 5.1436,
      "mean_token_accuracy": 0.2115332841873169,
      "num_tokens": 72310894.0,
      "step": 32390
    },
    {
      "entropy": 5.374290847778321,
      "epoch": 3.638288313584545,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003729513353931148,
      "loss": 5.03,
      "mean_token_accuracy": 0.21918028146028518,
      "num_tokens": 72322810.0,
      "step": 32395
    },
    {
      "entropy": 5.39676308631897,
      "epoch": 3.638849890492503,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003729151892882581,
      "loss": 5.0368,
      "mean_token_accuracy": 0.21922408938407897,
      "num_tokens": 72334509.0,
      "step": 32400
    },
    {
      "entropy": 5.299383687973022,
      "epoch": 3.6394114674004605,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003728790400657439,
      "loss": 5.0067,
      "mean_token_accuracy": 0.21406272649765015,
      "num_tokens": 72345036.0,
      "step": 32405
    },
    {
      "entropy": 5.375289344787598,
      "epoch": 3.639973044308418,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037284288772672315,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21222671270370483,
      "num_tokens": 72356029.0,
      "step": 32410
    },
    {
      "entropy": 5.400136184692383,
      "epoch": 3.640534621216376,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037280673227234687,
      "loss": 5.0671,
      "mean_token_accuracy": 0.21620844751596452,
      "num_tokens": 72366548.0,
      "step": 32415
    },
    {
      "entropy": 5.273833751678467,
      "epoch": 3.641096198124333,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003727705737037662,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21671319603919983,
      "num_tokens": 72377271.0,
      "step": 32420
    },
    {
      "entropy": 5.358374738693238,
      "epoch": 3.6416577750322907,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003727344120221326,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21629077792167664,
      "num_tokens": 72388651.0,
      "step": 32425
    },
    {
      "entropy": 5.313619184494018,
      "epoch": 3.6422193519402484,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037269824722859743,
      "loss": 5.081,
      "mean_token_accuracy": 0.21059762686491013,
      "num_tokens": 72400955.0,
      "step": 32430
    },
    {
      "entropy": 5.399298524856567,
      "epoch": 3.6427809288482056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003726620793243121,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21027046740055083,
      "num_tokens": 72412372.0,
      "step": 32435
    },
    {
      "entropy": 5.380302047729492,
      "epoch": 3.6433425057561633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003726259083104284,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21565556824207305,
      "num_tokens": 72422664.0,
      "step": 32440
    },
    {
      "entropy": 5.332033586502075,
      "epoch": 3.643904082664121,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037258973418809784,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21102367490530013,
      "num_tokens": 72434829.0,
      "step": 32445
    },
    {
      "entropy": 5.295136976242065,
      "epoch": 3.644465659572078,
      "grad_norm": 1.0,
      "learning_rate": 0.0003725535569584723,
      "loss": 4.9312,
      "mean_token_accuracy": 0.21832533031702042,
      "num_tokens": 72445140.0,
      "step": 32450
    },
    {
      "entropy": 5.313194704055786,
      "epoch": 3.645027236480036,
      "grad_norm": 1.0,
      "learning_rate": 0.00037251737662270367,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22916316539049147,
      "num_tokens": 72455281.0,
      "step": 32455
    },
    {
      "entropy": 5.33523416519165,
      "epoch": 3.6455888133879935,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000372481193181944,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21002114713191986,
      "num_tokens": 72466296.0,
      "step": 32460
    },
    {
      "entropy": 5.3231658935546875,
      "epoch": 3.646150390295951,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037244500663734524,
      "loss": 4.9901,
      "mean_token_accuracy": 0.2133605495095253,
      "num_tokens": 72477390.0,
      "step": 32465
    },
    {
      "entropy": 5.331514549255371,
      "epoch": 3.6467119672039088,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003724088169900597,
      "loss": 4.9272,
      "mean_token_accuracy": 0.2232219770550728,
      "num_tokens": 72488241.0,
      "step": 32470
    },
    {
      "entropy": 5.385822057723999,
      "epoch": 3.647273544111866,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037237262424123975,
      "loss": 5.0096,
      "mean_token_accuracy": 0.21734921336174012,
      "num_tokens": 72498823.0,
      "step": 32475
    },
    {
      "entropy": 5.376839828491211,
      "epoch": 3.6478351210198237,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037233642839203763,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20787446647882463,
      "num_tokens": 72509399.0,
      "step": 32480
    },
    {
      "entropy": 5.435761737823486,
      "epoch": 3.6483966979277813,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003723002294436059,
      "loss": 5.0855,
      "mean_token_accuracy": 0.20962174832820893,
      "num_tokens": 72519182.0,
      "step": 32485
    },
    {
      "entropy": 5.266895818710327,
      "epoch": 3.6489582748357385,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037226402739709704,
      "loss": 4.92,
      "mean_token_accuracy": 0.21698999851942063,
      "num_tokens": 72528852.0,
      "step": 32490
    },
    {
      "entropy": 5.316558122634888,
      "epoch": 3.649519851743696,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003722278222536639,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2137270078063011,
      "num_tokens": 72540477.0,
      "step": 32495
    },
    {
      "entropy": 5.414946746826172,
      "epoch": 3.650081428651654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037219161401445914,
      "loss": 5.0332,
      "mean_token_accuracy": 0.2119777351617813,
      "num_tokens": 72553404.0,
      "step": 32500
    },
    {
      "entropy": 5.387693166732788,
      "epoch": 3.6506430055596115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037215540268063573,
      "loss": 5.0427,
      "mean_token_accuracy": 0.2132757470011711,
      "num_tokens": 72564516.0,
      "step": 32505
    },
    {
      "entropy": 5.297772312164307,
      "epoch": 3.651204582467569,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003721191882533466,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2175482451915741,
      "num_tokens": 72575888.0,
      "step": 32510
    },
    {
      "entropy": 5.353543901443482,
      "epoch": 3.6517661593755264,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003720829707337449,
      "loss": 4.992,
      "mean_token_accuracy": 0.21309920400381088,
      "num_tokens": 72586629.0,
      "step": 32515
    },
    {
      "entropy": 5.372041845321656,
      "epoch": 3.652327736283484,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003720467501229836,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21766812056303025,
      "num_tokens": 72597264.0,
      "step": 32520
    },
    {
      "entropy": 5.369106960296631,
      "epoch": 3.6528893131914417,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003720105264222163,
      "loss": 4.9995,
      "mean_token_accuracy": 0.2159409776329994,
      "num_tokens": 72608450.0,
      "step": 32525
    },
    {
      "entropy": 5.305699157714844,
      "epoch": 3.653450890099399,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00037197429963259615,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20834420323371888,
      "num_tokens": 72620538.0,
      "step": 32530
    },
    {
      "entropy": 5.405397033691406,
      "epoch": 3.6540124670073566,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003719380697552767,
      "loss": 5.0862,
      "mean_token_accuracy": 0.21224789023399354,
      "num_tokens": 72632589.0,
      "step": 32535
    },
    {
      "entropy": 5.4171364307403564,
      "epoch": 3.6545740439153143,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003719018367914115,
      "loss": 5.0096,
      "mean_token_accuracy": 0.21378622055053711,
      "num_tokens": 72643749.0,
      "step": 32540
    },
    {
      "entropy": 5.395247602462769,
      "epoch": 3.6551356208232715,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003718656007421542,
      "loss": 5.0529,
      "mean_token_accuracy": 0.21652597486972808,
      "num_tokens": 72655485.0,
      "step": 32545
    },
    {
      "entropy": 5.408996677398681,
      "epoch": 3.655697197731229,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037182936160865866,
      "loss": 5.0553,
      "mean_token_accuracy": 0.21566933542490005,
      "num_tokens": 72666175.0,
      "step": 32550
    },
    {
      "entropy": 5.343943977355957,
      "epoch": 3.656258774639187,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003717931193920788,
      "loss": 4.9594,
      "mean_token_accuracy": 0.2137770876288414,
      "num_tokens": 72675858.0,
      "step": 32555
    },
    {
      "entropy": 5.282408571243286,
      "epoch": 3.6568203515471445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003717568740935683,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2177508443593979,
      "num_tokens": 72686623.0,
      "step": 32560
    },
    {
      "entropy": 5.3714423179626465,
      "epoch": 3.657381928455102,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003717206257142815,
      "loss": 4.9847,
      "mean_token_accuracy": 0.22171366810798646,
      "num_tokens": 72698002.0,
      "step": 32565
    },
    {
      "entropy": 5.34528079032898,
      "epoch": 3.6579435053630593,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003716843742553724,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21767104268074036,
      "num_tokens": 72708981.0,
      "step": 32570
    },
    {
      "entropy": 5.256150579452514,
      "epoch": 3.658505082271017,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003716481197179953,
      "loss": 5.0025,
      "mean_token_accuracy": 0.21651729047298432,
      "num_tokens": 72719764.0,
      "step": 32575
    },
    {
      "entropy": 5.511712551116943,
      "epoch": 3.6590666591789747,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037161186210330464,
      "loss": 5.2167,
      "mean_token_accuracy": 0.20582916289567948,
      "num_tokens": 72730915.0,
      "step": 32580
    },
    {
      "entropy": 5.4349205493927,
      "epoch": 3.659628236086932,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037157560141245485,
      "loss": 5.0709,
      "mean_token_accuracy": 0.21230919808149337,
      "num_tokens": 72741869.0,
      "step": 32585
    },
    {
      "entropy": 5.3999162197113035,
      "epoch": 3.6601898129948895,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037153933764660034,
      "loss": 5.0238,
      "mean_token_accuracy": 0.2150506779551506,
      "num_tokens": 72753638.0,
      "step": 32590
    },
    {
      "entropy": 5.4058009624481205,
      "epoch": 3.660751389902847,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000371503070806896,
      "loss": 5.0448,
      "mean_token_accuracy": 0.21725285202264785,
      "num_tokens": 72765505.0,
      "step": 32595
    },
    {
      "entropy": 5.334006929397583,
      "epoch": 3.661312966810805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037146680089449643,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22112863063812255,
      "num_tokens": 72776348.0,
      "step": 32600
    },
    {
      "entropy": 5.37630558013916,
      "epoch": 3.6618745437187625,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037143052791055654,
      "loss": 5.0897,
      "mean_token_accuracy": 0.21058931350708007,
      "num_tokens": 72787720.0,
      "step": 32605
    },
    {
      "entropy": 5.334755086898804,
      "epoch": 3.6624361206267197,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003713942518562311,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21372233033180238,
      "num_tokens": 72798947.0,
      "step": 32610
    },
    {
      "entropy": 5.412807369232178,
      "epoch": 3.6629976975346774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003713579727326754,
      "loss": 5.1168,
      "mean_token_accuracy": 0.2093987613916397,
      "num_tokens": 72810859.0,
      "step": 32615
    },
    {
      "entropy": 5.317181921005249,
      "epoch": 3.663559274442635,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037132169054104446,
      "loss": 4.9371,
      "mean_token_accuracy": 0.22443264126777648,
      "num_tokens": 72821089.0,
      "step": 32620
    },
    {
      "entropy": 5.270217370986939,
      "epoch": 3.6641208513505923,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037128540528249355,
      "loss": 4.9047,
      "mean_token_accuracy": 0.2201240509748459,
      "num_tokens": 72832226.0,
      "step": 32625
    },
    {
      "entropy": 5.359174013137817,
      "epoch": 3.66468242825855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000371249116958178,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2202112063765526,
      "num_tokens": 72843939.0,
      "step": 32630
    },
    {
      "entropy": 5.29600567817688,
      "epoch": 3.6652440051665076,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037121282556925326,
      "loss": 4.8645,
      "mean_token_accuracy": 0.2257383406162262,
      "num_tokens": 72854332.0,
      "step": 32635
    },
    {
      "entropy": 5.392191171646118,
      "epoch": 3.665805582074465,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003711765311168748,
      "loss": 5.1572,
      "mean_token_accuracy": 0.20914976745843888,
      "num_tokens": 72865530.0,
      "step": 32640
    },
    {
      "entropy": 5.389699172973633,
      "epoch": 3.6663671589824225,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037114023360219824,
      "loss": 5.0908,
      "mean_token_accuracy": 0.21136886328458787,
      "num_tokens": 72877683.0,
      "step": 32645
    },
    {
      "entropy": 5.377852964401245,
      "epoch": 3.66692873589038,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037110393302637937,
      "loss": 5.0246,
      "mean_token_accuracy": 0.22078874260187148,
      "num_tokens": 72888926.0,
      "step": 32650
    },
    {
      "entropy": 5.381099510192871,
      "epoch": 3.667490312798338,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037106762939057404,
      "loss": 5.0582,
      "mean_token_accuracy": 0.21053049862384796,
      "num_tokens": 72901318.0,
      "step": 32655
    },
    {
      "entropy": 5.466742849349975,
      "epoch": 3.6680518897062955,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003710313226959381,
      "loss": 5.0914,
      "mean_token_accuracy": 0.21864370852708817,
      "num_tokens": 72913177.0,
      "step": 32660
    },
    {
      "entropy": 5.442661952972412,
      "epoch": 3.6686134666142527,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003709950129436276,
      "loss": 5.1295,
      "mean_token_accuracy": 0.20580410659313203,
      "num_tokens": 72924322.0,
      "step": 32665
    },
    {
      "entropy": 5.339646244049073,
      "epoch": 3.6691750435222104,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003709587001347987,
      "loss": 4.995,
      "mean_token_accuracy": 0.22136670649051665,
      "num_tokens": 72936287.0,
      "step": 32670
    },
    {
      "entropy": 5.427655029296875,
      "epoch": 3.669736620430168,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037092238427060756,
      "loss": 5.0763,
      "mean_token_accuracy": 0.2081774428486824,
      "num_tokens": 72949262.0,
      "step": 32675
    },
    {
      "entropy": 5.427682590484619,
      "epoch": 3.6702981973381252,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003708860653522104,
      "loss": 5.0894,
      "mean_token_accuracy": 0.21382759660482406,
      "num_tokens": 72959283.0,
      "step": 32680
    },
    {
      "entropy": 5.32410078048706,
      "epoch": 3.670859774246083,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003708497433807639,
      "loss": 4.934,
      "mean_token_accuracy": 0.2243756502866745,
      "num_tokens": 72971063.0,
      "step": 32685
    },
    {
      "entropy": 5.304924535751343,
      "epoch": 3.6714213511540406,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037081341835742433,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22128563225269318,
      "num_tokens": 72980775.0,
      "step": 32690
    },
    {
      "entropy": 5.302289247512817,
      "epoch": 3.671982928061998,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003707770902833483,
      "loss": 4.969,
      "mean_token_accuracy": 0.22104980945587158,
      "num_tokens": 72992166.0,
      "step": 32695
    },
    {
      "entropy": 5.276807546615601,
      "epoch": 3.672544504969956,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003707407591596926,
      "loss": 4.916,
      "mean_token_accuracy": 0.21515513807535172,
      "num_tokens": 73003194.0,
      "step": 32700
    },
    {
      "entropy": 5.324234533309936,
      "epoch": 3.673106081877913,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000370704424987614,
      "loss": 5.0199,
      "mean_token_accuracy": 0.21842456459999085,
      "num_tokens": 73013375.0,
      "step": 32705
    },
    {
      "entropy": 5.393124961853028,
      "epoch": 3.6736676587858708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003706680877682694,
      "loss": 5.036,
      "mean_token_accuracy": 0.21864428967237473,
      "num_tokens": 73024532.0,
      "step": 32710
    },
    {
      "entropy": 5.339003324508667,
      "epoch": 3.6742292356938284,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037063174750281573,
      "loss": 4.9642,
      "mean_token_accuracy": 0.2214416339993477,
      "num_tokens": 73034721.0,
      "step": 32715
    },
    {
      "entropy": 5.316589498519898,
      "epoch": 3.6747908126017856,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003705954041924102,
      "loss": 4.9386,
      "mean_token_accuracy": 0.22162773460149765,
      "num_tokens": 73045894.0,
      "step": 32720
    },
    {
      "entropy": 5.180207109451294,
      "epoch": 3.6753523895097433,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037055905783821,
      "loss": 4.8217,
      "mean_token_accuracy": 0.23719478398561478,
      "num_tokens": 73057370.0,
      "step": 32725
    },
    {
      "entropy": 5.410839462280274,
      "epoch": 3.675913966417701,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037052270844137216,
      "loss": 5.0912,
      "mean_token_accuracy": 0.2135603666305542,
      "num_tokens": 73067700.0,
      "step": 32730
    },
    {
      "entropy": 5.312532377243042,
      "epoch": 3.676475543325658,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003704863560030543,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21994376480579375,
      "num_tokens": 73079612.0,
      "step": 32735
    },
    {
      "entropy": 5.368846035003662,
      "epoch": 3.677037120233616,
      "grad_norm": 1.0,
      "learning_rate": 0.0003704500005244139,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21279377043247222,
      "num_tokens": 73091060.0,
      "step": 32740
    },
    {
      "entropy": 5.407676458358765,
      "epoch": 3.6775986971415735,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037041364200660835,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21740828156471254,
      "num_tokens": 73102218.0,
      "step": 32745
    },
    {
      "entropy": 5.457616138458252,
      "epoch": 3.678160274049531,
      "grad_norm": 1.0,
      "learning_rate": 0.00037037728045079553,
      "loss": 5.0831,
      "mean_token_accuracy": 0.20908118337392806,
      "num_tokens": 73113669.0,
      "step": 32750
    },
    {
      "entropy": 5.381910610198974,
      "epoch": 3.678721850957489,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000370340915858133,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21981036365032197,
      "num_tokens": 73124765.0,
      "step": 32755
    },
    {
      "entropy": 5.36355996131897,
      "epoch": 3.679283427865446,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003703045482297788,
      "loss": 5.076,
      "mean_token_accuracy": 0.21208474785089493,
      "num_tokens": 73135278.0,
      "step": 32760
    },
    {
      "entropy": 5.243255615234375,
      "epoch": 3.6798450047734037,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003702681775668908,
      "loss": 4.833,
      "mean_token_accuracy": 0.22018363326787949,
      "num_tokens": 73145176.0,
      "step": 32765
    },
    {
      "entropy": 5.447045755386353,
      "epoch": 3.6804065816813614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003702318038706272,
      "loss": 5.1274,
      "mean_token_accuracy": 0.20888967216014862,
      "num_tokens": 73156001.0,
      "step": 32770
    },
    {
      "entropy": 5.2924768924713135,
      "epoch": 3.6809681585893186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003701954271421459,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21709701865911485,
      "num_tokens": 73167116.0,
      "step": 32775
    },
    {
      "entropy": 5.4066637516021725,
      "epoch": 3.6815297354972762,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037015904738260533,
      "loss": 5.0621,
      "mean_token_accuracy": 0.22130534648895264,
      "num_tokens": 73178750.0,
      "step": 32780
    },
    {
      "entropy": 5.326592254638672,
      "epoch": 3.682091312405234,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003701226645931637,
      "loss": 4.9627,
      "mean_token_accuracy": 0.22635652422904967,
      "num_tokens": 73189781.0,
      "step": 32785
    },
    {
      "entropy": 5.295560884475708,
      "epoch": 3.6826528893131916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003700862787749796,
      "loss": 4.931,
      "mean_token_accuracy": 0.22036688178777694,
      "num_tokens": 73200374.0,
      "step": 32790
    },
    {
      "entropy": 5.429730653762817,
      "epoch": 3.6832144662211492,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003700498899292115,
      "loss": 5.0844,
      "mean_token_accuracy": 0.2071483165025711,
      "num_tokens": 73211613.0,
      "step": 32795
    },
    {
      "entropy": 5.337763118743896,
      "epoch": 3.6837760431291064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003700134980570181,
      "loss": 4.9748,
      "mean_token_accuracy": 0.23014363199472426,
      "num_tokens": 73222309.0,
      "step": 32800
    },
    {
      "entropy": 5.468933582305908,
      "epoch": 3.684337620037064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003699771031595579,
      "loss": 5.2464,
      "mean_token_accuracy": 0.20184991806745528,
      "num_tokens": 73233956.0,
      "step": 32805
    },
    {
      "entropy": 5.4456202507019045,
      "epoch": 3.6848991969450218,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036994070523799005,
      "loss": 5.1035,
      "mean_token_accuracy": 0.2092838004231453,
      "num_tokens": 73245218.0,
      "step": 32810
    },
    {
      "entropy": 5.348458337783813,
      "epoch": 3.685460773852979,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036990430429347334,
      "loss": 4.9723,
      "mean_token_accuracy": 0.22643471360206605,
      "num_tokens": 73255830.0,
      "step": 32815
    },
    {
      "entropy": 5.303211307525634,
      "epoch": 3.6860223507609367,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036986790032716673,
      "loss": 4.8953,
      "mean_token_accuracy": 0.21931997835636138,
      "num_tokens": 73266384.0,
      "step": 32820
    },
    {
      "entropy": 5.345702362060547,
      "epoch": 3.6865839276688943,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003698314933402292,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21320143342018127,
      "num_tokens": 73276839.0,
      "step": 32825
    },
    {
      "entropy": 5.398399066925049,
      "epoch": 3.6871455045768515,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003697950833338204,
      "loss": 5.0736,
      "mean_token_accuracy": 0.2170320615172386,
      "num_tokens": 73287398.0,
      "step": 32830
    },
    {
      "entropy": 5.338272905349731,
      "epoch": 3.687707081484809,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036975867030909933,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21587843149900438,
      "num_tokens": 73297432.0,
      "step": 32835
    },
    {
      "entropy": 5.315424203872681,
      "epoch": 3.688268658392767,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036972225426722545,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22218521535396576,
      "num_tokens": 73308539.0,
      "step": 32840
    },
    {
      "entropy": 5.3108954429626465,
      "epoch": 3.6888302353007245,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003696858352093582,
      "loss": 4.9841,
      "mean_token_accuracy": 0.221356201171875,
      "num_tokens": 73319698.0,
      "step": 32845
    },
    {
      "entropy": 5.3234601497650145,
      "epoch": 3.689391812208682,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036964941313665715,
      "loss": 5.0337,
      "mean_token_accuracy": 0.21379416882991792,
      "num_tokens": 73329747.0,
      "step": 32850
    },
    {
      "entropy": 5.401691341400147,
      "epoch": 3.6899533891166394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003696129880502823,
      "loss": 5.0184,
      "mean_token_accuracy": 0.2223362520337105,
      "num_tokens": 73340523.0,
      "step": 32855
    },
    {
      "entropy": 5.571006202697754,
      "epoch": 3.690514966024597,
      "grad_norm": 1.046875,
      "learning_rate": 0.000369576559951393,
      "loss": 5.1014,
      "mean_token_accuracy": 0.20630527436733245,
      "num_tokens": 73352219.0,
      "step": 32860
    },
    {
      "entropy": 5.3385804176330565,
      "epoch": 3.6910765429325547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036954012884114946,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21635428965091705,
      "num_tokens": 73363157.0,
      "step": 32865
    },
    {
      "entropy": 5.337261772155761,
      "epoch": 3.691638119840512,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003695036947207115,
      "loss": 5.0763,
      "mean_token_accuracy": 0.21711685061454772,
      "num_tokens": 73375387.0,
      "step": 32870
    },
    {
      "entropy": 5.4419028759002686,
      "epoch": 3.6921996967484696,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036946725759123936,
      "loss": 5.1502,
      "mean_token_accuracy": 0.20825972259044648,
      "num_tokens": 73388056.0,
      "step": 32875
    },
    {
      "entropy": 5.448140096664429,
      "epoch": 3.6927612736564273,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003694308174538929,
      "loss": 5.0987,
      "mean_token_accuracy": 0.21698795408010482,
      "num_tokens": 73398510.0,
      "step": 32880
    },
    {
      "entropy": 5.370825147628784,
      "epoch": 3.693322850564385,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036939437430983274,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2208140268921852,
      "num_tokens": 73409382.0,
      "step": 32885
    },
    {
      "entropy": 5.299609756469726,
      "epoch": 3.6938844274723426,
      "grad_norm": 1.0,
      "learning_rate": 0.00036935792816021905,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21493352353572845,
      "num_tokens": 73419958.0,
      "step": 32890
    },
    {
      "entropy": 5.391295289993286,
      "epoch": 3.6944460043803,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036932147900621234,
      "loss": 4.9679,
      "mean_token_accuracy": 0.21798883974552155,
      "num_tokens": 73430982.0,
      "step": 32895
    },
    {
      "entropy": 5.3664226055145265,
      "epoch": 3.6950075812882575,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036928502684897314,
      "loss": 4.9895,
      "mean_token_accuracy": 0.22239900529384612,
      "num_tokens": 73442375.0,
      "step": 32900
    },
    {
      "entropy": 5.29531888961792,
      "epoch": 3.695569158196215,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00036924857168966216,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22790358066558838,
      "num_tokens": 73452212.0,
      "step": 32905
    },
    {
      "entropy": 5.280654573440552,
      "epoch": 3.6961307351041723,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003692121135294401,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21572890430688857,
      "num_tokens": 73462998.0,
      "step": 32910
    },
    {
      "entropy": 5.252621698379516,
      "epoch": 3.69669231201213,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036917565236946776,
      "loss": 4.9381,
      "mean_token_accuracy": 0.21879514306783676,
      "num_tokens": 73474387.0,
      "step": 32915
    },
    {
      "entropy": 5.350155162811279,
      "epoch": 3.6972538889200877,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036913918821090616,
      "loss": 4.9386,
      "mean_token_accuracy": 0.22192672342061998,
      "num_tokens": 73484524.0,
      "step": 32920
    },
    {
      "entropy": 5.337688732147217,
      "epoch": 3.697815465828045,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036910272105491625,
      "loss": 4.96,
      "mean_token_accuracy": 0.22790093272924422,
      "num_tokens": 73495286.0,
      "step": 32925
    },
    {
      "entropy": 5.299866819381714,
      "epoch": 3.6983770427360025,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00036906625090265925,
      "loss": 5.0034,
      "mean_token_accuracy": 0.2138687402009964,
      "num_tokens": 73507366.0,
      "step": 32930
    },
    {
      "entropy": 5.321167755126953,
      "epoch": 3.69893861964396,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036902977775529636,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2273053452372551,
      "num_tokens": 73517823.0,
      "step": 32935
    },
    {
      "entropy": 5.294930076599121,
      "epoch": 3.699500196551918,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003689933016139888,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21567903310060502,
      "num_tokens": 73529199.0,
      "step": 32940
    },
    {
      "entropy": 5.315556240081787,
      "epoch": 3.7000617734598755,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003689568224798981,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21435010880231858,
      "num_tokens": 73539845.0,
      "step": 32945
    },
    {
      "entropy": 5.418497037887573,
      "epoch": 3.7006233503678327,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003689203403541857,
      "loss": 5.0286,
      "mean_token_accuracy": 0.2127314105629921,
      "num_tokens": 73551331.0,
      "step": 32950
    },
    {
      "entropy": 5.373791933059692,
      "epoch": 3.7011849272757904,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036888385523801325,
      "loss": 5.0308,
      "mean_token_accuracy": 0.2199239820241928,
      "num_tokens": 73562155.0,
      "step": 32955
    },
    {
      "entropy": 5.284725332260132,
      "epoch": 3.701746504183748,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036884736713254247,
      "loss": 4.998,
      "mean_token_accuracy": 0.22226016968488693,
      "num_tokens": 73573014.0,
      "step": 32960
    },
    {
      "entropy": 5.320525312423706,
      "epoch": 3.7023080810917053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036881087603893515,
      "loss": 5.014,
      "mean_token_accuracy": 0.2107241317629814,
      "num_tokens": 73584919.0,
      "step": 32965
    },
    {
      "entropy": 5.326927137374878,
      "epoch": 3.702869657999663,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036877438195835307,
      "loss": 4.9662,
      "mean_token_accuracy": 0.2161330223083496,
      "num_tokens": 73596015.0,
      "step": 32970
    },
    {
      "entropy": 5.337098264694214,
      "epoch": 3.7034312349076206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003687378848919584,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22096004635095595,
      "num_tokens": 73606312.0,
      "step": 32975
    },
    {
      "entropy": 5.4395421028137205,
      "epoch": 3.7039928118155783,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036870138484091297,
      "loss": 5.1935,
      "mean_token_accuracy": 0.20167872458696365,
      "num_tokens": 73617411.0,
      "step": 32980
    },
    {
      "entropy": 5.4689970970153805,
      "epoch": 3.704554388723536,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036866488180637926,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21120749413967133,
      "num_tokens": 73627725.0,
      "step": 32985
    },
    {
      "entropy": 5.3699620246887205,
      "epoch": 3.705115965631493,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003686283757895193,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22116359025239946,
      "num_tokens": 73639130.0,
      "step": 32990
    },
    {
      "entropy": 5.218710565567017,
      "epoch": 3.705677542539451,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036859186679149555,
      "loss": 4.9323,
      "mean_token_accuracy": 0.2211493268609047,
      "num_tokens": 73650107.0,
      "step": 32995
    },
    {
      "entropy": 5.28927001953125,
      "epoch": 3.7062391194474085,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003685553548134705,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2219952866435051,
      "num_tokens": 73659735.0,
      "step": 33000
    },
    {
      "epoch": 3.7062391194474085,
      "eval_entropy": 5.134980522728409,
      "eval_loss": 5.134953498840332,
      "eval_mean_token_accuracy": 0.21828114114719044,
      "eval_num_tokens": 73659735.0,
      "eval_runtime": 23.8803,
      "eval_samples_per_second": 1298.688,
      "eval_steps_per_second": 162.352,
      "step": 33000
    },
    {
      "entropy": 5.297929191589356,
      "epoch": 3.7068006963553657,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003685188398566066,
      "loss": 4.9639,
      "mean_token_accuracy": 0.21889008283615113,
      "num_tokens": 73670969.0,
      "step": 33005
    },
    {
      "entropy": 5.370891475677491,
      "epoch": 3.7073622732633233,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003684823219220666,
      "loss": 5.0211,
      "mean_token_accuracy": 0.2257177338004112,
      "num_tokens": 73681513.0,
      "step": 33010
    },
    {
      "entropy": 5.431780290603638,
      "epoch": 3.707923850171281,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003684458010110133,
      "loss": 5.1014,
      "mean_token_accuracy": 0.2089888036251068,
      "num_tokens": 73692485.0,
      "step": 33015
    },
    {
      "entropy": 5.407169198989868,
      "epoch": 3.7084854270792382,
      "grad_norm": 1.125,
      "learning_rate": 0.0003684092771246094,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21806580722332,
      "num_tokens": 73702582.0,
      "step": 33020
    },
    {
      "entropy": 5.305132532119751,
      "epoch": 3.709047003987196,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003683727502640178,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22003330290317535,
      "num_tokens": 73713337.0,
      "step": 33025
    },
    {
      "entropy": 5.431763648986816,
      "epoch": 3.7096085808951536,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003683362204304017,
      "loss": 5.1473,
      "mean_token_accuracy": 0.2071835622191429,
      "num_tokens": 73724500.0,
      "step": 33030
    },
    {
      "entropy": 5.298002529144287,
      "epoch": 3.710170157803111,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003682996876249241,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22342725694179535,
      "num_tokens": 73734690.0,
      "step": 33035
    },
    {
      "entropy": 5.253892278671264,
      "epoch": 3.710731734711069,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003682631518487483,
      "loss": 4.841,
      "mean_token_accuracy": 0.22152213007211685,
      "num_tokens": 73744887.0,
      "step": 33040
    },
    {
      "entropy": 5.3777649879455565,
      "epoch": 3.711293311619026,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003682266131030375,
      "loss": 5.0856,
      "mean_token_accuracy": 0.21076125651597977,
      "num_tokens": 73757141.0,
      "step": 33045
    },
    {
      "entropy": 5.344167852401734,
      "epoch": 3.7118548885269838,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036819007138895525,
      "loss": 5.0093,
      "mean_token_accuracy": 0.22449839413166045,
      "num_tokens": 73767802.0,
      "step": 33050
    },
    {
      "entropy": 5.3354206562042235,
      "epoch": 3.7124164654349414,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000368153526707665,
      "loss": 5.0194,
      "mean_token_accuracy": 0.21402618885040284,
      "num_tokens": 73778278.0,
      "step": 33055
    },
    {
      "entropy": 5.313831090927124,
      "epoch": 3.7129780423428986,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003681169790603303,
      "loss": 4.9358,
      "mean_token_accuracy": 0.21787697076797485,
      "num_tokens": 73790065.0,
      "step": 33060
    },
    {
      "entropy": 5.3701554298400875,
      "epoch": 3.7135396192508563,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036808042844811484,
      "loss": 5.0013,
      "mean_token_accuracy": 0.22381238043308258,
      "num_tokens": 73800868.0,
      "step": 33065
    },
    {
      "entropy": 5.368294668197632,
      "epoch": 3.714101196158814,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036804387487218254,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22539226710796356,
      "num_tokens": 73811478.0,
      "step": 33070
    },
    {
      "entropy": 5.274724769592285,
      "epoch": 3.7146627730667716,
      "grad_norm": 0.84375,
      "learning_rate": 0.000368007318333697,
      "loss": 5.018,
      "mean_token_accuracy": 0.22523723393678666,
      "num_tokens": 73823618.0,
      "step": 33075
    },
    {
      "entropy": 5.331714391708374,
      "epoch": 3.7152243499747293,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003679707588338225,
      "loss": 4.9915,
      "mean_token_accuracy": 0.21232471764087676,
      "num_tokens": 73835268.0,
      "step": 33080
    },
    {
      "entropy": 5.368676614761353,
      "epoch": 3.7157859268826865,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000367934196373723,
      "loss": 4.9391,
      "mean_token_accuracy": 0.22048537135124208,
      "num_tokens": 73845877.0,
      "step": 33085
    },
    {
      "entropy": 5.429248905181884,
      "epoch": 3.716347503790644,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036789763095456264,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21320505887269975,
      "num_tokens": 73857038.0,
      "step": 33090
    },
    {
      "entropy": 5.330946063995361,
      "epoch": 3.716909080698602,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003678610625775056,
      "loss": 5.0136,
      "mean_token_accuracy": 0.21711913049221038,
      "num_tokens": 73868439.0,
      "step": 33095
    },
    {
      "entropy": 5.348852825164795,
      "epoch": 3.717470657606559,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036782449124371643,
      "loss": 5.0088,
      "mean_token_accuracy": 0.21376062631607057,
      "num_tokens": 73879427.0,
      "step": 33100
    },
    {
      "entropy": 5.352679491043091,
      "epoch": 3.7180322345145167,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036778791695435935,
      "loss": 4.9877,
      "mean_token_accuracy": 0.21826017498970032,
      "num_tokens": 73890369.0,
      "step": 33105
    },
    {
      "entropy": 5.373785972595215,
      "epoch": 3.7185938114224744,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036775133971059904,
      "loss": 5.0597,
      "mean_token_accuracy": 0.21338299214839934,
      "num_tokens": 73902524.0,
      "step": 33110
    },
    {
      "entropy": 5.252890586853027,
      "epoch": 3.7191553883304316,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003677147595136001,
      "loss": 4.8981,
      "mean_token_accuracy": 0.22793618589639664,
      "num_tokens": 73912163.0,
      "step": 33115
    },
    {
      "entropy": 5.332184553146362,
      "epoch": 3.7197169652383892,
      "grad_norm": 1.0,
      "learning_rate": 0.00036767817636452723,
      "loss": 4.92,
      "mean_token_accuracy": 0.2240413546562195,
      "num_tokens": 73923238.0,
      "step": 33120
    },
    {
      "entropy": 5.359105348587036,
      "epoch": 3.720278542146347,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036764159026454534,
      "loss": 5.0133,
      "mean_token_accuracy": 0.22455979734659196,
      "num_tokens": 73933833.0,
      "step": 33125
    },
    {
      "entropy": 5.380781126022339,
      "epoch": 3.7208401190543046,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003676050012148193,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21501425206661223,
      "num_tokens": 73944850.0,
      "step": 33130
    },
    {
      "entropy": 5.404418849945069,
      "epoch": 3.7214016959622622,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000367568409216514,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21749613881111146,
      "num_tokens": 73954490.0,
      "step": 33135
    },
    {
      "entropy": 5.382616853713989,
      "epoch": 3.7219632728702194,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003675318142707947,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21134851276874542,
      "num_tokens": 73966250.0,
      "step": 33140
    },
    {
      "entropy": 5.375850248336792,
      "epoch": 3.722524849778177,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003674952163788266,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22044012397527696,
      "num_tokens": 73976931.0,
      "step": 33145
    },
    {
      "entropy": 5.35321159362793,
      "epoch": 3.7230864266861348,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036745861554177486,
      "loss": 5.0624,
      "mean_token_accuracy": 0.21140566915273667,
      "num_tokens": 73989813.0,
      "step": 33150
    },
    {
      "entropy": 5.2578624248504635,
      "epoch": 3.723648003594092,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036742201176080496,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22542916983366013,
      "num_tokens": 74000961.0,
      "step": 33155
    },
    {
      "entropy": 5.335231018066406,
      "epoch": 3.7242095805020496,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003673854050370824,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2200678989291191,
      "num_tokens": 74011717.0,
      "step": 33160
    },
    {
      "entropy": 5.269906330108642,
      "epoch": 3.7247711574100073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036734879537177276,
      "loss": 4.938,
      "mean_token_accuracy": 0.2181604191660881,
      "num_tokens": 74022931.0,
      "step": 33165
    },
    {
      "entropy": 5.253876495361328,
      "epoch": 3.725332734317965,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003673121827660415,
      "loss": 4.8585,
      "mean_token_accuracy": 0.2228327661752701,
      "num_tokens": 74033110.0,
      "step": 33170
    },
    {
      "entropy": 5.28944993019104,
      "epoch": 3.7258943112259226,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036727556722105465,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21708866059780121,
      "num_tokens": 74043694.0,
      "step": 33175
    },
    {
      "entropy": 5.341071939468383,
      "epoch": 3.72645588813388,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036723894873797786,
      "loss": 4.9218,
      "mean_token_accuracy": 0.2255815491080284,
      "num_tokens": 74053997.0,
      "step": 33180
    },
    {
      "entropy": 5.352702856063843,
      "epoch": 3.7270174650418375,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036720232731797726,
      "loss": 5.0736,
      "mean_token_accuracy": 0.2094803422689438,
      "num_tokens": 74065562.0,
      "step": 33185
    },
    {
      "entropy": 5.455473184585571,
      "epoch": 3.727579041949795,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036716570296221873,
      "loss": 5.0896,
      "mean_token_accuracy": 0.215694397687912,
      "num_tokens": 74075988.0,
      "step": 33190
    },
    {
      "entropy": 5.346100187301635,
      "epoch": 3.7281406188577524,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036712907567186857,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21989684998989106,
      "num_tokens": 74087644.0,
      "step": 33195
    },
    {
      "entropy": 5.381581354141235,
      "epoch": 3.72870219576571,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003670924454480928,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21901889592409135,
      "num_tokens": 74097841.0,
      "step": 33200
    },
    {
      "entropy": 5.3245875358581545,
      "epoch": 3.7292637726736677,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003670558122920579,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21664955466985703,
      "num_tokens": 74108164.0,
      "step": 33205
    },
    {
      "entropy": 5.341843700408935,
      "epoch": 3.729825349581625,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036701917620493026,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21712402999401093,
      "num_tokens": 74120456.0,
      "step": 33210
    },
    {
      "entropy": 5.372651624679565,
      "epoch": 3.7303869264895826,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003669825371878764,
      "loss": 5.0909,
      "mean_token_accuracy": 0.2175146982073784,
      "num_tokens": 74131023.0,
      "step": 33215
    },
    {
      "entropy": 5.364157199859619,
      "epoch": 3.7309485033975403,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036694589524206275,
      "loss": 5.0757,
      "mean_token_accuracy": 0.2126437708735466,
      "num_tokens": 74141654.0,
      "step": 33220
    },
    {
      "entropy": 5.32896876335144,
      "epoch": 3.731510080305498,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003669092503686562,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2169465020298958,
      "num_tokens": 74152972.0,
      "step": 33225
    },
    {
      "entropy": 5.38514347076416,
      "epoch": 3.7320716572134556,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036687260256882347,
      "loss": 4.9929,
      "mean_token_accuracy": 0.212835593521595,
      "num_tokens": 74164725.0,
      "step": 33230
    },
    {
      "entropy": 5.251781749725342,
      "epoch": 3.732633234121413,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036683595184373147,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2249404937028885,
      "num_tokens": 74175356.0,
      "step": 33235
    },
    {
      "entropy": 5.232719564437867,
      "epoch": 3.7331948110293705,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003667992981945471,
      "loss": 4.897,
      "mean_token_accuracy": 0.22304559201002122,
      "num_tokens": 74186418.0,
      "step": 33240
    },
    {
      "entropy": 5.261322641372681,
      "epoch": 3.733756387937328,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036676264162243744,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21695178598165513,
      "num_tokens": 74199236.0,
      "step": 33245
    },
    {
      "entropy": 5.332357549667359,
      "epoch": 3.7343179648452853,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003667259821285698,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2279285177588463,
      "num_tokens": 74211475.0,
      "step": 33250
    },
    {
      "entropy": 5.2971068859100345,
      "epoch": 3.734879541753243,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003666893197141112,
      "loss": 4.9,
      "mean_token_accuracy": 0.22825241684913636,
      "num_tokens": 74223909.0,
      "step": 33255
    },
    {
      "entropy": 5.385779333114624,
      "epoch": 3.7354411186612007,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003666526543802291,
      "loss": 5.0816,
      "mean_token_accuracy": 0.2069343939423561,
      "num_tokens": 74235060.0,
      "step": 33260
    },
    {
      "entropy": 5.326198148727417,
      "epoch": 3.7360026955691583,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000366615986128091,
      "loss": 4.937,
      "mean_token_accuracy": 0.21986895501613618,
      "num_tokens": 74245930.0,
      "step": 33265
    },
    {
      "entropy": 5.330064201354981,
      "epoch": 3.736564272477116,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003665793149588644,
      "loss": 5.0227,
      "mean_token_accuracy": 0.20789432376623154,
      "num_tokens": 74257337.0,
      "step": 33270
    },
    {
      "entropy": 5.274132490158081,
      "epoch": 3.737125849385073,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036654264087371686,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22381270080804824,
      "num_tokens": 74268217.0,
      "step": 33275
    },
    {
      "entropy": 5.296833133697509,
      "epoch": 3.737687426293031,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00036650596387381615,
      "loss": 5.008,
      "mean_token_accuracy": 0.2140057548880577,
      "num_tokens": 74279842.0,
      "step": 33280
    },
    {
      "entropy": 5.343124914169311,
      "epoch": 3.7382490032009885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036646928396033,
      "loss": 5.0266,
      "mean_token_accuracy": 0.21006491035223007,
      "num_tokens": 74292328.0,
      "step": 33285
    },
    {
      "entropy": 5.418448543548584,
      "epoch": 3.7388105801089457,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036643260113442634,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21710605919361115,
      "num_tokens": 74302706.0,
      "step": 33290
    },
    {
      "entropy": 5.463385725021363,
      "epoch": 3.7393721570169034,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003663959153972732,
      "loss": 5.0878,
      "mean_token_accuracy": 0.20936229974031448,
      "num_tokens": 74314320.0,
      "step": 33295
    },
    {
      "entropy": 5.280169582366943,
      "epoch": 3.739933733924861,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003663592267500388,
      "loss": 4.8488,
      "mean_token_accuracy": 0.22713393419981004,
      "num_tokens": 74325324.0,
      "step": 33300
    },
    {
      "entropy": 5.324476194381714,
      "epoch": 3.7404953108328187,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003663225351938911,
      "loss": 5.0578,
      "mean_token_accuracy": 0.21340576112270354,
      "num_tokens": 74337623.0,
      "step": 33305
    },
    {
      "entropy": 5.2844452381134035,
      "epoch": 3.741056887740776,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003662858407299984,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2189347431063652,
      "num_tokens": 74348971.0,
      "step": 33310
    },
    {
      "entropy": 5.332478046417236,
      "epoch": 3.7416184646487336,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036624914335952924,
      "loss": 5.0307,
      "mean_token_accuracy": 0.2137461468577385,
      "num_tokens": 74360162.0,
      "step": 33315
    },
    {
      "entropy": 5.4756755352020265,
      "epoch": 3.7421800415566913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036621244308365195,
      "loss": 5.111,
      "mean_token_accuracy": 0.2141265019774437,
      "num_tokens": 74371180.0,
      "step": 33320
    },
    {
      "entropy": 5.328403806686401,
      "epoch": 3.742741618464649,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036617573990353503,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21520985662937164,
      "num_tokens": 74382854.0,
      "step": 33325
    },
    {
      "entropy": 5.348351240158081,
      "epoch": 3.743303195372606,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003661390338203472,
      "loss": 5.1179,
      "mean_token_accuracy": 0.20596520602703094,
      "num_tokens": 74394266.0,
      "step": 33330
    },
    {
      "entropy": 5.41334228515625,
      "epoch": 3.743864772280564,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0003661023248352572,
      "loss": 5.1181,
      "mean_token_accuracy": 0.20705696940422058,
      "num_tokens": 74406275.0,
      "step": 33335
    },
    {
      "entropy": 5.340884399414063,
      "epoch": 3.7444263491885215,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003660656129494339,
      "loss": 4.9422,
      "mean_token_accuracy": 0.21584748178720475,
      "num_tokens": 74417150.0,
      "step": 33340
    },
    {
      "entropy": 5.318228340148925,
      "epoch": 3.7449879260964787,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036602889816404613,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2254180908203125,
      "num_tokens": 74427997.0,
      "step": 33345
    },
    {
      "entropy": 5.397578144073487,
      "epoch": 3.7455495030044363,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036599218048026294,
      "loss": 5.0845,
      "mean_token_accuracy": 0.20841658264398574,
      "num_tokens": 74439278.0,
      "step": 33350
    },
    {
      "entropy": 5.363477182388306,
      "epoch": 3.746111079912394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036595545989925345,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21666987538337706,
      "num_tokens": 74450372.0,
      "step": 33355
    },
    {
      "entropy": 5.350135421752929,
      "epoch": 3.7466726568203517,
      "grad_norm": 1.0,
      "learning_rate": 0.00036591873642218676,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22061900645494462,
      "num_tokens": 74460040.0,
      "step": 33360
    },
    {
      "entropy": 5.364750528335572,
      "epoch": 3.7472342337283093,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036588201005023223,
      "loss": 5.0255,
      "mean_token_accuracy": 0.22008373141288756,
      "num_tokens": 74472291.0,
      "step": 33365
    },
    {
      "entropy": 5.375980424880981,
      "epoch": 3.7477958106362665,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003658452807845593,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21814474761486052,
      "num_tokens": 74482373.0,
      "step": 33370
    },
    {
      "entropy": 5.3062193393707275,
      "epoch": 3.748357387544224,
      "grad_norm": 1.0,
      "learning_rate": 0.00036580854862633747,
      "loss": 4.9884,
      "mean_token_accuracy": 0.2239741101861,
      "num_tokens": 74494240.0,
      "step": 33375
    },
    {
      "entropy": 5.390754461288452,
      "epoch": 3.748918964452182,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003657718135767362,
      "loss": 5.0297,
      "mean_token_accuracy": 0.22130923569202424,
      "num_tokens": 74504505.0,
      "step": 33380
    },
    {
      "entropy": 5.35042142868042,
      "epoch": 3.749480541360139,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003657350756369251,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22672809958457946,
      "num_tokens": 74515467.0,
      "step": 33385
    },
    {
      "entropy": 5.297260427474976,
      "epoch": 3.7500421182680967,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036569833480807403,
      "loss": 4.9639,
      "mean_token_accuracy": 0.21923446357250215,
      "num_tokens": 74527502.0,
      "step": 33390
    },
    {
      "entropy": 5.366000127792359,
      "epoch": 3.7506036951760544,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003656615910913528,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2117556244134903,
      "num_tokens": 74539491.0,
      "step": 33395
    },
    {
      "entropy": 5.376784420013427,
      "epoch": 3.751165272084012,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036562484448793135,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2131249025464058,
      "num_tokens": 74550847.0,
      "step": 33400
    },
    {
      "entropy": 5.390685033798218,
      "epoch": 3.7517268489919693,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003655880949989798,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21781370788812637,
      "num_tokens": 74561180.0,
      "step": 33405
    },
    {
      "entropy": 5.380361890792846,
      "epoch": 3.752288425899927,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036555134262566806,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21683878004550933,
      "num_tokens": 74571557.0,
      "step": 33410
    },
    {
      "entropy": 5.346853828430175,
      "epoch": 3.7528500028078846,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003655145873691665,
      "loss": 5.0338,
      "mean_token_accuracy": 0.21237218081951142,
      "num_tokens": 74582614.0,
      "step": 33415
    },
    {
      "entropy": 5.294654321670532,
      "epoch": 3.7534115797158423,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036547782923064537,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21536062210798262,
      "num_tokens": 74593176.0,
      "step": 33420
    },
    {
      "entropy": 5.338423156738282,
      "epoch": 3.7539731566237995,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036544106821127505,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22635134011507035,
      "num_tokens": 74604601.0,
      "step": 33425
    },
    {
      "entropy": 5.401857376098633,
      "epoch": 3.754534733531757,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000365404304312226,
      "loss": 5.0879,
      "mean_token_accuracy": 0.20920022577047348,
      "num_tokens": 74615591.0,
      "step": 33430
    },
    {
      "entropy": 5.330974483489991,
      "epoch": 3.755096310439715,
      "grad_norm": 1.0,
      "learning_rate": 0.00036536753753466897,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21911484152078628,
      "num_tokens": 74626905.0,
      "step": 33435
    },
    {
      "entropy": 5.379442501068115,
      "epoch": 3.755657887347672,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003653307678797744,
      "loss": 5.0224,
      "mean_token_accuracy": 0.21501513868570327,
      "num_tokens": 74638201.0,
      "step": 33440
    },
    {
      "entropy": 5.348442697525025,
      "epoch": 3.7562194642556297,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003652939953487131,
      "loss": 5.0602,
      "mean_token_accuracy": 0.21413756012916565,
      "num_tokens": 74649121.0,
      "step": 33445
    },
    {
      "entropy": 5.295163631439209,
      "epoch": 3.7567810411635874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036525721994265606,
      "loss": 4.8914,
      "mean_token_accuracy": 0.2232422947883606,
      "num_tokens": 74659482.0,
      "step": 33450
    },
    {
      "entropy": 5.243085956573486,
      "epoch": 3.757342618071545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003652204416627741,
      "loss": 4.8931,
      "mean_token_accuracy": 0.2254742830991745,
      "num_tokens": 74671587.0,
      "step": 33455
    },
    {
      "entropy": 5.362754487991333,
      "epoch": 3.7579041949795027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036518366051023824,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21980946958065034,
      "num_tokens": 74682753.0,
      "step": 33460
    },
    {
      "entropy": 5.311546516418457,
      "epoch": 3.75846577188746,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003651468764862197,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21651186048984528,
      "num_tokens": 74692670.0,
      "step": 33465
    },
    {
      "entropy": 5.282718372344971,
      "epoch": 3.7590273487954176,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036511008959188966,
      "loss": 5.0411,
      "mean_token_accuracy": 0.2124329388141632,
      "num_tokens": 74703832.0,
      "step": 33470
    },
    {
      "entropy": 5.329083013534546,
      "epoch": 3.759588925703375,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036507329982841935,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2222157284617424,
      "num_tokens": 74716577.0,
      "step": 33475
    },
    {
      "entropy": 5.399462413787842,
      "epoch": 3.7601505026113324,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003650365071969803,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21427347958087922,
      "num_tokens": 74727458.0,
      "step": 33480
    },
    {
      "entropy": 5.269257640838623,
      "epoch": 3.76071207951929,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003649997116987439,
      "loss": 4.8786,
      "mean_token_accuracy": 0.22561454325914382,
      "num_tokens": 74739049.0,
      "step": 33485
    },
    {
      "entropy": 5.314712762832642,
      "epoch": 3.7612736564272478,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003649629133348818,
      "loss": 5.0358,
      "mean_token_accuracy": 0.22253775149583815,
      "num_tokens": 74749509.0,
      "step": 33490
    },
    {
      "entropy": 5.360621356964112,
      "epoch": 3.7618352333352054,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00036492611210656574,
      "loss": 4.9868,
      "mean_token_accuracy": 0.22151993960142136,
      "num_tokens": 74762067.0,
      "step": 33495
    },
    {
      "entropy": 5.351818323135376,
      "epoch": 3.7623968102431626,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003648893080149673,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2268408939242363,
      "num_tokens": 74773106.0,
      "step": 33500
    },
    {
      "entropy": 5.371072101593017,
      "epoch": 3.7629583871511203,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00036485250106125844,
      "loss": 5.0792,
      "mean_token_accuracy": 0.21341005861759185,
      "num_tokens": 74785464.0,
      "step": 33505
    },
    {
      "entropy": 5.363032436370849,
      "epoch": 3.763519964059078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000364815691246611,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21929049491882324,
      "num_tokens": 74795939.0,
      "step": 33510
    },
    {
      "entropy": 5.315131044387817,
      "epoch": 3.7640815409670356,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036477887857219726,
      "loss": 4.9831,
      "mean_token_accuracy": 0.22232548594474794,
      "num_tokens": 74807036.0,
      "step": 33515
    },
    {
      "entropy": 5.3474085330963135,
      "epoch": 3.764643117874993,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036474206303918914,
      "loss": 5.0033,
      "mean_token_accuracy": 0.21888236999511718,
      "num_tokens": 74818402.0,
      "step": 33520
    },
    {
      "entropy": 5.324261093139649,
      "epoch": 3.7652046947829505,
      "grad_norm": 0.96875,
      "learning_rate": 0.000364705244648759,
      "loss": 4.9814,
      "mean_token_accuracy": 0.21829487085342408,
      "num_tokens": 74830246.0,
      "step": 33525
    },
    {
      "entropy": 5.29680290222168,
      "epoch": 3.765766271690908,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003646684234020789,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2230500802397728,
      "num_tokens": 74840690.0,
      "step": 33530
    },
    {
      "entropy": 5.39497480392456,
      "epoch": 3.7663278485988654,
      "grad_norm": 1.0,
      "learning_rate": 0.00036463159930032163,
      "loss": 4.9449,
      "mean_token_accuracy": 0.23221812695264815,
      "num_tokens": 74851807.0,
      "step": 33535
    },
    {
      "entropy": 5.326076364517212,
      "epoch": 3.766889425506823,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036459477234465935,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22159538716077803,
      "num_tokens": 74863028.0,
      "step": 33540
    },
    {
      "entropy": 5.3141998767852785,
      "epoch": 3.7674510024147807,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00036455794253626486,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22445466071367265,
      "num_tokens": 74874674.0,
      "step": 33545
    },
    {
      "entropy": 5.386721134185791,
      "epoch": 3.7680125793227384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036452110987631066,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20978295356035231,
      "num_tokens": 74885791.0,
      "step": 33550
    },
    {
      "entropy": 5.347938585281372,
      "epoch": 3.768574156230696,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036448427436596966,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21809386909008027,
      "num_tokens": 74897422.0,
      "step": 33555
    },
    {
      "entropy": 5.37655611038208,
      "epoch": 3.7691357331386532,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003644474360064146,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21207889914512634,
      "num_tokens": 74908680.0,
      "step": 33560
    },
    {
      "entropy": 5.339879322052002,
      "epoch": 3.769697310046611,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003644105947988185,
      "loss": 4.9864,
      "mean_token_accuracy": 0.22338478714227678,
      "num_tokens": 74919573.0,
      "step": 33565
    },
    {
      "entropy": 5.300648260116577,
      "epoch": 3.7702588869545686,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036437375074435446,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2191333070397377,
      "num_tokens": 74930507.0,
      "step": 33570
    },
    {
      "entropy": 5.329343461990357,
      "epoch": 3.770820463862526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003643369038441955,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21918968707323075,
      "num_tokens": 74942436.0,
      "step": 33575
    },
    {
      "entropy": 5.215061664581299,
      "epoch": 3.7713820407704834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036430005409951476,
      "loss": 4.873,
      "mean_token_accuracy": 0.22794498652219772,
      "num_tokens": 74953717.0,
      "step": 33580
    },
    {
      "entropy": 5.327362394332885,
      "epoch": 3.771943617678441,
      "grad_norm": 1.0,
      "learning_rate": 0.00036426320151148575,
      "loss": 5.0179,
      "mean_token_accuracy": 0.2222522810101509,
      "num_tokens": 74963878.0,
      "step": 33585
    },
    {
      "entropy": 5.360372018814087,
      "epoch": 3.7725051945863988,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036422634608128184,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2316638022661209,
      "num_tokens": 74976239.0,
      "step": 33590
    },
    {
      "entropy": 5.3766467571258545,
      "epoch": 3.773066771494356,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036418948781007636,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21437846422195433,
      "num_tokens": 74987694.0,
      "step": 33595
    },
    {
      "entropy": 5.379682302474976,
      "epoch": 3.7736283484023136,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036415262669904305,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21898936480283737,
      "num_tokens": 74998514.0,
      "step": 33600
    },
    {
      "entropy": 5.386404418945313,
      "epoch": 3.7741899253102713,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036411576274935546,
      "loss": 4.9899,
      "mean_token_accuracy": 0.21745138317346574,
      "num_tokens": 75010438.0,
      "step": 33605
    },
    {
      "entropy": 5.287320327758789,
      "epoch": 3.774751502218229,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036407889596218757,
      "loss": 4.8976,
      "mean_token_accuracy": 0.2214828610420227,
      "num_tokens": 75021280.0,
      "step": 33610
    },
    {
      "entropy": 5.262604761123657,
      "epoch": 3.775313079126186,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000364042026338713,
      "loss": 4.9107,
      "mean_token_accuracy": 0.22415657192468644,
      "num_tokens": 75032746.0,
      "step": 33615
    },
    {
      "entropy": 5.374963903427124,
      "epoch": 3.775874656034144,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00036400515388010574,
      "loss": 5.0946,
      "mean_token_accuracy": 0.21346224546432496,
      "num_tokens": 75044227.0,
      "step": 33620
    },
    {
      "entropy": 5.435530281066894,
      "epoch": 3.7764362329421015,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003639682785875399,
      "loss": 5.0884,
      "mean_token_accuracy": 0.21298457235097884,
      "num_tokens": 75053974.0,
      "step": 33625
    },
    {
      "entropy": 5.30731086730957,
      "epoch": 3.7769978098500587,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036393140046218957,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2232253447175026,
      "num_tokens": 75064539.0,
      "step": 33630
    },
    {
      "entropy": 5.341500663757325,
      "epoch": 3.7775593867580164,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00036389451950522895,
      "loss": 4.943,
      "mean_token_accuracy": 0.2232791781425476,
      "num_tokens": 75075229.0,
      "step": 33635
    },
    {
      "entropy": 5.416627311706543,
      "epoch": 3.778120963665974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036385763571783236,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21555475443601607,
      "num_tokens": 75085347.0,
      "step": 33640
    },
    {
      "entropy": 5.388874626159668,
      "epoch": 3.7786825405739317,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003638207491011742,
      "loss": 4.9859,
      "mean_token_accuracy": 0.221958926320076,
      "num_tokens": 75096141.0,
      "step": 33645
    },
    {
      "entropy": 5.337704038619995,
      "epoch": 3.7792441174818894,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003637838596564289,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2268048956990242,
      "num_tokens": 75107749.0,
      "step": 33650
    },
    {
      "entropy": 5.326472902297974,
      "epoch": 3.7798056943898466,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036374696738477105,
      "loss": 5.0031,
      "mean_token_accuracy": 0.21911714673042298,
      "num_tokens": 75118644.0,
      "step": 33655
    },
    {
      "entropy": 5.297040224075317,
      "epoch": 3.7803672712978043,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003637100722873754,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2139420837163925,
      "num_tokens": 75130534.0,
      "step": 33660
    },
    {
      "entropy": 5.410743761062622,
      "epoch": 3.780928848205762,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003636731743654166,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22056004106998445,
      "num_tokens": 75142159.0,
      "step": 33665
    },
    {
      "entropy": 5.380608797073364,
      "epoch": 3.781490425113719,
      "grad_norm": 1.0,
      "learning_rate": 0.00036363627362006955,
      "loss": 4.9388,
      "mean_token_accuracy": 0.2237120106816292,
      "num_tokens": 75153603.0,
      "step": 33670
    },
    {
      "entropy": 5.260287094116211,
      "epoch": 3.782052002021677,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003635993700525092,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22387394607067107,
      "num_tokens": 75164538.0,
      "step": 33675
    },
    {
      "entropy": 5.337713527679443,
      "epoch": 3.7826135789296345,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036356246366391054,
      "loss": 5.0384,
      "mean_token_accuracy": 0.21903070360422133,
      "num_tokens": 75175030.0,
      "step": 33680
    },
    {
      "entropy": 5.457322597503662,
      "epoch": 3.783175155837592,
      "grad_norm": 1.0,
      "learning_rate": 0.00036352555445544864,
      "loss": 5.0442,
      "mean_token_accuracy": 0.22232666462659836,
      "num_tokens": 75187116.0,
      "step": 33685
    },
    {
      "entropy": 5.386841106414795,
      "epoch": 3.78373673274555,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036348864242829874,
      "loss": 4.954,
      "mean_token_accuracy": 0.22358518987894058,
      "num_tokens": 75197247.0,
      "step": 33690
    },
    {
      "entropy": 5.269994306564331,
      "epoch": 3.784298309653507,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036345172758363623,
      "loss": 4.9959,
      "mean_token_accuracy": 0.2207615166902542,
      "num_tokens": 75208531.0,
      "step": 33695
    },
    {
      "entropy": 5.326676845550537,
      "epoch": 3.7848598865614647,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003634148099226364,
      "loss": 5.06,
      "mean_token_accuracy": 0.2155129238963127,
      "num_tokens": 75221061.0,
      "step": 33700
    },
    {
      "entropy": 5.359136629104614,
      "epoch": 3.7854214634694223,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036337788944647466,
      "loss": 4.9551,
      "mean_token_accuracy": 0.22132600098848343,
      "num_tokens": 75232897.0,
      "step": 33705
    },
    {
      "entropy": 5.4366497039794925,
      "epoch": 3.7859830403773795,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003633409661563267,
      "loss": 5.0316,
      "mean_token_accuracy": 0.21436087638139725,
      "num_tokens": 75245228.0,
      "step": 33710
    },
    {
      "entropy": 5.357966375350952,
      "epoch": 3.786544617285337,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036330404005336807,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2196967974305153,
      "num_tokens": 75256486.0,
      "step": 33715
    },
    {
      "entropy": 5.302194404602051,
      "epoch": 3.787106194193295,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003632671111387746,
      "loss": 4.9502,
      "mean_token_accuracy": 0.22759285420179368,
      "num_tokens": 75266439.0,
      "step": 33720
    },
    {
      "entropy": 5.346197843551636,
      "epoch": 3.787667771101252,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036323017941372203,
      "loss": 4.9862,
      "mean_token_accuracy": 0.22110812813043595,
      "num_tokens": 75276550.0,
      "step": 33725
    },
    {
      "entropy": 5.291669654846191,
      "epoch": 3.7882293480092097,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003631932448793864,
      "loss": 4.989,
      "mean_token_accuracy": 0.21875206679105758,
      "num_tokens": 75287061.0,
      "step": 33730
    },
    {
      "entropy": 5.375171518325805,
      "epoch": 3.7887909249171674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003631563075369436,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21434696912765502,
      "num_tokens": 75298393.0,
      "step": 33735
    },
    {
      "entropy": 5.341448211669922,
      "epoch": 3.789352501825125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036311936738756977,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21580106168985366,
      "num_tokens": 75308153.0,
      "step": 33740
    },
    {
      "entropy": 5.279288053512573,
      "epoch": 3.7899140787330827,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003630824244324411,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22393419742584228,
      "num_tokens": 75319383.0,
      "step": 33745
    },
    {
      "entropy": 5.357103252410889,
      "epoch": 3.79047565564104,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003630454786727339,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21319998502731324,
      "num_tokens": 75329736.0,
      "step": 33750
    },
    {
      "entropy": 5.450018119812012,
      "epoch": 3.7910372325489976,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003630085301096245,
      "loss": 5.1023,
      "mean_token_accuracy": 0.20908478200435637,
      "num_tokens": 75341451.0,
      "step": 33755
    },
    {
      "entropy": 5.242102336883545,
      "epoch": 3.7915988094569553,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036297157874428933,
      "loss": 4.8568,
      "mean_token_accuracy": 0.2239123538136482,
      "num_tokens": 75352664.0,
      "step": 33760
    },
    {
      "entropy": 5.339789152145386,
      "epoch": 3.7921603863649125,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036293462457790505,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20972893089056016,
      "num_tokens": 75362847.0,
      "step": 33765
    },
    {
      "entropy": 5.329607057571411,
      "epoch": 3.79272196327287,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036289766761164814,
      "loss": 5.0253,
      "mean_token_accuracy": 0.21408329755067826,
      "num_tokens": 75373344.0,
      "step": 33770
    },
    {
      "entropy": 5.312130117416382,
      "epoch": 3.793283540180828,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036286070784669543,
      "loss": 4.975,
      "mean_token_accuracy": 0.21716601103544236,
      "num_tokens": 75384262.0,
      "step": 33775
    },
    {
      "entropy": 5.409479331970215,
      "epoch": 3.7938451170887855,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036282374528422373,
      "loss": 5.026,
      "mean_token_accuracy": 0.22436884939670562,
      "num_tokens": 75395817.0,
      "step": 33780
    },
    {
      "entropy": 5.294265556335449,
      "epoch": 3.794406693996743,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036278677992540993,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22832961678504943,
      "num_tokens": 75406452.0,
      "step": 33785
    },
    {
      "entropy": 5.3391355037689205,
      "epoch": 3.7949682709047003,
      "grad_norm": 0.984375,
      "learning_rate": 0.000362749811771431,
      "loss": 4.9948,
      "mean_token_accuracy": 0.21619293093681335,
      "num_tokens": 75418638.0,
      "step": 33790
    },
    {
      "entropy": 5.393720388412476,
      "epoch": 3.795529847812658,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003627128408234641,
      "loss": 4.9979,
      "mean_token_accuracy": 0.2214144378900528,
      "num_tokens": 75430024.0,
      "step": 33795
    },
    {
      "entropy": 5.419979238510132,
      "epoch": 3.7960914247206157,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003626758670826861,
      "loss": 5.141,
      "mean_token_accuracy": 0.2130550280213356,
      "num_tokens": 75442641.0,
      "step": 33800
    },
    {
      "entropy": 5.394854545593262,
      "epoch": 3.796653001628573,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003626388905502747,
      "loss": 5.0808,
      "mean_token_accuracy": 0.2144799068570137,
      "num_tokens": 75453245.0,
      "step": 33805
    },
    {
      "entropy": 5.345727586746216,
      "epoch": 3.7972145785365305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000362601911227407,
      "loss": 4.9548,
      "mean_token_accuracy": 0.217292220890522,
      "num_tokens": 75464243.0,
      "step": 33810
    },
    {
      "entropy": 5.343364572525024,
      "epoch": 3.797776155444488,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003625649291152604,
      "loss": 4.9869,
      "mean_token_accuracy": 0.2220085382461548,
      "num_tokens": 75474348.0,
      "step": 33815
    },
    {
      "entropy": 5.351348447799682,
      "epoch": 3.7983377323524454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036252794421501257,
      "loss": 5.0602,
      "mean_token_accuracy": 0.22099349796772003,
      "num_tokens": 75486903.0,
      "step": 33820
    },
    {
      "entropy": 5.3609991550445555,
      "epoch": 3.798899309260403,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036249095652784096,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21950208842754365,
      "num_tokens": 75498198.0,
      "step": 33825
    },
    {
      "entropy": 5.33484115600586,
      "epoch": 3.7994608861683608,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036245396605492344,
      "loss": 5.026,
      "mean_token_accuracy": 0.2140579864382744,
      "num_tokens": 75509787.0,
      "step": 33830
    },
    {
      "entropy": 5.285181951522827,
      "epoch": 3.8000224630763184,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003624169727974376,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22876724004745483,
      "num_tokens": 75521706.0,
      "step": 33835
    },
    {
      "entropy": 5.3039140701293945,
      "epoch": 3.800584039984276,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036237997675656167,
      "loss": 5.0125,
      "mean_token_accuracy": 0.22284515649080278,
      "num_tokens": 75532432.0,
      "step": 33840
    },
    {
      "entropy": 5.460378694534302,
      "epoch": 3.8011456168922333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003623429779334732,
      "loss": 5.097,
      "mean_token_accuracy": 0.20875541865825653,
      "num_tokens": 75543473.0,
      "step": 33845
    },
    {
      "entropy": 5.3762613296508786,
      "epoch": 3.801707193800191,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036230597632935053,
      "loss": 4.9547,
      "mean_token_accuracy": 0.2239344298839569,
      "num_tokens": 75554065.0,
      "step": 33850
    },
    {
      "entropy": 5.2829936981201175,
      "epoch": 3.8022687707081486,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003622689719453717,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2276305839419365,
      "num_tokens": 75564823.0,
      "step": 33855
    },
    {
      "entropy": 5.299142551422119,
      "epoch": 3.802830347616106,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00036223196478271495,
      "loss": 5.0544,
      "mean_token_accuracy": 0.21453292965888976,
      "num_tokens": 75576062.0,
      "step": 33860
    },
    {
      "entropy": 5.383856105804443,
      "epoch": 3.8033919245240635,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003621949548425585,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21656244844198227,
      "num_tokens": 75587038.0,
      "step": 33865
    },
    {
      "entropy": 5.360256195068359,
      "epoch": 3.803953501432021,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000362157942126081,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21545762121677398,
      "num_tokens": 75598306.0,
      "step": 33870
    },
    {
      "entropy": 5.334410667419434,
      "epoch": 3.804515078339979,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003621209266344608,
      "loss": 4.9945,
      "mean_token_accuracy": 0.22085247933864594,
      "num_tokens": 75609432.0,
      "step": 33875
    },
    {
      "entropy": 5.2666271209716795,
      "epoch": 3.8050766552479365,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003620839083688765,
      "loss": 4.8998,
      "mean_token_accuracy": 0.22624107599258422,
      "num_tokens": 75619936.0,
      "step": 33880
    },
    {
      "entropy": 5.372866582870484,
      "epoch": 3.8056382321558937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003620468873305068,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22565543949604033,
      "num_tokens": 75630271.0,
      "step": 33885
    },
    {
      "entropy": 5.330533075332641,
      "epoch": 3.8061998090638514,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036200986352053036,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2226680338382721,
      "num_tokens": 75639859.0,
      "step": 33890
    },
    {
      "entropy": 5.326581954956055,
      "epoch": 3.806761385971809,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036197283694012615,
      "loss": 4.9595,
      "mean_token_accuracy": 0.22294404208660126,
      "num_tokens": 75650610.0,
      "step": 33895
    },
    {
      "entropy": 5.378152704238891,
      "epoch": 3.8073229628797662,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003619358075904732,
      "loss": 4.9711,
      "mean_token_accuracy": 0.21770093590021133,
      "num_tokens": 75661604.0,
      "step": 33900
    },
    {
      "entropy": 5.383124351501465,
      "epoch": 3.807884539787724,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00036189877547275023,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2166071876883507,
      "num_tokens": 75674173.0,
      "step": 33905
    },
    {
      "entropy": 5.383337736129761,
      "epoch": 3.8084461166956816,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036186174058813666,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21839091777801514,
      "num_tokens": 75684579.0,
      "step": 33910
    },
    {
      "entropy": 5.3164268970489506,
      "epoch": 3.8090076936036388,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036182470293781154,
      "loss": 4.9735,
      "mean_token_accuracy": 0.22771539986133577,
      "num_tokens": 75696022.0,
      "step": 33915
    },
    {
      "entropy": 5.374322319030762,
      "epoch": 3.8095692705115964,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003617876625229543,
      "loss": 5.0325,
      "mean_token_accuracy": 0.2172514721751213,
      "num_tokens": 75708255.0,
      "step": 33920
    },
    {
      "entropy": 5.298708057403564,
      "epoch": 3.810130847419554,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003617506193447442,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2253175362944603,
      "num_tokens": 75719069.0,
      "step": 33925
    },
    {
      "entropy": 5.26147608757019,
      "epoch": 3.8106924243275118,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003617135734043606,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22868882566690446,
      "num_tokens": 75730078.0,
      "step": 33930
    },
    {
      "entropy": 5.28793420791626,
      "epoch": 3.8112540012354694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003616765247029833,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21778174340724946,
      "num_tokens": 75741910.0,
      "step": 33935
    },
    {
      "entropy": 5.313697671890258,
      "epoch": 3.8118155781434266,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036163947324179186,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21482069343328475,
      "num_tokens": 75752261.0,
      "step": 33940
    },
    {
      "entropy": 5.3549542903900145,
      "epoch": 3.8123771550513843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036160241902196587,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21821476072072982,
      "num_tokens": 75762442.0,
      "step": 33945
    },
    {
      "entropy": 5.296267986297607,
      "epoch": 3.812938731959342,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036156536204468535,
      "loss": 4.9094,
      "mean_token_accuracy": 0.2197491019964218,
      "num_tokens": 75773550.0,
      "step": 33950
    },
    {
      "entropy": 5.41727499961853,
      "epoch": 3.813500308867299,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036152830231113013,
      "loss": 5.082,
      "mean_token_accuracy": 0.21125413626432418,
      "num_tokens": 75785730.0,
      "step": 33955
    },
    {
      "entropy": 5.2589795112609865,
      "epoch": 3.814061885775257,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003614912398224802,
      "loss": 4.7953,
      "mean_token_accuracy": 0.22894089818000793,
      "num_tokens": 75795868.0,
      "step": 33960
    },
    {
      "entropy": 5.313398742675782,
      "epoch": 3.8146234626832145,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003614541745799156,
      "loss": 5.0866,
      "mean_token_accuracy": 0.21187548488378524,
      "num_tokens": 75808318.0,
      "step": 33965
    },
    {
      "entropy": 5.367811870574951,
      "epoch": 3.815185039591172,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003614171065846166,
      "loss": 5.017,
      "mean_token_accuracy": 0.21753253042697906,
      "num_tokens": 75819930.0,
      "step": 33970
    },
    {
      "entropy": 5.465809392929077,
      "epoch": 3.81574661649913,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003613800358377635,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21721564531326293,
      "num_tokens": 75831814.0,
      "step": 33975
    },
    {
      "entropy": 5.335186004638672,
      "epoch": 3.816308193407087,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003613429623405364,
      "loss": 4.9768,
      "mean_token_accuracy": 0.21474671959877015,
      "num_tokens": 75842610.0,
      "step": 33980
    },
    {
      "entropy": 5.281474924087524,
      "epoch": 3.8168697703150447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036130588609411597,
      "loss": 5.0553,
      "mean_token_accuracy": 0.21472473293542862,
      "num_tokens": 75854582.0,
      "step": 33985
    },
    {
      "entropy": 5.345591831207275,
      "epoch": 3.8174313472230024,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036126880709968264,
      "loss": 4.931,
      "mean_token_accuracy": 0.22597001641988754,
      "num_tokens": 75865715.0,
      "step": 33990
    },
    {
      "entropy": 5.388801717758179,
      "epoch": 3.8179929241309596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036123172535841705,
      "loss": 4.9555,
      "mean_token_accuracy": 0.22396664023399354,
      "num_tokens": 75876348.0,
      "step": 33995
    },
    {
      "entropy": 5.386597299575806,
      "epoch": 3.8185545010389172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003611946408714998,
      "loss": 5.0493,
      "mean_token_accuracy": 0.21565669924020767,
      "num_tokens": 75888316.0,
      "step": 34000
    },
    {
      "entropy": 5.333922910690307,
      "epoch": 3.819116077946875,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036115755364011186,
      "loss": 4.9829,
      "mean_token_accuracy": 0.22253286987543106,
      "num_tokens": 75898872.0,
      "step": 34005
    },
    {
      "entropy": 5.397887706756592,
      "epoch": 3.819677654854832,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000361120463665434,
      "loss": 5.0423,
      "mean_token_accuracy": 0.21444333344697952,
      "num_tokens": 75909768.0,
      "step": 34010
    },
    {
      "entropy": 5.40838394165039,
      "epoch": 3.82023923176279,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003610833709486471,
      "loss": 5.0409,
      "mean_token_accuracy": 0.2120228201150894,
      "num_tokens": 75921488.0,
      "step": 34015
    },
    {
      "entropy": 5.349220180511475,
      "epoch": 3.8208008086707475,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003610462754909323,
      "loss": 4.9794,
      "mean_token_accuracy": 0.21451006978750228,
      "num_tokens": 75933619.0,
      "step": 34020
    },
    {
      "entropy": 5.320299434661865,
      "epoch": 3.821362385578705,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036100917729347073,
      "loss": 4.9,
      "mean_token_accuracy": 0.2216912806034088,
      "num_tokens": 75943369.0,
      "step": 34025
    },
    {
      "entropy": 5.387540817260742,
      "epoch": 3.8219239624866628,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036097207635744363,
      "loss": 5.0826,
      "mean_token_accuracy": 0.21220593303442,
      "num_tokens": 75954793.0,
      "step": 34030
    },
    {
      "entropy": 5.305421924591064,
      "epoch": 3.82248553939462,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036093497268403215,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22392136007547378,
      "num_tokens": 75965765.0,
      "step": 34035
    },
    {
      "entropy": 5.309311819076538,
      "epoch": 3.8230471163025777,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000360897866274418,
      "loss": 4.9656,
      "mean_token_accuracy": 0.22120587825775145,
      "num_tokens": 75976435.0,
      "step": 34040
    },
    {
      "entropy": 5.360985279083252,
      "epoch": 3.8236086932105353,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036086075712978233,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21701428294181824,
      "num_tokens": 75987379.0,
      "step": 34045
    },
    {
      "entropy": 5.423221635818481,
      "epoch": 3.8241702701184925,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036082364525130684,
      "loss": 5.1176,
      "mean_token_accuracy": 0.20473133027553558,
      "num_tokens": 75999073.0,
      "step": 34050
    },
    {
      "entropy": 5.355525302886963,
      "epoch": 3.82473184702645,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003607865306401732,
      "loss": 4.9639,
      "mean_token_accuracy": 0.21527751088142394,
      "num_tokens": 76009490.0,
      "step": 34055
    },
    {
      "entropy": 5.3245758533477785,
      "epoch": 3.825293423934408,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036074941329756327,
      "loss": 5.0052,
      "mean_token_accuracy": 0.22115110158920287,
      "num_tokens": 76021616.0,
      "step": 34060
    },
    {
      "entropy": 5.332374572753906,
      "epoch": 3.8258550008423655,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003607122932246587,
      "loss": 4.997,
      "mean_token_accuracy": 0.22266539484262465,
      "num_tokens": 76031031.0,
      "step": 34065
    },
    {
      "entropy": 5.254789066314697,
      "epoch": 3.826416577750323,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00036067517042264147,
      "loss": 4.8282,
      "mean_token_accuracy": 0.2270468607544899,
      "num_tokens": 76041433.0,
      "step": 34070
    },
    {
      "entropy": 5.341865825653076,
      "epoch": 3.8269781546582804,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003606380448926936,
      "loss": 4.998,
      "mean_token_accuracy": 0.21849806159734725,
      "num_tokens": 76052359.0,
      "step": 34075
    },
    {
      "entropy": 5.352876996994018,
      "epoch": 3.827539731566238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003606009166359972,
      "loss": 4.9423,
      "mean_token_accuracy": 0.22541649937629699,
      "num_tokens": 76062857.0,
      "step": 34080
    },
    {
      "entropy": 5.472993659973144,
      "epoch": 3.8281013084741957,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036056378565373434,
      "loss": 5.1585,
      "mean_token_accuracy": 0.20510734766721725,
      "num_tokens": 76074937.0,
      "step": 34085
    },
    {
      "entropy": 5.2900432586669925,
      "epoch": 3.828662885382153,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003605266519470874,
      "loss": 4.915,
      "mean_token_accuracy": 0.2211875006556511,
      "num_tokens": 76087434.0,
      "step": 34090
    },
    {
      "entropy": 5.31575837135315,
      "epoch": 3.8292244622901106,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003604895155172386,
      "loss": 4.9557,
      "mean_token_accuracy": 0.22189746648073197,
      "num_tokens": 76098244.0,
      "step": 34095
    },
    {
      "entropy": 5.297691345214844,
      "epoch": 3.8297860391980683,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003604523763653706,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22503688633441926,
      "num_tokens": 76108798.0,
      "step": 34100
    },
    {
      "entropy": 5.372882270812989,
      "epoch": 3.8303476161060255,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036041523449266573,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22199310213327408,
      "num_tokens": 76120776.0,
      "step": 34105
    },
    {
      "entropy": 5.355411338806152,
      "epoch": 3.830909193013983,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036037808990030667,
      "loss": 5.0469,
      "mean_token_accuracy": 0.22123882323503494,
      "num_tokens": 76133562.0,
      "step": 34110
    },
    {
      "entropy": 5.263192129135132,
      "epoch": 3.831470769921941,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003603409425894761,
      "loss": 4.9903,
      "mean_token_accuracy": 0.2226415604352951,
      "num_tokens": 76146071.0,
      "step": 34115
    },
    {
      "entropy": 5.353691291809082,
      "epoch": 3.8320323468298985,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003603037925613568,
      "loss": 4.9239,
      "mean_token_accuracy": 0.22817468494176865,
      "num_tokens": 76157568.0,
      "step": 34120
    },
    {
      "entropy": 5.33496732711792,
      "epoch": 3.832593923737856,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003602666398171317,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22977200001478196,
      "num_tokens": 76168853.0,
      "step": 34125
    },
    {
      "entropy": 5.211858892440796,
      "epoch": 3.8331555006458133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036022948435798373,
      "loss": 4.8569,
      "mean_token_accuracy": 0.22130682468414306,
      "num_tokens": 76178480.0,
      "step": 34130
    },
    {
      "entropy": 5.244050645828247,
      "epoch": 3.833717077553771,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000360192326185096,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22588102966547013,
      "num_tokens": 76189031.0,
      "step": 34135
    },
    {
      "entropy": 5.20930061340332,
      "epoch": 3.8342786544617287,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003601551652996514,
      "loss": 4.8527,
      "mean_token_accuracy": 0.2241537094116211,
      "num_tokens": 76199345.0,
      "step": 34140
    },
    {
      "entropy": 5.3511070728302,
      "epoch": 3.834840231369686,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003601180017028334,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21012949347496032,
      "num_tokens": 76211023.0,
      "step": 34145
    },
    {
      "entropy": 5.43844165802002,
      "epoch": 3.8354018082776435,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003600808353958252,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21390115469694138,
      "num_tokens": 76221562.0,
      "step": 34150
    },
    {
      "entropy": 5.377560472488403,
      "epoch": 3.835963385185601,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003600436663798102,
      "loss": 5.0947,
      "mean_token_accuracy": 0.2088586449623108,
      "num_tokens": 76235113.0,
      "step": 34155
    },
    {
      "entropy": 5.373347043991089,
      "epoch": 3.836524962093559,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036000649465597185,
      "loss": 4.9767,
      "mean_token_accuracy": 0.22686143666505815,
      "num_tokens": 76245804.0,
      "step": 34160
    },
    {
      "entropy": 5.340077018737793,
      "epoch": 3.8370865390015165,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035996932022549377,
      "loss": 4.9573,
      "mean_token_accuracy": 0.226337006688118,
      "num_tokens": 76257275.0,
      "step": 34165
    },
    {
      "entropy": 5.308880233764649,
      "epoch": 3.8376481159094737,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00035993214308955955,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21919851154088973,
      "num_tokens": 76268401.0,
      "step": 34170
    },
    {
      "entropy": 5.2743120193481445,
      "epoch": 3.8382096928174314,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000359894963249353,
      "loss": 4.9009,
      "mean_token_accuracy": 0.21955974996089936,
      "num_tokens": 76278943.0,
      "step": 34175
    },
    {
      "entropy": 5.332439708709717,
      "epoch": 3.838771269725389,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035985778070605785,
      "loss": 4.9804,
      "mean_token_accuracy": 0.22518874257802962,
      "num_tokens": 76289899.0,
      "step": 34180
    },
    {
      "entropy": 5.397191286087036,
      "epoch": 3.8393328466333463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003598205954608581,
      "loss": 5.0668,
      "mean_token_accuracy": 0.21639354079961776,
      "num_tokens": 76300574.0,
      "step": 34185
    },
    {
      "entropy": 5.307239723205567,
      "epoch": 3.839894423541304,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00035978340751493767,
      "loss": 5.025,
      "mean_token_accuracy": 0.21680848002433778,
      "num_tokens": 76312789.0,
      "step": 34190
    },
    {
      "entropy": 5.282459115982055,
      "epoch": 3.8404560004492616,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035974621686948063,
      "loss": 4.8967,
      "mean_token_accuracy": 0.22004208713769913,
      "num_tokens": 76322911.0,
      "step": 34195
    },
    {
      "entropy": 5.338252353668213,
      "epoch": 3.841017577357219,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003597090235256712,
      "loss": 5.0265,
      "mean_token_accuracy": 0.21764661371707916,
      "num_tokens": 76333699.0,
      "step": 34200
    },
    {
      "entropy": 5.21779236793518,
      "epoch": 3.8415791542651765,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003596718274846936,
      "loss": 4.902,
      "mean_token_accuracy": 0.2212449923157692,
      "num_tokens": 76344405.0,
      "step": 34205
    },
    {
      "entropy": 5.2961455345153805,
      "epoch": 3.842140731173134,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003596346287477322,
      "loss": 4.9983,
      "mean_token_accuracy": 0.21882383972406388,
      "num_tokens": 76356403.0,
      "step": 34210
    },
    {
      "entropy": 5.347577857971191,
      "epoch": 3.842702308081092,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003595974273159714,
      "loss": 4.9391,
      "mean_token_accuracy": 0.2180602341890335,
      "num_tokens": 76368216.0,
      "step": 34215
    },
    {
      "entropy": 5.4728484630584715,
      "epoch": 3.8432638849890495,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003595602231905957,
      "loss": 5.1732,
      "mean_token_accuracy": 0.21075690686702728,
      "num_tokens": 76379996.0,
      "step": 34220
    },
    {
      "entropy": 5.2909965991973875,
      "epoch": 3.8438254618970067,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035952301637278973,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22745050936937333,
      "num_tokens": 76391409.0,
      "step": 34225
    },
    {
      "entropy": 5.3159387588500975,
      "epoch": 3.8443870388049644,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035948580686373803,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21929509192705154,
      "num_tokens": 76402111.0,
      "step": 34230
    },
    {
      "entropy": 5.29626579284668,
      "epoch": 3.844948615712922,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035944859466462554,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22586579173803328,
      "num_tokens": 76413225.0,
      "step": 34235
    },
    {
      "entropy": 5.34407868385315,
      "epoch": 3.8455101926208792,
      "grad_norm": 0.984375,
      "learning_rate": 0.000359411379776637,
      "loss": 4.9884,
      "mean_token_accuracy": 0.218968565762043,
      "num_tokens": 76423236.0,
      "step": 34240
    },
    {
      "entropy": 5.3959883689880375,
      "epoch": 3.846071769528837,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035937416220095745,
      "loss": 5.0245,
      "mean_token_accuracy": 0.22425228655338286,
      "num_tokens": 76435599.0,
      "step": 34245
    },
    {
      "entropy": 5.2845948219299315,
      "epoch": 3.8466333464367946,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003593369419387718,
      "loss": 4.9794,
      "mean_token_accuracy": 0.2224591627717018,
      "num_tokens": 76446182.0,
      "step": 34250
    },
    {
      "entropy": 5.297406482696533,
      "epoch": 3.847194923344752,
      "grad_norm": 0.84375,
      "learning_rate": 0.0003592997189912652,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22103478163480758,
      "num_tokens": 76459151.0,
      "step": 34255
    },
    {
      "entropy": 5.385849475860596,
      "epoch": 3.84775650025271,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035926249335962296,
      "loss": 5.0557,
      "mean_token_accuracy": 0.2154635891318321,
      "num_tokens": 76471235.0,
      "step": 34260
    },
    {
      "entropy": 5.290727853775024,
      "epoch": 3.848318077160667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035922526504503016,
      "loss": 4.9387,
      "mean_token_accuracy": 0.2178479790687561,
      "num_tokens": 76481699.0,
      "step": 34265
    },
    {
      "entropy": 5.366047716140747,
      "epoch": 3.8488796540686248,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003591880340486723,
      "loss": 5.0577,
      "mean_token_accuracy": 0.21498210579156876,
      "num_tokens": 76493038.0,
      "step": 34270
    },
    {
      "entropy": 5.294825172424316,
      "epoch": 3.8494412309765824,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003591508003717347,
      "loss": 4.9409,
      "mean_token_accuracy": 0.223392952978611,
      "num_tokens": 76503601.0,
      "step": 34275
    },
    {
      "entropy": 5.318000268936157,
      "epoch": 3.8500028078845396,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035911356401540306,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22591498345136643,
      "num_tokens": 76513914.0,
      "step": 34280
    },
    {
      "entropy": 5.318132781982422,
      "epoch": 3.8505643847924973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003590763249808629,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2282802939414978,
      "num_tokens": 76524924.0,
      "step": 34285
    },
    {
      "entropy": 5.252163743972778,
      "epoch": 3.851125961700455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035903908326929995,
      "loss": 4.9068,
      "mean_token_accuracy": 0.2254076898097992,
      "num_tokens": 76536468.0,
      "step": 34290
    },
    {
      "entropy": 5.374836683273315,
      "epoch": 3.851687538608412,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035900183888190005,
      "loss": 5.0282,
      "mean_token_accuracy": 0.22161675989627838,
      "num_tokens": 76548320.0,
      "step": 34295
    },
    {
      "entropy": 5.336897230148315,
      "epoch": 3.85224911551637,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003589645918198489,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22136852741241456,
      "num_tokens": 76560482.0,
      "step": 34300
    },
    {
      "entropy": 5.351146554946899,
      "epoch": 3.8528106924243275,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035892734208433264,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21467004865407943,
      "num_tokens": 76571743.0,
      "step": 34305
    },
    {
      "entropy": 5.276084518432617,
      "epoch": 3.853372269332285,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035889008967653725,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2131062477827072,
      "num_tokens": 76582545.0,
      "step": 34310
    },
    {
      "entropy": 5.3727256774902346,
      "epoch": 3.853933846240243,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003588528345976489,
      "loss": 5.0526,
      "mean_token_accuracy": 0.21507383584976197,
      "num_tokens": 76594104.0,
      "step": 34315
    },
    {
      "entropy": 5.372976875305175,
      "epoch": 3.8544954231482,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003588155768488537,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21563293635845185,
      "num_tokens": 76603414.0,
      "step": 34320
    },
    {
      "entropy": 5.374476051330566,
      "epoch": 3.8550570000561577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035877831643133803,
      "loss": 5.0229,
      "mean_token_accuracy": 0.21743345856666565,
      "num_tokens": 76614098.0,
      "step": 34325
    },
    {
      "entropy": 5.466065311431885,
      "epoch": 3.8556185769641154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003587410533462883,
      "loss": 5.1374,
      "mean_token_accuracy": 0.2166756808757782,
      "num_tokens": 76625179.0,
      "step": 34330
    },
    {
      "entropy": 5.291461515426636,
      "epoch": 3.8561801538720726,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035870378759489104,
      "loss": 4.8606,
      "mean_token_accuracy": 0.2264542043209076,
      "num_tokens": 76636710.0,
      "step": 34335
    },
    {
      "entropy": 5.290535306930542,
      "epoch": 3.8567417307800302,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003586665191783326,
      "loss": 4.9998,
      "mean_token_accuracy": 0.21129338592290878,
      "num_tokens": 76647585.0,
      "step": 34340
    },
    {
      "entropy": 5.347399425506592,
      "epoch": 3.857303307687988,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003586292480977998,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2124169498682022,
      "num_tokens": 76658376.0,
      "step": 34345
    },
    {
      "entropy": 5.384722423553467,
      "epoch": 3.8578648845959456,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035859197435447927,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22697136998176576,
      "num_tokens": 76669237.0,
      "step": 34350
    },
    {
      "entropy": 5.367992210388183,
      "epoch": 3.8584264615039032,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035855469794955787,
      "loss": 5.0957,
      "mean_token_accuracy": 0.20734002590179443,
      "num_tokens": 76681067.0,
      "step": 34355
    },
    {
      "entropy": 5.408801794052124,
      "epoch": 3.8589880384118604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035851741888422254,
      "loss": 4.9837,
      "mean_token_accuracy": 0.2224266082048416,
      "num_tokens": 76690858.0,
      "step": 34360
    },
    {
      "entropy": 5.387903690338135,
      "epoch": 3.859549615319818,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003584801371596601,
      "loss": 5.0553,
      "mean_token_accuracy": 0.2148429811000824,
      "num_tokens": 76701571.0,
      "step": 34365
    },
    {
      "entropy": 5.3273707866668705,
      "epoch": 3.8601111922277758,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003584428527770578,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21457699835300445,
      "num_tokens": 76712565.0,
      "step": 34370
    },
    {
      "entropy": 5.359912014007568,
      "epoch": 3.860672769135733,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035840556573760266,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21210582703351974,
      "num_tokens": 76724484.0,
      "step": 34375
    },
    {
      "entropy": 5.312947797775268,
      "epoch": 3.8612343460436906,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003583682760424819,
      "loss": 4.8897,
      "mean_token_accuracy": 0.22590861916542054,
      "num_tokens": 76735596.0,
      "step": 34380
    },
    {
      "entropy": 5.413711071014404,
      "epoch": 3.8617959229516483,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035833098369288303,
      "loss": 5.0667,
      "mean_token_accuracy": 0.21579882502555847,
      "num_tokens": 76747160.0,
      "step": 34385
    },
    {
      "entropy": 5.327003860473633,
      "epoch": 3.8623574998596055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035829368868999325,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21669704467058182,
      "num_tokens": 76758136.0,
      "step": 34390
    },
    {
      "entropy": 5.296048927307129,
      "epoch": 3.862919076767563,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035825639103500005,
      "loss": 4.9796,
      "mean_token_accuracy": 0.21815905421972276,
      "num_tokens": 76768167.0,
      "step": 34395
    },
    {
      "entropy": 5.438165712356567,
      "epoch": 3.863480653675521,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035821909072909113,
      "loss": 5.1214,
      "mean_token_accuracy": 0.21192978620529174,
      "num_tokens": 76779717.0,
      "step": 34400
    },
    {
      "entropy": 5.243574905395508,
      "epoch": 3.8640422305834785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035818178777345417,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21840597838163375,
      "num_tokens": 76790837.0,
      "step": 34405
    },
    {
      "entropy": 5.356860113143921,
      "epoch": 3.864603807491436,
      "grad_norm": 1.0,
      "learning_rate": 0.0003581444821692767,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22023036032915116,
      "num_tokens": 76801159.0,
      "step": 34410
    },
    {
      "entropy": 5.366121625900268,
      "epoch": 3.8651653843993934,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003581071739177467,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21985567659139632,
      "num_tokens": 76812409.0,
      "step": 34415
    },
    {
      "entropy": 5.4340386390686035,
      "epoch": 3.865726961307351,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035806986302005205,
      "loss": 5.0775,
      "mean_token_accuracy": 0.2129431650042534,
      "num_tokens": 76824360.0,
      "step": 34420
    },
    {
      "entropy": 5.394805574417115,
      "epoch": 3.8662885382153087,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003580325494773808,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22332290410995484,
      "num_tokens": 76834725.0,
      "step": 34425
    },
    {
      "entropy": 5.275696659088135,
      "epoch": 3.866850115123266,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035799523329092093,
      "loss": 4.9783,
      "mean_token_accuracy": 0.22392223477363588,
      "num_tokens": 76846099.0,
      "step": 34430
    },
    {
      "entropy": 5.3988707065582275,
      "epoch": 3.8674116920312236,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035795791446186073,
      "loss": 5.0347,
      "mean_token_accuracy": 0.21507358253002168,
      "num_tokens": 76857229.0,
      "step": 34435
    },
    {
      "entropy": 5.438662815093994,
      "epoch": 3.8679732689391813,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003579205929913882,
      "loss": 4.9931,
      "mean_token_accuracy": 0.22638604044914246,
      "num_tokens": 76868609.0,
      "step": 34440
    },
    {
      "entropy": 5.401218700408935,
      "epoch": 3.868534845847139,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003578832688806919,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22060953676700593,
      "num_tokens": 76879539.0,
      "step": 34445
    },
    {
      "entropy": 5.365166854858399,
      "epoch": 3.8690964227550966,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003578459421309602,
      "loss": 5.0251,
      "mean_token_accuracy": 0.21152975410223007,
      "num_tokens": 76889333.0,
      "step": 34450
    },
    {
      "entropy": 5.33002815246582,
      "epoch": 3.869657999663054,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003578086127433816,
      "loss": 5.0751,
      "mean_token_accuracy": 0.2153173044323921,
      "num_tokens": 76901334.0,
      "step": 34455
    },
    {
      "entropy": 5.425039768218994,
      "epoch": 3.8702195765710115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035777128071914454,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2244469329714775,
      "num_tokens": 76913758.0,
      "step": 34460
    },
    {
      "entropy": 5.398364734649658,
      "epoch": 3.870781153478969,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003577339460594378,
      "loss": 5.0256,
      "mean_token_accuracy": 0.21481548100709916,
      "num_tokens": 76925317.0,
      "step": 34465
    },
    {
      "entropy": 5.365930414199829,
      "epoch": 3.8713427303869263,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003576966087654502,
      "loss": 5.0827,
      "mean_token_accuracy": 0.21622089594602584,
      "num_tokens": 76937169.0,
      "step": 34470
    },
    {
      "entropy": 5.322516298294067,
      "epoch": 3.871904307294884,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035765926883837054,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21853512674570083,
      "num_tokens": 76948958.0,
      "step": 34475
    },
    {
      "entropy": 5.405957984924316,
      "epoch": 3.8724658842028417,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003576219262793877,
      "loss": 5.0723,
      "mean_token_accuracy": 0.2142027050256729,
      "num_tokens": 76960686.0,
      "step": 34480
    },
    {
      "entropy": 5.340802955627441,
      "epoch": 3.873027461110799,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035758458108969056,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22139831334352494,
      "num_tokens": 76971127.0,
      "step": 34485
    },
    {
      "entropy": 5.293386363983155,
      "epoch": 3.8735890380187565,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003575472332704684,
      "loss": 4.8313,
      "mean_token_accuracy": 0.2363441213965416,
      "num_tokens": 76982647.0,
      "step": 34490
    },
    {
      "entropy": 5.372289705276489,
      "epoch": 3.874150614926714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035750988282291037,
      "loss": 5.0559,
      "mean_token_accuracy": 0.22276147603988647,
      "num_tokens": 76994274.0,
      "step": 34495
    },
    {
      "entropy": 5.304782962799072,
      "epoch": 3.874712191834672,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035747252974820565,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22307076156139374,
      "num_tokens": 77005435.0,
      "step": 34500
    },
    {
      "entropy": 5.319442844390869,
      "epoch": 3.8752737687426295,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035743517404754354,
      "loss": 4.9397,
      "mean_token_accuracy": 0.21738810688257218,
      "num_tokens": 77016913.0,
      "step": 34505
    },
    {
      "entropy": 5.327348089218139,
      "epoch": 3.8758353456505867,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003573978157221136,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21784308850765227,
      "num_tokens": 77027340.0,
      "step": 34510
    },
    {
      "entropy": 5.325407791137695,
      "epoch": 3.8763969225585444,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035736045477310517,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21725043505430222,
      "num_tokens": 77039465.0,
      "step": 34515
    },
    {
      "entropy": 5.371528577804566,
      "epoch": 3.876958499466502,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035732309120170786,
      "loss": 5.0376,
      "mean_token_accuracy": 0.21463114470243455,
      "num_tokens": 77050503.0,
      "step": 34520
    },
    {
      "entropy": 5.473624658584595,
      "epoch": 3.8775200763744593,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035728572500911155,
      "loss": 4.9595,
      "mean_token_accuracy": 0.22070858925580977,
      "num_tokens": 77060212.0,
      "step": 34525
    },
    {
      "entropy": 5.411468935012818,
      "epoch": 3.878081653282417,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035724835619650573,
      "loss": 5.0723,
      "mean_token_accuracy": 0.21515158712863922,
      "num_tokens": 77071134.0,
      "step": 34530
    },
    {
      "entropy": 5.392029094696045,
      "epoch": 3.8786432301903746,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003572109847650804,
      "loss": 5.0266,
      "mean_token_accuracy": 0.22550059705972672,
      "num_tokens": 77081844.0,
      "step": 34535
    },
    {
      "entropy": 5.363567638397217,
      "epoch": 3.8792048070983323,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035717361071602546,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22274601757526397,
      "num_tokens": 77092891.0,
      "step": 34540
    },
    {
      "entropy": 5.294776678085327,
      "epoch": 3.87976638400629,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035713623405053083,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22207600772380828,
      "num_tokens": 77104620.0,
      "step": 34545
    },
    {
      "entropy": 5.327441596984864,
      "epoch": 3.880327960914247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035709885476978666,
      "loss": 5.0206,
      "mean_token_accuracy": 0.21455124020576477,
      "num_tokens": 77115249.0,
      "step": 34550
    },
    {
      "entropy": 5.365397214889526,
      "epoch": 3.880889537822205,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035706147287498303,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21623631566762924,
      "num_tokens": 77127287.0,
      "step": 34555
    },
    {
      "entropy": 5.460450983047485,
      "epoch": 3.8814511147301625,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035702408836731043,
      "loss": 4.9835,
      "mean_token_accuracy": 0.22453559786081315,
      "num_tokens": 77139022.0,
      "step": 34560
    },
    {
      "entropy": 5.3107527732849125,
      "epoch": 3.8820126916381197,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000356986701247959,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21536468267440795,
      "num_tokens": 77148594.0,
      "step": 34565
    },
    {
      "entropy": 5.293627166748047,
      "epoch": 3.8825742685460773,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003569493115181192,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2237549230456352,
      "num_tokens": 77158882.0,
      "step": 34570
    },
    {
      "entropy": 5.332336807250977,
      "epoch": 3.883135845454035,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003569119191789815,
      "loss": 4.999,
      "mean_token_accuracy": 0.21486303061246873,
      "num_tokens": 77169721.0,
      "step": 34575
    },
    {
      "entropy": 5.369428110122681,
      "epoch": 3.8836974223619922,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035687452423173655,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21325239092111586,
      "num_tokens": 77181223.0,
      "step": 34580
    },
    {
      "entropy": 5.327916860580444,
      "epoch": 3.88425899926995,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000356837126677575,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21363091617822647,
      "num_tokens": 77193000.0,
      "step": 34585
    },
    {
      "entropy": 5.339040756225586,
      "epoch": 3.8848205761779075,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003567997265176876,
      "loss": 4.949,
      "mean_token_accuracy": 0.21859423369169234,
      "num_tokens": 77203721.0,
      "step": 34590
    },
    {
      "entropy": 5.289082479476929,
      "epoch": 3.885382153085865,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035676232375326513,
      "loss": 4.819,
      "mean_token_accuracy": 0.2308942675590515,
      "num_tokens": 77214601.0,
      "step": 34595
    },
    {
      "entropy": 5.381571483612061,
      "epoch": 3.885943729993823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035672491838549867,
      "loss": 4.9867,
      "mean_token_accuracy": 0.21910925805568696,
      "num_tokens": 77225478.0,
      "step": 34600
    },
    {
      "entropy": 5.385287427902222,
      "epoch": 3.88650530690178,
      "grad_norm": 1.015625,
      "learning_rate": 0.000356687510415579,
      "loss": 5.063,
      "mean_token_accuracy": 0.21933630257844924,
      "num_tokens": 77237601.0,
      "step": 34605
    },
    {
      "entropy": 5.378786420822143,
      "epoch": 3.8870668838097377,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003566500998446973,
      "loss": 5.0697,
      "mean_token_accuracy": 0.21607993096113204,
      "num_tokens": 77248137.0,
      "step": 34610
    },
    {
      "entropy": 5.425086498260498,
      "epoch": 3.8876284607176954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035661268667404485,
      "loss": 5.0687,
      "mean_token_accuracy": 0.21746052652597428,
      "num_tokens": 77260366.0,
      "step": 34615
    },
    {
      "entropy": 5.436946296691895,
      "epoch": 3.8881900376256526,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003565752709048127,
      "loss": 5.0287,
      "mean_token_accuracy": 0.21868159174919127,
      "num_tokens": 77270301.0,
      "step": 34620
    },
    {
      "entropy": 5.378354883193969,
      "epoch": 3.8887516145336103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003565378525381924,
      "loss": 5.0531,
      "mean_token_accuracy": 0.21499278247356415,
      "num_tokens": 77282824.0,
      "step": 34625
    },
    {
      "entropy": 5.3479057312011715,
      "epoch": 3.889313191441568,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003565004315753752,
      "loss": 4.99,
      "mean_token_accuracy": 0.22019737511873244,
      "num_tokens": 77293532.0,
      "step": 34630
    },
    {
      "entropy": 5.403587341308594,
      "epoch": 3.8898747683495256,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035646300801755266,
      "loss": 4.9817,
      "mean_token_accuracy": 0.2219406023621559,
      "num_tokens": 77304713.0,
      "step": 34635
    },
    {
      "entropy": 5.3187775135040285,
      "epoch": 3.8904363452574833,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035642558186591635,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21719868779182433,
      "num_tokens": 77317666.0,
      "step": 34640
    },
    {
      "entropy": 5.280806398391723,
      "epoch": 3.8909979221654405,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003563881531216579,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22085976153612136,
      "num_tokens": 77328217.0,
      "step": 34645
    },
    {
      "entropy": 5.385529947280884,
      "epoch": 3.891559499073398,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003563507217859691,
      "loss": 4.9873,
      "mean_token_accuracy": 0.22658690214157104,
      "num_tokens": 77339607.0,
      "step": 34650
    },
    {
      "entropy": 5.440567922592163,
      "epoch": 3.892121075981356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003563132878600419,
      "loss": 5.0705,
      "mean_token_accuracy": 0.21355245411396026,
      "num_tokens": 77351651.0,
      "step": 34655
    },
    {
      "entropy": 5.478644800186157,
      "epoch": 3.892682652889313,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003562758513450678,
      "loss": 5.111,
      "mean_token_accuracy": 0.21229639947414397,
      "num_tokens": 77364431.0,
      "step": 34660
    },
    {
      "entropy": 5.309645080566407,
      "epoch": 3.8932442297972707,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003562384122422393,
      "loss": 4.9945,
      "mean_token_accuracy": 0.22049230039119722,
      "num_tokens": 77375426.0,
      "step": 34665
    },
    {
      "entropy": 5.310685491561889,
      "epoch": 3.8938058067052284,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003562009705527481,
      "loss": 4.984,
      "mean_token_accuracy": 0.222299924492836,
      "num_tokens": 77386794.0,
      "step": 34670
    },
    {
      "entropy": 5.382521152496338,
      "epoch": 3.8943673836131856,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035616352627778665,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2231205940246582,
      "num_tokens": 77398155.0,
      "step": 34675
    },
    {
      "entropy": 5.438284254074096,
      "epoch": 3.8949289605211432,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035612607941854695,
      "loss": 5.1391,
      "mean_token_accuracy": 0.20671844184398652,
      "num_tokens": 77408922.0,
      "step": 34680
    },
    {
      "entropy": 5.312597370147705,
      "epoch": 3.895490537429101,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003560886299762214,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22600263655185698,
      "num_tokens": 77419084.0,
      "step": 34685
    },
    {
      "entropy": 5.2731733322143555,
      "epoch": 3.8960521143370586,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003560511779520025,
      "loss": 4.8938,
      "mean_token_accuracy": 0.2297250047326088,
      "num_tokens": 77430141.0,
      "step": 34690
    },
    {
      "entropy": 5.372419691085815,
      "epoch": 3.896613691245016,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003560137233470826,
      "loss": 5.0977,
      "mean_token_accuracy": 0.22026560008525847,
      "num_tokens": 77440754.0,
      "step": 34695
    },
    {
      "entropy": 5.326091480255127,
      "epoch": 3.8971752681529734,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035597626616265434,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21802209615707396,
      "num_tokens": 77453532.0,
      "step": 34700
    },
    {
      "entropy": 5.354774045944214,
      "epoch": 3.897736845060931,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035593880639991037,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22288752496242523,
      "num_tokens": 77464987.0,
      "step": 34705
    },
    {
      "entropy": 5.355606126785278,
      "epoch": 3.8982984219688888,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035590134406004336,
      "loss": 4.9931,
      "mean_token_accuracy": 0.21443939357995986,
      "num_tokens": 77477559.0,
      "step": 34710
    },
    {
      "entropy": 5.31189227104187,
      "epoch": 3.898859998876846,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003558638791442463,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2153637230396271,
      "num_tokens": 77489863.0,
      "step": 34715
    },
    {
      "entropy": 5.282545948028565,
      "epoch": 3.8994215757848036,
      "grad_norm": 1.0,
      "learning_rate": 0.0003558264116537119,
      "loss": 4.9221,
      "mean_token_accuracy": 0.2196336567401886,
      "num_tokens": 77501119.0,
      "step": 34720
    },
    {
      "entropy": 5.3356256008148195,
      "epoch": 3.8999831526927613,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003557889415896332,
      "loss": 4.9997,
      "mean_token_accuracy": 0.22867706865072251,
      "num_tokens": 77512530.0,
      "step": 34725
    },
    {
      "entropy": 5.370674514770508,
      "epoch": 3.900544729600719,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035575146895320327,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21908026933670044,
      "num_tokens": 77524624.0,
      "step": 34730
    },
    {
      "entropy": 5.386878776550293,
      "epoch": 3.9011063065086766,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003557139937456152,
      "loss": 5.006,
      "mean_token_accuracy": 0.21558476239442825,
      "num_tokens": 77536567.0,
      "step": 34735
    },
    {
      "entropy": 5.322687816619873,
      "epoch": 3.901667883416634,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035567651596806236,
      "loss": 4.9601,
      "mean_token_accuracy": 0.22049544602632523,
      "num_tokens": 77547061.0,
      "step": 34740
    },
    {
      "entropy": 5.345395183563232,
      "epoch": 3.9022294603245915,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00035563903562173786,
      "loss": 5.0078,
      "mean_token_accuracy": 0.22194723635911942,
      "num_tokens": 77558783.0,
      "step": 34745
    },
    {
      "entropy": 5.288995456695557,
      "epoch": 3.902791037232549,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00035560155270783527,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22381065040826797,
      "num_tokens": 77570382.0,
      "step": 34750
    },
    {
      "entropy": 5.355011415481568,
      "epoch": 3.9033526141405064,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003555640672275479,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21910130530595778,
      "num_tokens": 77581750.0,
      "step": 34755
    },
    {
      "entropy": 5.352764129638672,
      "epoch": 3.903914191048464,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003555265791820695,
      "loss": 4.9955,
      "mean_token_accuracy": 0.2259882226586342,
      "num_tokens": 77593784.0,
      "step": 34760
    },
    {
      "entropy": 5.237539052963257,
      "epoch": 3.9044757679564217,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003554890885725935,
      "loss": 4.8812,
      "mean_token_accuracy": 0.22495695650577546,
      "num_tokens": 77604993.0,
      "step": 34765
    },
    {
      "entropy": 5.30055799484253,
      "epoch": 3.905037344864379,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003554515954003137,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22281993329524993,
      "num_tokens": 77615977.0,
      "step": 34770
    },
    {
      "entropy": 5.300332641601562,
      "epoch": 3.9055989217723366,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003554140996664239,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22245702147483826,
      "num_tokens": 77625724.0,
      "step": 34775
    },
    {
      "entropy": 5.29795470237732,
      "epoch": 3.9061604986802942,
      "grad_norm": 1.046875,
      "learning_rate": 0.000355376601372118,
      "loss": 4.9665,
      "mean_token_accuracy": 0.22179109156131743,
      "num_tokens": 77635924.0,
      "step": 34780
    },
    {
      "entropy": 5.369157838821411,
      "epoch": 3.906722075588252,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003553391005185899,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21407967656850815,
      "num_tokens": 77646142.0,
      "step": 34785
    },
    {
      "entropy": 5.343344354629517,
      "epoch": 3.9072836524962096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003553015971070338,
      "loss": 4.964,
      "mean_token_accuracy": 0.22329080998897552,
      "num_tokens": 77657500.0,
      "step": 34790
    },
    {
      "entropy": 5.416098308563233,
      "epoch": 3.907845229404167,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003552640911386436,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21359740644693376,
      "num_tokens": 77668916.0,
      "step": 34795
    },
    {
      "entropy": 5.3855578899383545,
      "epoch": 3.9084068063121244,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035522658261461366,
      "loss": 5.0269,
      "mean_token_accuracy": 0.20899223834276198,
      "num_tokens": 77680395.0,
      "step": 34800
    },
    {
      "entropy": 5.295803594589233,
      "epoch": 3.908968383220082,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003551890715361382,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22267122715711593,
      "num_tokens": 77691309.0,
      "step": 34805
    },
    {
      "entropy": 5.353998613357544,
      "epoch": 3.9095299601280393,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003551515579044116,
      "loss": 4.948,
      "mean_token_accuracy": 0.22546534538269042,
      "num_tokens": 77702922.0,
      "step": 34810
    },
    {
      "entropy": 5.372947931289673,
      "epoch": 3.910091537035997,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00035511404172062827,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2166505813598633,
      "num_tokens": 77714252.0,
      "step": 34815
    },
    {
      "entropy": 5.285768699645996,
      "epoch": 3.9106531139439547,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003550765229859828,
      "loss": 4.882,
      "mean_token_accuracy": 0.22304289042949677,
      "num_tokens": 77724473.0,
      "step": 34820
    },
    {
      "entropy": 5.2749042987823485,
      "epoch": 3.9112146908519123,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003550390017016698,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21660188287496568,
      "num_tokens": 77735103.0,
      "step": 34825
    },
    {
      "entropy": 5.4561093807220455,
      "epoch": 3.91177626775987,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035500147786888387,
      "loss": 5.0966,
      "mean_token_accuracy": 0.21477610170841216,
      "num_tokens": 77747370.0,
      "step": 34830
    },
    {
      "entropy": 5.337905979156494,
      "epoch": 3.912337844667827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035496395148881994,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21486195921897888,
      "num_tokens": 77759183.0,
      "step": 34835
    },
    {
      "entropy": 5.381987571716309,
      "epoch": 3.912899421575785,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035492642256267277,
      "loss": 4.966,
      "mean_token_accuracy": 0.22381928265094758,
      "num_tokens": 77770130.0,
      "step": 34840
    },
    {
      "entropy": 5.27836332321167,
      "epoch": 3.9134609984837425,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035488889109163725,
      "loss": 4.8474,
      "mean_token_accuracy": 0.23484774827957153,
      "num_tokens": 77780830.0,
      "step": 34845
    },
    {
      "entropy": 5.31723051071167,
      "epoch": 3.9140225753916997,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003548513570769085,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21503322422504426,
      "num_tokens": 77791265.0,
      "step": 34850
    },
    {
      "entropy": 5.22876763343811,
      "epoch": 3.9145841522996574,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003548138205196816,
      "loss": 4.9042,
      "mean_token_accuracy": 0.2245846137404442,
      "num_tokens": 77802899.0,
      "step": 34855
    },
    {
      "entropy": 5.232019329071045,
      "epoch": 3.915145729207615,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003547762814211517,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22694508880376815,
      "num_tokens": 77815034.0,
      "step": 34860
    },
    {
      "entropy": 5.371186542510986,
      "epoch": 3.9157073061155723,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000354738739782514,
      "loss": 5.0139,
      "mean_token_accuracy": 0.2130333214998245,
      "num_tokens": 77826305.0,
      "step": 34865
    },
    {
      "entropy": 5.372768449783325,
      "epoch": 3.91626888302353,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000354701195604964,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2186853975057602,
      "num_tokens": 77837698.0,
      "step": 34870
    },
    {
      "entropy": 5.332578039169311,
      "epoch": 3.9168304599314876,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003546636488896969,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22395929247140883,
      "num_tokens": 77848630.0,
      "step": 34875
    },
    {
      "entropy": 5.367195653915405,
      "epoch": 3.9173920368394453,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003546260996379083,
      "loss": 5.0564,
      "mean_token_accuracy": 0.21050926446914672,
      "num_tokens": 77859832.0,
      "step": 34880
    },
    {
      "entropy": 5.39934287071228,
      "epoch": 3.917953613747403,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000354588547850794,
      "loss": 5.0944,
      "mean_token_accuracy": 0.21282154321670532,
      "num_tokens": 77871071.0,
      "step": 34885
    },
    {
      "entropy": 5.3235045909881595,
      "epoch": 3.91851519065536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003545509935295493,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21860711127519608,
      "num_tokens": 77881323.0,
      "step": 34890
    },
    {
      "entropy": 5.374647760391236,
      "epoch": 3.919076767563318,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003545134366753701,
      "loss": 5.0242,
      "mean_token_accuracy": 0.21758922934532166,
      "num_tokens": 77893754.0,
      "step": 34895
    },
    {
      "entropy": 5.351257276535034,
      "epoch": 3.9196383444712755,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003544758772894523,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2209628328680992,
      "num_tokens": 77905408.0,
      "step": 34900
    },
    {
      "entropy": 5.241050291061401,
      "epoch": 3.9201999213792327,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035443831537299177,
      "loss": 4.8341,
      "mean_token_accuracy": 0.23224117755889892,
      "num_tokens": 77915839.0,
      "step": 34905
    },
    {
      "entropy": 5.315095138549805,
      "epoch": 3.9207614982871903,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003544007509271844,
      "loss": 4.931,
      "mean_token_accuracy": 0.22335672825574876,
      "num_tokens": 77926626.0,
      "step": 34910
    },
    {
      "entropy": 5.381797409057617,
      "epoch": 3.921323075195148,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035436318395322636,
      "loss": 5.1326,
      "mean_token_accuracy": 0.2024238958954811,
      "num_tokens": 77938781.0,
      "step": 34915
    },
    {
      "entropy": 5.418303394317627,
      "epoch": 3.9218846521031057,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003543256144523137,
      "loss": 5.0186,
      "mean_token_accuracy": 0.2163691148161888,
      "num_tokens": 77949370.0,
      "step": 34920
    },
    {
      "entropy": 5.329764461517334,
      "epoch": 3.9224462290110633,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003542880424256428,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21960351318120958,
      "num_tokens": 77961396.0,
      "step": 34925
    },
    {
      "entropy": 5.405273675918579,
      "epoch": 3.9230078059190205,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003542504678744099,
      "loss": 5.0571,
      "mean_token_accuracy": 0.21164146363735198,
      "num_tokens": 77972811.0,
      "step": 34930
    },
    {
      "entropy": 5.314858770370483,
      "epoch": 3.923569382826978,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003542128907998113,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22843219637870787,
      "num_tokens": 77984452.0,
      "step": 34935
    },
    {
      "entropy": 5.291006326675415,
      "epoch": 3.924130959734936,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003541753112030435,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22351561933755876,
      "num_tokens": 77995963.0,
      "step": 34940
    },
    {
      "entropy": 5.273133039474487,
      "epoch": 3.924692536642893,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035413772908530317,
      "loss": 4.861,
      "mean_token_accuracy": 0.2281592458486557,
      "num_tokens": 78006638.0,
      "step": 34945
    },
    {
      "entropy": 5.22492151260376,
      "epoch": 3.9252541135508507,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003541001444477869,
      "loss": 4.8735,
      "mean_token_accuracy": 0.2345397874712944,
      "num_tokens": 78017982.0,
      "step": 34950
    },
    {
      "entropy": 5.263615846633911,
      "epoch": 3.9258156904588084,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035406255729169124,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22397965639829637,
      "num_tokens": 78028827.0,
      "step": 34955
    },
    {
      "entropy": 5.3835996150970455,
      "epoch": 3.9263772673667656,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003540249676182131,
      "loss": 5.0162,
      "mean_token_accuracy": 0.22074820846319199,
      "num_tokens": 78040021.0,
      "step": 34960
    },
    {
      "entropy": 5.405057191848755,
      "epoch": 3.9269388442747233,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035398737542854936,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20998377948999405,
      "num_tokens": 78050827.0,
      "step": 34965
    },
    {
      "entropy": 5.348815298080444,
      "epoch": 3.927500421182681,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003539497807238969,
      "loss": 4.9992,
      "mean_token_accuracy": 0.2160017505288124,
      "num_tokens": 78062241.0,
      "step": 34970
    },
    {
      "entropy": 5.42320876121521,
      "epoch": 3.9280619980906386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035391218350545287,
      "loss": 5.0611,
      "mean_token_accuracy": 0.21313002854585647,
      "num_tokens": 78073611.0,
      "step": 34975
    },
    {
      "entropy": 5.328111171722412,
      "epoch": 3.9286235749985963,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035387458377441434,
      "loss": 4.8997,
      "mean_token_accuracy": 0.225292606651783,
      "num_tokens": 78084212.0,
      "step": 34980
    },
    {
      "entropy": 5.305535364151001,
      "epoch": 3.9291851519065535,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003538369815319784,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21925136297941208,
      "num_tokens": 78094696.0,
      "step": 34985
    },
    {
      "entropy": 5.281348609924317,
      "epoch": 3.929746728814511,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035379937677934245,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22395480722188948,
      "num_tokens": 78106204.0,
      "step": 34990
    },
    {
      "entropy": 5.415148496627808,
      "epoch": 3.930308305722469,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003537617695177037,
      "loss": 5.1229,
      "mean_token_accuracy": 0.21057090759277344,
      "num_tokens": 78118660.0,
      "step": 34995
    },
    {
      "entropy": 5.316461753845215,
      "epoch": 3.930869882630426,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035372415974825974,
      "loss": 4.883,
      "mean_token_accuracy": 0.22155152559280394,
      "num_tokens": 78129507.0,
      "step": 35000
    },
    {
      "entropy": 5.397950029373169,
      "epoch": 3.9314314595383837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035368654747220795,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21713715344667434,
      "num_tokens": 78140137.0,
      "step": 35005
    },
    {
      "entropy": 5.375945711135865,
      "epoch": 3.9319930364463413,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035364893269074606,
      "loss": 5.0487,
      "mean_token_accuracy": 0.2095806673169136,
      "num_tokens": 78152484.0,
      "step": 35010
    },
    {
      "entropy": 5.423652982711792,
      "epoch": 3.932554613354299,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035361131540507157,
      "loss": 5.0626,
      "mean_token_accuracy": 0.21409189701080322,
      "num_tokens": 78164559.0,
      "step": 35015
    },
    {
      "entropy": 5.338519191741943,
      "epoch": 3.9331161902622567,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003535736956163823,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21765755414962767,
      "num_tokens": 78175236.0,
      "step": 35020
    },
    {
      "entropy": 5.38746657371521,
      "epoch": 3.933677767170214,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035353607332587613,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21620997041463852,
      "num_tokens": 78186753.0,
      "step": 35025
    },
    {
      "entropy": 5.460506629943848,
      "epoch": 3.9342393440781716,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003534984485347509,
      "loss": 5.1641,
      "mean_token_accuracy": 0.20795054882764816,
      "num_tokens": 78199207.0,
      "step": 35030
    },
    {
      "entropy": 5.536807584762573,
      "epoch": 3.934800920986129,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003534608212442046,
      "loss": 5.189,
      "mean_token_accuracy": 0.20577844530344008,
      "num_tokens": 78212177.0,
      "step": 35035
    },
    {
      "entropy": 5.404459142684937,
      "epoch": 3.9353624978940864,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035342319145543533,
      "loss": 4.9738,
      "mean_token_accuracy": 0.21205434650182725,
      "num_tokens": 78223925.0,
      "step": 35040
    },
    {
      "entropy": 5.354441928863525,
      "epoch": 3.935924074802044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035338555916964126,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21574638336896895,
      "num_tokens": 78234593.0,
      "step": 35045
    },
    {
      "entropy": 5.337656450271607,
      "epoch": 3.9364856517100018,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035334792438802056,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21589228212833406,
      "num_tokens": 78245457.0,
      "step": 35050
    },
    {
      "entropy": 5.357191896438598,
      "epoch": 3.937047228617959,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003533102871117716,
      "loss": 4.972,
      "mean_token_accuracy": 0.2216845765709877,
      "num_tokens": 78256394.0,
      "step": 35055
    },
    {
      "entropy": 5.431589555740357,
      "epoch": 3.9376088055259166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003532726473420927,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21491375267505647,
      "num_tokens": 78268889.0,
      "step": 35060
    },
    {
      "entropy": 5.307046604156494,
      "epoch": 3.9381703824338743,
      "grad_norm": 1.125,
      "learning_rate": 0.0003532350050801823,
      "loss": 4.931,
      "mean_token_accuracy": 0.21831417679786683,
      "num_tokens": 78280441.0,
      "step": 35065
    },
    {
      "entropy": 5.354033422470093,
      "epoch": 3.938731959341832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035319736032723905,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21797335594892503,
      "num_tokens": 78291366.0,
      "step": 35070
    },
    {
      "entropy": 5.423819255828858,
      "epoch": 3.9392935362497896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003531597130844616,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21057135611772537,
      "num_tokens": 78301541.0,
      "step": 35075
    },
    {
      "entropy": 5.419572639465332,
      "epoch": 3.939855113157747,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035312206335304843,
      "loss": 5.0247,
      "mean_token_accuracy": 0.2156011864542961,
      "num_tokens": 78311782.0,
      "step": 35080
    },
    {
      "entropy": 5.313531827926636,
      "epoch": 3.9404166900657045,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035308441113419855,
      "loss": 4.973,
      "mean_token_accuracy": 0.21382860839366913,
      "num_tokens": 78321636.0,
      "step": 35085
    },
    {
      "entropy": 5.462100219726563,
      "epoch": 3.940978266973662,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035304675642911074,
      "loss": 5.065,
      "mean_token_accuracy": 0.21820736229419707,
      "num_tokens": 78332882.0,
      "step": 35090
    },
    {
      "entropy": 5.4580240726470945,
      "epoch": 3.9415398438816194,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003530090992389838,
      "loss": 5.1077,
      "mean_token_accuracy": 0.21289365589618683,
      "num_tokens": 78345122.0,
      "step": 35095
    },
    {
      "entropy": 5.424880123138427,
      "epoch": 3.942101420789577,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000352971439565017,
      "loss": 5.0437,
      "mean_token_accuracy": 0.21755089312791825,
      "num_tokens": 78355730.0,
      "step": 35100
    },
    {
      "entropy": 5.33050765991211,
      "epoch": 3.9426629976975347,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035293377740840934,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22511381953954696,
      "num_tokens": 78366651.0,
      "step": 35105
    },
    {
      "entropy": 5.31247501373291,
      "epoch": 3.9432245746054924,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035289611277036,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22682802230119706,
      "num_tokens": 78377633.0,
      "step": 35110
    },
    {
      "entropy": 5.351218271255493,
      "epoch": 3.94378615151345,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003528584456520681,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2156555250287056,
      "num_tokens": 78389038.0,
      "step": 35115
    },
    {
      "entropy": 5.311679220199585,
      "epoch": 3.9443477284214072,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035282077605473317,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2086957722902298,
      "num_tokens": 78401045.0,
      "step": 35120
    },
    {
      "entropy": 5.354515314102173,
      "epoch": 3.944909305329365,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003527831039795545,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20779362320899963,
      "num_tokens": 78413013.0,
      "step": 35125
    },
    {
      "entropy": 5.284785604476928,
      "epoch": 3.9454708822373226,
      "grad_norm": 1.0,
      "learning_rate": 0.0003527454294277316,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22698598206043244,
      "num_tokens": 78423794.0,
      "step": 35130
    },
    {
      "entropy": 5.332569646835327,
      "epoch": 3.94603245914528,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003527077524004642,
      "loss": 4.9551,
      "mean_token_accuracy": 0.22336965054273605,
      "num_tokens": 78434902.0,
      "step": 35135
    },
    {
      "entropy": 5.347472286224365,
      "epoch": 3.9465940360532374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035267007289895176,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2202248677611351,
      "num_tokens": 78446114.0,
      "step": 35140
    },
    {
      "entropy": 5.342621183395385,
      "epoch": 3.947155612961195,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035263239092439417,
      "loss": 4.9139,
      "mean_token_accuracy": 0.228553931415081,
      "num_tokens": 78456882.0,
      "step": 35145
    },
    {
      "entropy": 5.3503124713897705,
      "epoch": 3.9477171898691523,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035259470647799104,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22682185471057892,
      "num_tokens": 78467416.0,
      "step": 35150
    },
    {
      "entropy": 5.274475383758545,
      "epoch": 3.94827876677711,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035255701956094236,
      "loss": 4.9135,
      "mean_token_accuracy": 0.22802285104990005,
      "num_tokens": 78477364.0,
      "step": 35155
    },
    {
      "entropy": 5.311997556686402,
      "epoch": 3.9488403436850676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035251933017444814,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21983757764101028,
      "num_tokens": 78488396.0,
      "step": 35160
    },
    {
      "entropy": 5.349977922439575,
      "epoch": 3.9494019205930253,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035248163831970837,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21643757224082946,
      "num_tokens": 78498586.0,
      "step": 35165
    },
    {
      "entropy": 5.398192310333252,
      "epoch": 3.949963497500983,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035244394399792304,
      "loss": 4.9437,
      "mean_token_accuracy": 0.2248475506901741,
      "num_tokens": 78509664.0,
      "step": 35170
    },
    {
      "entropy": 5.479954433441162,
      "epoch": 3.95052507440894,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003524062472102927,
      "loss": 5.1665,
      "mean_token_accuracy": 0.21790126264095305,
      "num_tokens": 78521950.0,
      "step": 35175
    },
    {
      "entropy": 5.369222974777221,
      "epoch": 3.951086651316898,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003523685479580174,
      "loss": 5.0618,
      "mean_token_accuracy": 0.2112535610795021,
      "num_tokens": 78532943.0,
      "step": 35180
    },
    {
      "entropy": 5.277353143692016,
      "epoch": 3.9516482282248555,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003523308462422974,
      "loss": 4.8606,
      "mean_token_accuracy": 0.2257692039012909,
      "num_tokens": 78544042.0,
      "step": 35185
    },
    {
      "entropy": 5.304129314422608,
      "epoch": 3.9522098051328127,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003522931420643334,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22221378833055497,
      "num_tokens": 78554812.0,
      "step": 35190
    },
    {
      "entropy": 5.262289428710938,
      "epoch": 3.9527713820407704,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003522554354253257,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2225796490907669,
      "num_tokens": 78565036.0,
      "step": 35195
    },
    {
      "entropy": 5.206549501419067,
      "epoch": 3.953332958948728,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035221772632647503,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22893341034650802,
      "num_tokens": 78575420.0,
      "step": 35200
    },
    {
      "entropy": 5.386279201507568,
      "epoch": 3.9538945358566857,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035218001476898194,
      "loss": 5.0296,
      "mean_token_accuracy": 0.22270840257406235,
      "num_tokens": 78586670.0,
      "step": 35205
    },
    {
      "entropy": 5.27560567855835,
      "epoch": 3.9544561127646434,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003521423007540472,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22069099098443984,
      "num_tokens": 78598048.0,
      "step": 35210
    },
    {
      "entropy": 5.314364004135132,
      "epoch": 3.9550176896726006,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003521045842828718,
      "loss": 5.0255,
      "mean_token_accuracy": 0.21850430220365524,
      "num_tokens": 78610364.0,
      "step": 35215
    },
    {
      "entropy": 5.360458278656006,
      "epoch": 3.9555792665805583,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003520668653566565,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2133463904261589,
      "num_tokens": 78622117.0,
      "step": 35220
    },
    {
      "entropy": 5.343085861206054,
      "epoch": 3.956140843488516,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003520291439766024,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2217706948518753,
      "num_tokens": 78632894.0,
      "step": 35225
    },
    {
      "entropy": 5.331859397888183,
      "epoch": 3.956702420396473,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035199142014391034,
      "loss": 4.953,
      "mean_token_accuracy": 0.22255935817956923,
      "num_tokens": 78643851.0,
      "step": 35230
    },
    {
      "entropy": 5.340222120285034,
      "epoch": 3.957263997304431,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035195369385978165,
      "loss": 5.0032,
      "mean_token_accuracy": 0.21705972254276276,
      "num_tokens": 78655097.0,
      "step": 35235
    },
    {
      "entropy": 5.2687132358551025,
      "epoch": 3.9578255742123885,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00035191596512541747,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22267247289419173,
      "num_tokens": 78667572.0,
      "step": 35240
    },
    {
      "entropy": 5.4215349674224855,
      "epoch": 3.9583871511203457,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003518782339420192,
      "loss": 5.0161,
      "mean_token_accuracy": 0.21417624205350877,
      "num_tokens": 78678858.0,
      "step": 35245
    },
    {
      "entropy": 5.318108177185058,
      "epoch": 3.9589487280283033,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035184050031078815,
      "loss": 4.8339,
      "mean_token_accuracy": 0.2234713226556778,
      "num_tokens": 78688074.0,
      "step": 35250
    },
    {
      "entropy": 5.228628015518188,
      "epoch": 3.959510304936261,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035180276423292566,
      "loss": 4.9223,
      "mean_token_accuracy": 0.21468404233455657,
      "num_tokens": 78698805.0,
      "step": 35255
    },
    {
      "entropy": 5.181595230102539,
      "epoch": 3.9600718818442187,
      "grad_norm": 1.0,
      "learning_rate": 0.0003517650257096334,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22290850430727005,
      "num_tokens": 78710636.0,
      "step": 35260
    },
    {
      "entropy": 5.292023038864135,
      "epoch": 3.9606334587521763,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035172728474211306,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2233498752117157,
      "num_tokens": 78721464.0,
      "step": 35265
    },
    {
      "entropy": 5.395468616485596,
      "epoch": 3.9611950356601335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035168954133156613,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21887556612491607,
      "num_tokens": 78733306.0,
      "step": 35270
    },
    {
      "entropy": 5.42512617111206,
      "epoch": 3.961756612568091,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003516517954791944,
      "loss": 5.0836,
      "mean_token_accuracy": 0.21382859945297242,
      "num_tokens": 78745254.0,
      "step": 35275
    },
    {
      "entropy": 5.309650039672851,
      "epoch": 3.962318189476049,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035161404718619986,
      "loss": 4.911,
      "mean_token_accuracy": 0.22636700570583343,
      "num_tokens": 78755796.0,
      "step": 35280
    },
    {
      "entropy": 5.290368461608887,
      "epoch": 3.962879766384006,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003515762964537844,
      "loss": 4.9801,
      "mean_token_accuracy": 0.21700799614191055,
      "num_tokens": 78767394.0,
      "step": 35285
    },
    {
      "entropy": 5.367566013336182,
      "epoch": 3.9634413432919637,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003515385432831499,
      "loss": 5.0192,
      "mean_token_accuracy": 0.22364734560251237,
      "num_tokens": 78779017.0,
      "step": 35290
    },
    {
      "entropy": 5.415853214263916,
      "epoch": 3.9640029201999214,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003515007876754984,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21351849138736725,
      "num_tokens": 78789964.0,
      "step": 35295
    },
    {
      "entropy": 5.33529953956604,
      "epoch": 3.964564497107879,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003514630296320323,
      "loss": 4.996,
      "mean_token_accuracy": 0.21875876635313035,
      "num_tokens": 78801657.0,
      "step": 35300
    },
    {
      "entropy": 5.393939352035522,
      "epoch": 3.9651260740158367,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003514252691539536,
      "loss": 5.1088,
      "mean_token_accuracy": 0.22044118344783784,
      "num_tokens": 78812396.0,
      "step": 35305
    },
    {
      "entropy": 5.310418605804443,
      "epoch": 3.965687650923794,
      "grad_norm": 1.0,
      "learning_rate": 0.00035138750624246466,
      "loss": 4.9838,
      "mean_token_accuracy": 0.2133146271109581,
      "num_tokens": 78823711.0,
      "step": 35310
    },
    {
      "entropy": 5.36651463508606,
      "epoch": 3.9662492278317516,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000351349740898768,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21718451678752898,
      "num_tokens": 78835394.0,
      "step": 35315
    },
    {
      "entropy": 5.32029619216919,
      "epoch": 3.9668108047397093,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003513119731240659,
      "loss": 4.9729,
      "mean_token_accuracy": 0.22351225167512895,
      "num_tokens": 78846180.0,
      "step": 35320
    },
    {
      "entropy": 5.290055847167968,
      "epoch": 3.9673723816476665,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035127420291956104,
      "loss": 4.9539,
      "mean_token_accuracy": 0.22187353521585465,
      "num_tokens": 78856537.0,
      "step": 35325
    },
    {
      "entropy": 5.328797817230225,
      "epoch": 3.967933958555624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003512364302864559,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21677847057580948,
      "num_tokens": 78867416.0,
      "step": 35330
    },
    {
      "entropy": 5.349091911315918,
      "epoch": 3.968495535463582,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003511986552259533,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2165856510400772,
      "num_tokens": 78878725.0,
      "step": 35335
    },
    {
      "entropy": 5.339759635925293,
      "epoch": 3.969057112371539,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035116087773925594,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21742845922708512,
      "num_tokens": 78889659.0,
      "step": 35340
    },
    {
      "entropy": 5.249183988571167,
      "epoch": 3.9696186892794967,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003511230978275667,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22357446253299712,
      "num_tokens": 78900056.0,
      "step": 35345
    },
    {
      "entropy": 5.301784372329712,
      "epoch": 3.9701802661874543,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035108531549208846,
      "loss": 4.959,
      "mean_token_accuracy": 0.22608496397733688,
      "num_tokens": 78911343.0,
      "step": 35350
    },
    {
      "entropy": 5.336179876327515,
      "epoch": 3.970741843095412,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035104753073402433,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21938952505588533,
      "num_tokens": 78921580.0,
      "step": 35355
    },
    {
      "entropy": 5.366355085372925,
      "epoch": 3.9713034200033697,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003510097435545774,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22171651273965837,
      "num_tokens": 78932810.0,
      "step": 35360
    },
    {
      "entropy": 5.332621240615845,
      "epoch": 3.971864996911327,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035097195395495057,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21959864795207978,
      "num_tokens": 78945100.0,
      "step": 35365
    },
    {
      "entropy": 5.388849210739136,
      "epoch": 3.9724265738192845,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003509341619363474,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21557445973157882,
      "num_tokens": 78957284.0,
      "step": 35370
    },
    {
      "entropy": 5.391395616531372,
      "epoch": 3.972988150727242,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035089636749997095,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2293539524078369,
      "num_tokens": 78968639.0,
      "step": 35375
    },
    {
      "entropy": 5.25002121925354,
      "epoch": 3.9735497276351994,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035085857064702487,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22301720231771469,
      "num_tokens": 78978323.0,
      "step": 35380
    },
    {
      "entropy": 5.337507677078247,
      "epoch": 3.974111304543157,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003508207713787123,
      "loss": 5.0145,
      "mean_token_accuracy": 0.2187003791332245,
      "num_tokens": 78989193.0,
      "step": 35385
    },
    {
      "entropy": 5.295000648498535,
      "epoch": 3.9746728814511147,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035078296969623716,
      "loss": 4.8337,
      "mean_token_accuracy": 0.23241523057222366,
      "num_tokens": 78999640.0,
      "step": 35390
    },
    {
      "entropy": 5.25071005821228,
      "epoch": 3.9752344583590724,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035074516560080273,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22764895260334014,
      "num_tokens": 79010468.0,
      "step": 35395
    },
    {
      "entropy": 5.272427272796631,
      "epoch": 3.97579603526703,
      "grad_norm": 1.0,
      "learning_rate": 0.0003507073590936129,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21107348948717117,
      "num_tokens": 79021399.0,
      "step": 35400
    },
    {
      "entropy": 5.299168586730957,
      "epoch": 3.9763576121749873,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035066955017587135,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2204820141196251,
      "num_tokens": 79031643.0,
      "step": 35405
    },
    {
      "entropy": 5.475370264053344,
      "epoch": 3.976919189082945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000350631738848782,
      "loss": 5.1365,
      "mean_token_accuracy": 0.21205174028873444,
      "num_tokens": 79041911.0,
      "step": 35410
    },
    {
      "entropy": 5.413081312179566,
      "epoch": 3.9774807659909026,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003505939251135488,
      "loss": 5.0805,
      "mean_token_accuracy": 0.2110368400812149,
      "num_tokens": 79054698.0,
      "step": 35415
    },
    {
      "entropy": 5.486343288421631,
      "epoch": 3.97804234289886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003505561089713756,
      "loss": 5.1043,
      "mean_token_accuracy": 0.2069486603140831,
      "num_tokens": 79065855.0,
      "step": 35420
    },
    {
      "entropy": 5.476979207992554,
      "epoch": 3.9786039198068175,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003505182904234667,
      "loss": 5.0669,
      "mean_token_accuracy": 0.2124079167842865,
      "num_tokens": 79077332.0,
      "step": 35425
    },
    {
      "entropy": 5.352615690231323,
      "epoch": 3.979165496714775,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003504804694710261,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21859672665596008,
      "num_tokens": 79088996.0,
      "step": 35430
    },
    {
      "entropy": 5.342862224578857,
      "epoch": 3.9797270736227324,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003504426461152581,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2223045065999031,
      "num_tokens": 79100828.0,
      "step": 35435
    },
    {
      "entropy": 5.312425899505615,
      "epoch": 3.98028865053069,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003504048203573669,
      "loss": 4.9387,
      "mean_token_accuracy": 0.222202305495739,
      "num_tokens": 79112297.0,
      "step": 35440
    },
    {
      "entropy": 5.350941801071167,
      "epoch": 3.9808502274386477,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003503669921985571,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21543802618980407,
      "num_tokens": 79122795.0,
      "step": 35445
    },
    {
      "entropy": 5.310916948318481,
      "epoch": 3.9814118043466054,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000350329161640033,
      "loss": 4.961,
      "mean_token_accuracy": 0.21896079927682877,
      "num_tokens": 79133652.0,
      "step": 35450
    },
    {
      "entropy": 5.328591394424438,
      "epoch": 3.981973381254563,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035029132868299914,
      "loss": 4.9378,
      "mean_token_accuracy": 0.224832683801651,
      "num_tokens": 79144107.0,
      "step": 35455
    },
    {
      "entropy": 5.35670166015625,
      "epoch": 3.9825349581625202,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035025349332866017,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21856579929590225,
      "num_tokens": 79155900.0,
      "step": 35460
    },
    {
      "entropy": 5.356700229644775,
      "epoch": 3.983096535070478,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00035021565557822085,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2162473350763321,
      "num_tokens": 79167425.0,
      "step": 35465
    },
    {
      "entropy": 5.363798427581787,
      "epoch": 3.9836581119784356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003501778154328858,
      "loss": 5.0609,
      "mean_token_accuracy": 0.2220173954963684,
      "num_tokens": 79178880.0,
      "step": 35470
    },
    {
      "entropy": 5.3401697158813475,
      "epoch": 3.9842196888863928,
      "grad_norm": 0.875,
      "learning_rate": 0.0003501399728938599,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22661043405532838,
      "num_tokens": 79190210.0,
      "step": 35475
    },
    {
      "entropy": 5.31838903427124,
      "epoch": 3.9847812657943504,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035010212796234827,
      "loss": 4.9423,
      "mean_token_accuracy": 0.2223057195544243,
      "num_tokens": 79200497.0,
      "step": 35480
    },
    {
      "entropy": 5.280216360092163,
      "epoch": 3.985342842702308,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003500642806395557,
      "loss": 4.9589,
      "mean_token_accuracy": 0.22691920548677444,
      "num_tokens": 79212305.0,
      "step": 35485
    },
    {
      "entropy": 5.426902008056641,
      "epoch": 3.9859044196102658,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003500264309266873,
      "loss": 4.9723,
      "mean_token_accuracy": 0.22065745741128923,
      "num_tokens": 79222968.0,
      "step": 35490
    },
    {
      "entropy": 5.423346042633057,
      "epoch": 3.9864659965182234,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034998857882494823,
      "loss": 5.0106,
      "mean_token_accuracy": 0.2183726578950882,
      "num_tokens": 79235064.0,
      "step": 35495
    },
    {
      "entropy": 5.365753841400147,
      "epoch": 3.9870275734261806,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034995072433554377,
      "loss": 4.962,
      "mean_token_accuracy": 0.21878813505172728,
      "num_tokens": 79245761.0,
      "step": 35500
    },
    {
      "entropy": 5.320128011703491,
      "epoch": 3.9875891503341383,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003499128674596792,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2271315947175026,
      "num_tokens": 79256070.0,
      "step": 35505
    },
    {
      "entropy": 5.333623838424683,
      "epoch": 3.988150727242096,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034987500819855976,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21460554599761963,
      "num_tokens": 79266322.0,
      "step": 35510
    },
    {
      "entropy": 5.4118382930755615,
      "epoch": 3.988712304150053,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034983714655339115,
      "loss": 5.0461,
      "mean_token_accuracy": 0.21961329281330108,
      "num_tokens": 79277272.0,
      "step": 35515
    },
    {
      "entropy": 5.40453143119812,
      "epoch": 3.989273881058011,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003497992825253787,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21363062411546707,
      "num_tokens": 79288463.0,
      "step": 35520
    },
    {
      "entropy": 5.393286418914795,
      "epoch": 3.9898354579659685,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034976141611572817,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22115608006715776,
      "num_tokens": 79298655.0,
      "step": 35525
    },
    {
      "entropy": 5.316653680801392,
      "epoch": 3.9903970348739257,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00034972354732564506,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22825133800506592,
      "num_tokens": 79309731.0,
      "step": 35530
    },
    {
      "entropy": 5.332367372512818,
      "epoch": 3.9909586117818834,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00034968567615633516,
      "loss": 5.0021,
      "mean_token_accuracy": 0.22008467018604277,
      "num_tokens": 79321712.0,
      "step": 35535
    },
    {
      "entropy": 5.264921188354492,
      "epoch": 3.991520188689841,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034964780260900445,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22396388798952102,
      "num_tokens": 79333159.0,
      "step": 35540
    },
    {
      "entropy": 5.399532175064087,
      "epoch": 3.9920817655977987,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034960992668485875,
      "loss": 4.9752,
      "mean_token_accuracy": 0.2214301973581314,
      "num_tokens": 79344012.0,
      "step": 35545
    },
    {
      "entropy": 5.347935199737549,
      "epoch": 3.9926433425057564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034957204838510406,
      "loss": 5.0309,
      "mean_token_accuracy": 0.21926185190677644,
      "num_tokens": 79355162.0,
      "step": 35550
    },
    {
      "entropy": 5.324479007720948,
      "epoch": 3.9932049194137136,
      "grad_norm": 1.0,
      "learning_rate": 0.00034953416771094645,
      "loss": 4.9017,
      "mean_token_accuracy": 0.2238995462656021,
      "num_tokens": 79365605.0,
      "step": 35555
    },
    {
      "entropy": 5.312754917144775,
      "epoch": 3.9937664963216712,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000349496284663592,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21817588061094284,
      "num_tokens": 79376548.0,
      "step": 35560
    },
    {
      "entropy": 5.343481397628784,
      "epoch": 3.994328073229629,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034945839924424696,
      "loss": 4.897,
      "mean_token_accuracy": 0.22986447364091872,
      "num_tokens": 79388352.0,
      "step": 35565
    },
    {
      "entropy": 5.330991125106811,
      "epoch": 3.994889650137586,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003494205114541176,
      "loss": 4.959,
      "mean_token_accuracy": 0.22196414321660995,
      "num_tokens": 79399110.0,
      "step": 35570
    },
    {
      "entropy": 5.34568305015564,
      "epoch": 3.995451227045544,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034938262129441026,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21151814758777618,
      "num_tokens": 79410350.0,
      "step": 35575
    },
    {
      "entropy": 5.352759599685669,
      "epoch": 3.9960128039535014,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003493447287663315,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22151091396808625,
      "num_tokens": 79421015.0,
      "step": 35580
    },
    {
      "entropy": 5.363454055786133,
      "epoch": 3.996574380861459,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003493068338710877,
      "loss": 5.029,
      "mean_token_accuracy": 0.2223750188946724,
      "num_tokens": 79432200.0,
      "step": 35585
    },
    {
      "entropy": 5.39575834274292,
      "epoch": 3.9971359577694168,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003492689366098855,
      "loss": 5.0842,
      "mean_token_accuracy": 0.21461691707372665,
      "num_tokens": 79445121.0,
      "step": 35590
    },
    {
      "entropy": 5.309022617340088,
      "epoch": 3.997697534677374,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034923103698393147,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22232760787010192,
      "num_tokens": 79456533.0,
      "step": 35595
    },
    {
      "entropy": 5.300660610198975,
      "epoch": 3.9982591115853316,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003491931349944325,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22674213945865632,
      "num_tokens": 79466365.0,
      "step": 35600
    },
    {
      "entropy": 5.408140659332275,
      "epoch": 3.9988206884932893,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034915523064259533,
      "loss": 5.0609,
      "mean_token_accuracy": 0.21128993332386017,
      "num_tokens": 79477203.0,
      "step": 35605
    },
    {
      "entropy": 5.256291198730469,
      "epoch": 3.9993822654012465,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003491173239296268,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21757977604866027,
      "num_tokens": 79490062.0,
      "step": 35610
    },
    {
      "entropy": 5.402770853042602,
      "epoch": 3.999943842309204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000349079414856734,
      "loss": 5.025,
      "mean_token_accuracy": 0.21764834374189376,
      "num_tokens": 79501727.0,
      "step": 35615
    },
    {
      "entropy": 5.350081973605686,
      "epoch": 4.000449261526366,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003490415034251239,
      "loss": 4.8058,
      "mean_token_accuracy": 0.23239662746588388,
      "num_tokens": 79510304.0,
      "step": 35620
    },
    {
      "entropy": 5.362659168243408,
      "epoch": 4.001010838434324,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003490035896360036,
      "loss": 4.9398,
      "mean_token_accuracy": 0.21907162368297578,
      "num_tokens": 79521811.0,
      "step": 35625
    },
    {
      "entropy": 5.310119724273681,
      "epoch": 4.001572415342281,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034896567349058027,
      "loss": 4.8311,
      "mean_token_accuracy": 0.23497136533260346,
      "num_tokens": 79532756.0,
      "step": 35630
    },
    {
      "entropy": 5.3906646251678465,
      "epoch": 4.002133992250239,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034892775499006126,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21891241222620011,
      "num_tokens": 79544945.0,
      "step": 35635
    },
    {
      "entropy": 5.36221752166748,
      "epoch": 4.002695569158196,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003488898341356538,
      "loss": 5.0011,
      "mean_token_accuracy": 0.22100998610258102,
      "num_tokens": 79555374.0,
      "step": 35640
    },
    {
      "entropy": 5.3484093189239506,
      "epoch": 4.0032571460661535,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00034885191092856545,
      "loss": 4.8984,
      "mean_token_accuracy": 0.22541451156139375,
      "num_tokens": 79566598.0,
      "step": 35645
    },
    {
      "entropy": 5.390035963058471,
      "epoch": 4.003818722974112,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003488139853700036,
      "loss": 4.901,
      "mean_token_accuracy": 0.23107196539640426,
      "num_tokens": 79577859.0,
      "step": 35650
    },
    {
      "entropy": 5.313963317871094,
      "epoch": 4.004380299882069,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034877605746117576,
      "loss": 4.8815,
      "mean_token_accuracy": 0.2243813991546631,
      "num_tokens": 79588704.0,
      "step": 35655
    },
    {
      "entropy": 5.187617635726928,
      "epoch": 4.004941876790026,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003487381272032897,
      "loss": 4.7392,
      "mean_token_accuracy": 0.2395566537976265,
      "num_tokens": 79599493.0,
      "step": 35660
    },
    {
      "entropy": 5.34735255241394,
      "epoch": 4.005503453697984,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000348700194597553,
      "loss": 4.8921,
      "mean_token_accuracy": 0.23031046837568284,
      "num_tokens": 79610305.0,
      "step": 35665
    },
    {
      "entropy": 5.415438222885132,
      "epoch": 4.006065030605941,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00034866225964517354,
      "loss": 4.9923,
      "mean_token_accuracy": 0.22049729526042938,
      "num_tokens": 79622077.0,
      "step": 35670
    },
    {
      "entropy": 5.4395647048950195,
      "epoch": 4.006626607513899,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003486243223473592,
      "loss": 5.0318,
      "mean_token_accuracy": 0.22212042659521103,
      "num_tokens": 79634999.0,
      "step": 35675
    },
    {
      "entropy": 5.324721097946167,
      "epoch": 4.007188184421857,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034858638270531786,
      "loss": 4.8146,
      "mean_token_accuracy": 0.23227564990520477,
      "num_tokens": 79646152.0,
      "step": 35680
    },
    {
      "entropy": 5.337189197540283,
      "epoch": 4.007749761329814,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034854844072025755,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22656266689300536,
      "num_tokens": 79657941.0,
      "step": 35685
    },
    {
      "entropy": 5.30948281288147,
      "epoch": 4.008311338237772,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034851049639338637,
      "loss": 4.8159,
      "mean_token_accuracy": 0.23552124500274657,
      "num_tokens": 79667713.0,
      "step": 35690
    },
    {
      "entropy": 5.3425657749176025,
      "epoch": 4.008872915145729,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003484725497259125,
      "loss": 4.878,
      "mean_token_accuracy": 0.2266979619860649,
      "num_tokens": 79679351.0,
      "step": 35695
    },
    {
      "entropy": 5.384876346588134,
      "epoch": 4.009434492053686,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034843460071904407,
      "loss": 5.0146,
      "mean_token_accuracy": 0.21544950753450393,
      "num_tokens": 79689406.0,
      "step": 35700
    },
    {
      "entropy": 5.3059299945831295,
      "epoch": 4.0099960689616445,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034839664937398964,
      "loss": 4.8135,
      "mean_token_accuracy": 0.2257569268345833,
      "num_tokens": 79700231.0,
      "step": 35705
    },
    {
      "entropy": 5.323565912246704,
      "epoch": 4.010557645869602,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003483586956919572,
      "loss": 4.901,
      "mean_token_accuracy": 0.22826377004384996,
      "num_tokens": 79711721.0,
      "step": 35710
    },
    {
      "entropy": 5.214164304733276,
      "epoch": 4.011119222777559,
      "grad_norm": 1.0,
      "learning_rate": 0.0003483207396741556,
      "loss": 4.7843,
      "mean_token_accuracy": 0.2241688370704651,
      "num_tokens": 79721795.0,
      "step": 35715
    },
    {
      "entropy": 5.414013767242432,
      "epoch": 4.011680799685517,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003482827813217932,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21992823779582976,
      "num_tokens": 79734356.0,
      "step": 35720
    },
    {
      "entropy": 5.2959051609039305,
      "epoch": 4.012242376593474,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034824482063607856,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22849412262439728,
      "num_tokens": 79745847.0,
      "step": 35725
    },
    {
      "entropy": 5.2945802211761475,
      "epoch": 4.012803953501432,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003482068576182204,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2229880303144455,
      "num_tokens": 79756580.0,
      "step": 35730
    },
    {
      "entropy": 5.378960514068604,
      "epoch": 4.01336553040939,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034816889226942755,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22178370058536528,
      "num_tokens": 79767912.0,
      "step": 35735
    },
    {
      "entropy": 5.407948589324951,
      "epoch": 4.013927107317347,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003481309245909088,
      "loss": 4.9987,
      "mean_token_accuracy": 0.22079693377017975,
      "num_tokens": 79778572.0,
      "step": 35740
    },
    {
      "entropy": 5.326175737380981,
      "epoch": 4.014488684225305,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003480929545838731,
      "loss": 4.8567,
      "mean_token_accuracy": 0.23198736310005189,
      "num_tokens": 79789237.0,
      "step": 35745
    },
    {
      "entropy": 5.4087996006011965,
      "epoch": 4.015050261133262,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003480549822495293,
      "loss": 5.074,
      "mean_token_accuracy": 0.21715234220027924,
      "num_tokens": 79800271.0,
      "step": 35750
    },
    {
      "entropy": 5.38117036819458,
      "epoch": 4.015611838041219,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00034801700758908653,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22003119587898254,
      "num_tokens": 79811378.0,
      "step": 35755
    },
    {
      "entropy": 5.366988706588745,
      "epoch": 4.0161734149491775,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034797903060375396,
      "loss": 4.9734,
      "mean_token_accuracy": 0.22263508141040803,
      "num_tokens": 79822795.0,
      "step": 35760
    },
    {
      "entropy": 5.3617630958557125,
      "epoch": 4.016734991857135,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003479410512947407,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21751251369714736,
      "num_tokens": 79833331.0,
      "step": 35765
    },
    {
      "entropy": 5.285930347442627,
      "epoch": 4.017296568765093,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034790306966325606,
      "loss": 4.8733,
      "mean_token_accuracy": 0.22795773595571517,
      "num_tokens": 79844079.0,
      "step": 35770
    },
    {
      "entropy": 5.363169574737549,
      "epoch": 4.01785814567305,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003478650857105095,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2281282901763916,
      "num_tokens": 79855275.0,
      "step": 35775
    },
    {
      "entropy": 5.361121749877929,
      "epoch": 4.018419722581007,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003478270994377102,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22439662218093873,
      "num_tokens": 79866147.0,
      "step": 35780
    },
    {
      "entropy": 5.347488737106323,
      "epoch": 4.018981299488965,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034778911084606793,
      "loss": 4.8816,
      "mean_token_accuracy": 0.2261527433991432,
      "num_tokens": 79876791.0,
      "step": 35785
    },
    {
      "entropy": 5.34072642326355,
      "epoch": 4.0195428763969225,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003477511199367922,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22676541805267333,
      "num_tokens": 79887817.0,
      "step": 35790
    },
    {
      "entropy": 5.2713642597198485,
      "epoch": 4.02010445330488,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00034771312671109256,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2210535943508148,
      "num_tokens": 79899055.0,
      "step": 35795
    },
    {
      "entropy": 5.3012939453125,
      "epoch": 4.020666030212838,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003476751311701787,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22702090442180634,
      "num_tokens": 79912066.0,
      "step": 35800
    },
    {
      "entropy": 5.398511219024658,
      "epoch": 4.021227607120795,
      "grad_norm": 1.0,
      "learning_rate": 0.0003476371333152605,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21970458328723907,
      "num_tokens": 79922845.0,
      "step": 35805
    },
    {
      "entropy": 5.354999589920044,
      "epoch": 4.021789184028753,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034759913314754776,
      "loss": 4.9484,
      "mean_token_accuracy": 0.2208826094865799,
      "num_tokens": 79934732.0,
      "step": 35810
    },
    {
      "entropy": 5.391208219528198,
      "epoch": 4.02235076093671,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034756113066825045,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2246980145573616,
      "num_tokens": 79945182.0,
      "step": 35815
    },
    {
      "entropy": 5.373191881179809,
      "epoch": 4.022912337844668,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003475231258785787,
      "loss": 4.973,
      "mean_token_accuracy": 0.22054938673973085,
      "num_tokens": 79956929.0,
      "step": 35820
    },
    {
      "entropy": 5.332314825057983,
      "epoch": 4.023473914752626,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003474851187797424,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21810028851032257,
      "num_tokens": 79968361.0,
      "step": 35825
    },
    {
      "entropy": 5.337249803543091,
      "epoch": 4.024035491660583,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034744710937295185,
      "loss": 4.954,
      "mean_token_accuracy": 0.22067830264568328,
      "num_tokens": 79978833.0,
      "step": 35830
    },
    {
      "entropy": 5.302785110473633,
      "epoch": 4.02459706856854,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034740909765941725,
      "loss": 4.9175,
      "mean_token_accuracy": 0.22347631752490998,
      "num_tokens": 79989997.0,
      "step": 35835
    },
    {
      "entropy": 5.283454322814942,
      "epoch": 4.025158645476498,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003473710836403489,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22082237303256988,
      "num_tokens": 80001218.0,
      "step": 35840
    },
    {
      "entropy": 5.464446067810059,
      "epoch": 4.0257202223844555,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003473330673169572,
      "loss": 5.0486,
      "mean_token_accuracy": 0.2181730479001999,
      "num_tokens": 80013906.0,
      "step": 35845
    },
    {
      "entropy": 5.347693634033203,
      "epoch": 4.026281799292413,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003472950486904525,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21553568989038469,
      "num_tokens": 80026384.0,
      "step": 35850
    },
    {
      "entropy": 5.34577054977417,
      "epoch": 4.026843376200371,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003472570277620455,
      "loss": 4.8617,
      "mean_token_accuracy": 0.23103515803813934,
      "num_tokens": 80037718.0,
      "step": 35855
    },
    {
      "entropy": 5.346845245361328,
      "epoch": 4.027404953108328,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003472190045329466,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22339367419481276,
      "num_tokens": 80049081.0,
      "step": 35860
    },
    {
      "entropy": 5.254318475723267,
      "epoch": 4.027966530016286,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003471809790043667,
      "loss": 4.8749,
      "mean_token_accuracy": 0.225941401720047,
      "num_tokens": 80060435.0,
      "step": 35865
    },
    {
      "entropy": 5.36871976852417,
      "epoch": 4.028528106924243,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034714295117751635,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22826303094625472,
      "num_tokens": 80070657.0,
      "step": 35870
    },
    {
      "entropy": 5.294485282897949,
      "epoch": 4.029089683832201,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034710492105360653,
      "loss": 4.8304,
      "mean_token_accuracy": 0.23098201900720597,
      "num_tokens": 80080265.0,
      "step": 35875
    },
    {
      "entropy": 5.312833547592163,
      "epoch": 4.029651260740159,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003470668886338481,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22407266497612,
      "num_tokens": 80091917.0,
      "step": 35880
    },
    {
      "entropy": 5.296406841278076,
      "epoch": 4.030212837648116,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034702885391945184,
      "loss": 4.918,
      "mean_token_accuracy": 0.21938593089580535,
      "num_tokens": 80103919.0,
      "step": 35885
    },
    {
      "entropy": 5.452708148956299,
      "epoch": 4.030774414556073,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000346990816911629,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2213690012693405,
      "num_tokens": 80115926.0,
      "step": 35890
    },
    {
      "entropy": 5.31996603012085,
      "epoch": 4.031335991464031,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003469527776115906,
      "loss": 4.8645,
      "mean_token_accuracy": 0.22716038823127746,
      "num_tokens": 80127007.0,
      "step": 35895
    },
    {
      "entropy": 5.253648042678833,
      "epoch": 4.031897568371988,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034691473602054793,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22362604290246962,
      "num_tokens": 80138442.0,
      "step": 35900
    },
    {
      "entropy": 5.3670188903808596,
      "epoch": 4.032459145279946,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003468766921397121,
      "loss": 4.9274,
      "mean_token_accuracy": 0.223621965944767,
      "num_tokens": 80148395.0,
      "step": 35905
    },
    {
      "entropy": 5.363602256774902,
      "epoch": 4.033020722187904,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003468386459702946,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22143317461013795,
      "num_tokens": 80159348.0,
      "step": 35910
    },
    {
      "entropy": 5.358467292785645,
      "epoch": 4.033582299095861,
      "grad_norm": 1.0,
      "learning_rate": 0.00034680059751350664,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2279469832777977,
      "num_tokens": 80169772.0,
      "step": 35915
    },
    {
      "entropy": 5.322882413864136,
      "epoch": 4.034143876003819,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034676254677055984,
      "loss": 4.8959,
      "mean_token_accuracy": 0.22922544181346893,
      "num_tokens": 80180538.0,
      "step": 35920
    },
    {
      "entropy": 5.4054646492004395,
      "epoch": 4.034705452911776,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034672449374266575,
      "loss": 5.0806,
      "mean_token_accuracy": 0.21007412075996398,
      "num_tokens": 80191693.0,
      "step": 35925
    },
    {
      "entropy": 5.339513206481934,
      "epoch": 4.0352670298197335,
      "grad_norm": 1.0,
      "learning_rate": 0.0003466864384310359,
      "loss": 4.8567,
      "mean_token_accuracy": 0.23149998039007186,
      "num_tokens": 80201923.0,
      "step": 35930
    },
    {
      "entropy": 5.374654388427734,
      "epoch": 4.035828606727692,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00034664838083688223,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22639152109622956,
      "num_tokens": 80213588.0,
      "step": 35935
    },
    {
      "entropy": 5.301891326904297,
      "epoch": 4.036390183635649,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034661032096141615,
      "loss": 4.883,
      "mean_token_accuracy": 0.22512781769037246,
      "num_tokens": 80224780.0,
      "step": 35940
    },
    {
      "entropy": 5.420991992950439,
      "epoch": 4.036951760543606,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003465722588058497,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21716538518667222,
      "num_tokens": 80235399.0,
      "step": 35945
    },
    {
      "entropy": 5.323662376403808,
      "epoch": 4.037513337451564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034653419437139475,
      "loss": 4.8772,
      "mean_token_accuracy": 0.2288653701543808,
      "num_tokens": 80245019.0,
      "step": 35950
    },
    {
      "entropy": 5.366043996810913,
      "epoch": 4.038074914359521,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003464961276592634,
      "loss": 4.9885,
      "mean_token_accuracy": 0.22577728927135468,
      "num_tokens": 80256650.0,
      "step": 35955
    },
    {
      "entropy": 5.305973386764526,
      "epoch": 4.0386364912674795,
      "grad_norm": 0.875,
      "learning_rate": 0.00034645805867066756,
      "loss": 4.8736,
      "mean_token_accuracy": 0.2206096976995468,
      "num_tokens": 80267463.0,
      "step": 35960
    },
    {
      "entropy": 5.355551528930664,
      "epoch": 4.039198068175437,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034641998740681933,
      "loss": 4.9515,
      "mean_token_accuracy": 0.22348917573690413,
      "num_tokens": 80279751.0,
      "step": 35965
    },
    {
      "entropy": 5.328068590164184,
      "epoch": 4.039759645083394,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034638191386893115,
      "loss": 4.8893,
      "mean_token_accuracy": 0.23066311180591584,
      "num_tokens": 80291249.0,
      "step": 35970
    },
    {
      "entropy": 5.332080268859864,
      "epoch": 4.040321221991352,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034634383805821504,
      "loss": 4.8588,
      "mean_token_accuracy": 0.23518844097852706,
      "num_tokens": 80302082.0,
      "step": 35975
    },
    {
      "entropy": 5.231941890716553,
      "epoch": 4.040882798899309,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034630575997588353,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23936370760202408,
      "num_tokens": 80312470.0,
      "step": 35980
    },
    {
      "entropy": 5.297620820999145,
      "epoch": 4.0414443758072665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000346267679623149,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23594695776700975,
      "num_tokens": 80322562.0,
      "step": 35985
    },
    {
      "entropy": 5.24401273727417,
      "epoch": 4.042005952715225,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034622959700122383,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22688017934560775,
      "num_tokens": 80333434.0,
      "step": 35990
    },
    {
      "entropy": 5.226737689971924,
      "epoch": 4.042567529623182,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003461915121113206,
      "loss": 4.8267,
      "mean_token_accuracy": 0.23215852975845336,
      "num_tokens": 80345028.0,
      "step": 35995
    },
    {
      "entropy": 5.263515949249268,
      "epoch": 4.04312910653114,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003461534249546521,
      "loss": 4.8303,
      "mean_token_accuracy": 0.2303904786705971,
      "num_tokens": 80356536.0,
      "step": 36000
    },
    {
      "epoch": 4.04312910653114,
      "eval_entropy": 5.10593157550063,
      "eval_loss": 5.10050106048584,
      "eval_mean_token_accuracy": 0.22143534857079267,
      "eval_num_tokens": 80356536.0,
      "eval_runtime": 23.9558,
      "eval_samples_per_second": 1294.59,
      "eval_steps_per_second": 161.839,
      "step": 36000
    },
    {
      "entropy": 5.359178161621093,
      "epoch": 4.043690683439097,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000346115335532431,
      "loss": 4.9782,
      "mean_token_accuracy": 0.22344818264245986,
      "num_tokens": 80366874.0,
      "step": 36005
    },
    {
      "entropy": 5.41416745185852,
      "epoch": 4.044252260347054,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034607724384587006,
      "loss": 4.8933,
      "mean_token_accuracy": 0.21368961185216903,
      "num_tokens": 80378015.0,
      "step": 36010
    },
    {
      "entropy": 5.2436353206634525,
      "epoch": 4.044813837255012,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034603914989618197,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22186829894781113,
      "num_tokens": 80388423.0,
      "step": 36015
    },
    {
      "entropy": 5.303881978988647,
      "epoch": 4.04537541416297,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003460010536845799,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22972507774829865,
      "num_tokens": 80399985.0,
      "step": 36020
    },
    {
      "entropy": 5.314314985275269,
      "epoch": 4.045936991070927,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003459629552122767,
      "loss": 4.9082,
      "mean_token_accuracy": 0.2273220419883728,
      "num_tokens": 80410653.0,
      "step": 36025
    },
    {
      "entropy": 5.313387155532837,
      "epoch": 4.046498567978885,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003459248544804855,
      "loss": 4.912,
      "mean_token_accuracy": 0.2211352825164795,
      "num_tokens": 80421944.0,
      "step": 36030
    },
    {
      "entropy": 5.4333281993865965,
      "epoch": 4.047060144886842,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003458867514904194,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2197570249438286,
      "num_tokens": 80433509.0,
      "step": 36035
    },
    {
      "entropy": 5.410898447036743,
      "epoch": 4.047621721794799,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003458486462432916,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2215388908982277,
      "num_tokens": 80444148.0,
      "step": 36040
    },
    {
      "entropy": 5.351002216339111,
      "epoch": 4.0481832987027575,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034581053874031554,
      "loss": 4.952,
      "mean_token_accuracy": 0.22047373354434968,
      "num_tokens": 80455495.0,
      "step": 36045
    },
    {
      "entropy": 5.318565368652344,
      "epoch": 4.048744875610715,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003457724289827044,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2189271256327629,
      "num_tokens": 80465846.0,
      "step": 36050
    },
    {
      "entropy": 5.398042488098144,
      "epoch": 4.049306452518673,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003457343169716716,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21851759701967238,
      "num_tokens": 80477172.0,
      "step": 36055
    },
    {
      "entropy": 5.335282278060913,
      "epoch": 4.04986802942663,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034569620270843087,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2303847700357437,
      "num_tokens": 80487174.0,
      "step": 36060
    },
    {
      "entropy": 5.3667747497558596,
      "epoch": 4.050429606334587,
      "grad_norm": 1.0,
      "learning_rate": 0.00034565808619419554,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22330591678619385,
      "num_tokens": 80497748.0,
      "step": 36065
    },
    {
      "entropy": 5.260837459564209,
      "epoch": 4.050991183242545,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003456199674301793,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22931792140007018,
      "num_tokens": 80508713.0,
      "step": 36070
    },
    {
      "entropy": 5.3427032947540285,
      "epoch": 4.051552760150503,
      "grad_norm": 1.046875,
      "learning_rate": 0.000345581846417596,
      "loss": 4.9,
      "mean_token_accuracy": 0.22784536927938462,
      "num_tokens": 80518813.0,
      "step": 36075
    },
    {
      "entropy": 5.246268224716187,
      "epoch": 4.05211433705846,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034554372315765927,
      "loss": 4.8835,
      "mean_token_accuracy": 0.22304878383874893,
      "num_tokens": 80529541.0,
      "step": 36080
    },
    {
      "entropy": 5.346269178390503,
      "epoch": 4.052675913966418,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003455055976515831,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22363602370023727,
      "num_tokens": 80541942.0,
      "step": 36085
    },
    {
      "entropy": 5.414978122711181,
      "epoch": 4.053237490874375,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034546746990058125,
      "loss": 4.9659,
      "mean_token_accuracy": 0.2186441108584404,
      "num_tokens": 80552477.0,
      "step": 36090
    },
    {
      "entropy": 5.337516212463379,
      "epoch": 4.053799067782332,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034542933990586795,
      "loss": 4.974,
      "mean_token_accuracy": 0.21640294343233107,
      "num_tokens": 80563976.0,
      "step": 36095
    },
    {
      "entropy": 5.344397830963135,
      "epoch": 4.0543606446902904,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034539120766865713,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21911509037017823,
      "num_tokens": 80575397.0,
      "step": 36100
    },
    {
      "entropy": 5.297573137283325,
      "epoch": 4.054922221598248,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003453530731901628,
      "loss": 4.9168,
      "mean_token_accuracy": 0.2219690814614296,
      "num_tokens": 80587500.0,
      "step": 36105
    },
    {
      "entropy": 5.3101075172424315,
      "epoch": 4.055483798506206,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003453149364715995,
      "loss": 4.8325,
      "mean_token_accuracy": 0.22964439988136293,
      "num_tokens": 80598216.0,
      "step": 36110
    },
    {
      "entropy": 5.299663543701172,
      "epoch": 4.056045375414163,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034527679751418126,
      "loss": 4.8865,
      "mean_token_accuracy": 0.223801825940609,
      "num_tokens": 80609759.0,
      "step": 36115
    },
    {
      "entropy": 5.376327848434448,
      "epoch": 4.05660695232212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003452386563191226,
      "loss": 4.8982,
      "mean_token_accuracy": 0.22585987001657487,
      "num_tokens": 80620587.0,
      "step": 36120
    },
    {
      "entropy": 5.341989755630493,
      "epoch": 4.057168529230078,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034520051288763784,
      "loss": 5.017,
      "mean_token_accuracy": 0.21274927407503127,
      "num_tokens": 80631882.0,
      "step": 36125
    },
    {
      "entropy": 5.368930006027222,
      "epoch": 4.0577301061380355,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003451623672209415,
      "loss": 4.976,
      "mean_token_accuracy": 0.22589125186204911,
      "num_tokens": 80644878.0,
      "step": 36130
    },
    {
      "entropy": 5.246311521530151,
      "epoch": 4.058291683045993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034512421932024825,
      "loss": 4.7666,
      "mean_token_accuracy": 0.233259217441082,
      "num_tokens": 80655641.0,
      "step": 36135
    },
    {
      "entropy": 5.312887191772461,
      "epoch": 4.058853259953951,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003450860691867726,
      "loss": 4.964,
      "mean_token_accuracy": 0.21721645891666413,
      "num_tokens": 80666883.0,
      "step": 36140
    },
    {
      "entropy": 5.331695795059204,
      "epoch": 4.059414836861908,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034504791682172934,
      "loss": 4.9007,
      "mean_token_accuracy": 0.22148784548044204,
      "num_tokens": 80678082.0,
      "step": 36145
    },
    {
      "entropy": 5.343256092071533,
      "epoch": 4.059976413769866,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003450097622263333,
      "loss": 4.9693,
      "mean_token_accuracy": 0.21765907108783722,
      "num_tokens": 80691553.0,
      "step": 36150
    },
    {
      "entropy": 5.250389528274536,
      "epoch": 4.060537990677823,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034497160540179924,
      "loss": 4.7988,
      "mean_token_accuracy": 0.2354423075914383,
      "num_tokens": 80702353.0,
      "step": 36155
    },
    {
      "entropy": 5.247003793716431,
      "epoch": 4.061099567585781,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034493344634934205,
      "loss": 4.8906,
      "mean_token_accuracy": 0.22810729295015336,
      "num_tokens": 80712091.0,
      "step": 36160
    },
    {
      "entropy": 5.383491659164429,
      "epoch": 4.061661144493739,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003448952850701769,
      "loss": 5.0248,
      "mean_token_accuracy": 0.22045061737298965,
      "num_tokens": 80723283.0,
      "step": 36165
    },
    {
      "entropy": 5.40281400680542,
      "epoch": 4.062222721401696,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034485712156551873,
      "loss": 4.8616,
      "mean_token_accuracy": 0.23471304029226303,
      "num_tokens": 80733434.0,
      "step": 36170
    },
    {
      "entropy": 5.351315307617187,
      "epoch": 4.062784298309653,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034481895583658267,
      "loss": 5.0156,
      "mean_token_accuracy": 0.2111069977283478,
      "num_tokens": 80743236.0,
      "step": 36175
    },
    {
      "entropy": 5.356589078903198,
      "epoch": 4.063345875217611,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034478078788458406,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22012976109981536,
      "num_tokens": 80754821.0,
      "step": 36180
    },
    {
      "entropy": 5.26940279006958,
      "epoch": 4.0639074521255685,
      "grad_norm": 0.96875,
      "learning_rate": 0.000344742617710738,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22598658949136735,
      "num_tokens": 80765755.0,
      "step": 36185
    },
    {
      "entropy": 5.4358312606811525,
      "epoch": 4.064469029033527,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034470444531626003,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2127585232257843,
      "num_tokens": 80776062.0,
      "step": 36190
    },
    {
      "entropy": 5.392159938812256,
      "epoch": 4.065030605941484,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034466627070236545,
      "loss": 4.9136,
      "mean_token_accuracy": 0.21950702965259553,
      "num_tokens": 80787005.0,
      "step": 36195
    },
    {
      "entropy": 5.302799129486084,
      "epoch": 4.065592182849441,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003446280938702698,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22878555804491044,
      "num_tokens": 80798949.0,
      "step": 36200
    },
    {
      "entropy": 5.352507209777832,
      "epoch": 4.066153759757399,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034458991482118865,
      "loss": 4.883,
      "mean_token_accuracy": 0.22353299260139464,
      "num_tokens": 80809087.0,
      "step": 36205
    },
    {
      "entropy": 5.284424209594727,
      "epoch": 4.066715336665356,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003445517335563376,
      "loss": 4.887,
      "mean_token_accuracy": 0.2294287234544754,
      "num_tokens": 80819764.0,
      "step": 36210
    },
    {
      "entropy": 5.273963499069214,
      "epoch": 4.067276913573314,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034451355007693237,
      "loss": 4.8802,
      "mean_token_accuracy": 0.222270230948925,
      "num_tokens": 80830123.0,
      "step": 36215
    },
    {
      "entropy": 5.310305118560791,
      "epoch": 4.067838490481272,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003444753643841888,
      "loss": 4.8014,
      "mean_token_accuracy": 0.228645958006382,
      "num_tokens": 80841222.0,
      "step": 36220
    },
    {
      "entropy": 5.3834411144256595,
      "epoch": 4.068400067389229,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034443717647932265,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2211059510707855,
      "num_tokens": 80854432.0,
      "step": 36225
    },
    {
      "entropy": 5.35668978691101,
      "epoch": 4.068961644297186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003443989863635498,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21854736655950546,
      "num_tokens": 80864915.0,
      "step": 36230
    },
    {
      "entropy": 5.2936736106872555,
      "epoch": 4.069523221205144,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00034436079403808647,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2189372643828392,
      "num_tokens": 80877239.0,
      "step": 36235
    },
    {
      "entropy": 5.488775062561035,
      "epoch": 4.070084798113101,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00034432259950414846,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20623173266649247,
      "num_tokens": 80890618.0,
      "step": 36240
    },
    {
      "entropy": 5.439596223831177,
      "epoch": 4.0706463750210595,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000344284402762952,
      "loss": 4.943,
      "mean_token_accuracy": 0.22089067697525025,
      "num_tokens": 80902445.0,
      "step": 36245
    },
    {
      "entropy": 5.348625040054321,
      "epoch": 4.071207951929017,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003442462038157132,
      "loss": 4.95,
      "mean_token_accuracy": 0.21522913873195648,
      "num_tokens": 80913519.0,
      "step": 36250
    },
    {
      "entropy": 5.31230936050415,
      "epoch": 4.071769528836974,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003442080026636485,
      "loss": 4.9238,
      "mean_token_accuracy": 0.215563540160656,
      "num_tokens": 80924800.0,
      "step": 36255
    },
    {
      "entropy": 5.352876663208008,
      "epoch": 4.072331105744932,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003441697993079742,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22767739742994308,
      "num_tokens": 80935812.0,
      "step": 36260
    },
    {
      "entropy": 5.3814905166625975,
      "epoch": 4.072892682652889,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003441315937499066,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22601717561483384,
      "num_tokens": 80946456.0,
      "step": 36265
    },
    {
      "entropy": 5.346848106384277,
      "epoch": 4.0734542595608465,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034409338599066224,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22367979139089583,
      "num_tokens": 80957983.0,
      "step": 36270
    },
    {
      "entropy": 5.368098402023316,
      "epoch": 4.074015836468805,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003440551760314577,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21383372694253922,
      "num_tokens": 80969093.0,
      "step": 36275
    },
    {
      "entropy": 5.35685019493103,
      "epoch": 4.074577413376762,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034401696387350954,
      "loss": 4.898,
      "mean_token_accuracy": 0.23036142587661743,
      "num_tokens": 80980511.0,
      "step": 36280
    },
    {
      "entropy": 5.28838963508606,
      "epoch": 4.075138990284719,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003439787495180345,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22649962604045867,
      "num_tokens": 80992437.0,
      "step": 36285
    },
    {
      "entropy": 5.3275219917297365,
      "epoch": 4.075700567192677,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034394053296624936,
      "loss": 4.8837,
      "mean_token_accuracy": 0.22668827027082444,
      "num_tokens": 81003105.0,
      "step": 36290
    },
    {
      "entropy": 5.314283943176269,
      "epoch": 4.076262144100634,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003439023142193709,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22405313849449157,
      "num_tokens": 81015141.0,
      "step": 36295
    },
    {
      "entropy": 5.253846645355225,
      "epoch": 4.0768237210085925,
      "grad_norm": 1.109375,
      "learning_rate": 0.000343864093278616,
      "loss": 4.893,
      "mean_token_accuracy": 0.22387565225362777,
      "num_tokens": 81025875.0,
      "step": 36300
    },
    {
      "entropy": 5.333708953857422,
      "epoch": 4.07738529791655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003438258701452017,
      "loss": 4.9157,
      "mean_token_accuracy": 0.22685809582471847,
      "num_tokens": 81036963.0,
      "step": 36305
    },
    {
      "entropy": 5.312244701385498,
      "epoch": 4.077946874824507,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003437876448203449,
      "loss": 4.8574,
      "mean_token_accuracy": 0.22804497927427292,
      "num_tokens": 81048660.0,
      "step": 36310
    },
    {
      "entropy": 5.258388948440552,
      "epoch": 4.078508451732465,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003437494173052629,
      "loss": 4.9057,
      "mean_token_accuracy": 0.21852595061063768,
      "num_tokens": 81060071.0,
      "step": 36315
    },
    {
      "entropy": 5.312449836730957,
      "epoch": 4.079070028640422,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003437111876011728,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22572489976882934,
      "num_tokens": 81071126.0,
      "step": 36320
    },
    {
      "entropy": 5.26862325668335,
      "epoch": 4.0796316055483794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034367295570929176,
      "loss": 4.8719,
      "mean_token_accuracy": 0.22447004169225693,
      "num_tokens": 81081349.0,
      "step": 36325
    },
    {
      "entropy": 5.246787881851196,
      "epoch": 4.0801931824563376,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003436347216308373,
      "loss": 4.7825,
      "mean_token_accuracy": 0.22882500737905503,
      "num_tokens": 81092804.0,
      "step": 36330
    },
    {
      "entropy": 5.263952541351318,
      "epoch": 4.080754759364295,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003435964853670265,
      "loss": 4.867,
      "mean_token_accuracy": 0.23249226361513137,
      "num_tokens": 81103539.0,
      "step": 36335
    },
    {
      "entropy": 5.3492615699768065,
      "epoch": 4.081316336272253,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00034355824691907714,
      "loss": 4.9724,
      "mean_token_accuracy": 0.22277533113956452,
      "num_tokens": 81115586.0,
      "step": 36340
    },
    {
      "entropy": 5.294626474380493,
      "epoch": 4.08187791318021,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034352000628820655,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22429246753454207,
      "num_tokens": 81125354.0,
      "step": 36345
    },
    {
      "entropy": 5.3012491226196286,
      "epoch": 4.082439490088167,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003434817634756324,
      "loss": 4.8326,
      "mean_token_accuracy": 0.23545102030038834,
      "num_tokens": 81136890.0,
      "step": 36350
    },
    {
      "entropy": 5.30880765914917,
      "epoch": 4.083001066996125,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003434435184825724,
      "loss": 4.9276,
      "mean_token_accuracy": 0.22578466087579727,
      "num_tokens": 81147539.0,
      "step": 36355
    },
    {
      "entropy": 5.296538972854615,
      "epoch": 4.083562643904083,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034340527131024415,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22424944937229158,
      "num_tokens": 81160529.0,
      "step": 36360
    },
    {
      "entropy": 5.399967908859253,
      "epoch": 4.08412422081204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003433670219598656,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22058828622102739,
      "num_tokens": 81171067.0,
      "step": 36365
    },
    {
      "entropy": 5.343110656738281,
      "epoch": 4.084685797719998,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003433287704326545,
      "loss": 4.8751,
      "mean_token_accuracy": 0.22623569816350936,
      "num_tokens": 81182383.0,
      "step": 36370
    },
    {
      "entropy": 5.2593122005462645,
      "epoch": 4.085247374627955,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003432905167298288,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22965872436761856,
      "num_tokens": 81192939.0,
      "step": 36375
    },
    {
      "entropy": 5.302305889129639,
      "epoch": 4.085808951535913,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034325226085260666,
      "loss": 4.879,
      "mean_token_accuracy": 0.22699808180332184,
      "num_tokens": 81204110.0,
      "step": 36380
    },
    {
      "entropy": 5.300213766098023,
      "epoch": 4.0863705284438705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000343214002802206,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22306997776031495,
      "num_tokens": 81214519.0,
      "step": 36385
    },
    {
      "entropy": 5.226606464385986,
      "epoch": 4.086932105351828,
      "grad_norm": 1.0,
      "learning_rate": 0.00034317574257984514,
      "loss": 4.778,
      "mean_token_accuracy": 0.23381907045841216,
      "num_tokens": 81225427.0,
      "step": 36390
    },
    {
      "entropy": 5.328241586685181,
      "epoch": 4.087493682259786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000343137480186742,
      "loss": 4.9381,
      "mean_token_accuracy": 0.21319423019886016,
      "num_tokens": 81236450.0,
      "step": 36395
    },
    {
      "entropy": 5.364252376556396,
      "epoch": 4.088055259167743,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034309921562411527,
      "loss": 4.9049,
      "mean_token_accuracy": 0.22523159384727479,
      "num_tokens": 81248302.0,
      "step": 36400
    },
    {
      "entropy": 5.2345983505249025,
      "epoch": 4.0886168360757,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034306094889318295,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23168878853321076,
      "num_tokens": 81258332.0,
      "step": 36405
    },
    {
      "entropy": 5.2078385829925535,
      "epoch": 4.089178412983658,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003430226799951637,
      "loss": 4.865,
      "mean_token_accuracy": 0.22297257483005523,
      "num_tokens": 81270009.0,
      "step": 36410
    },
    {
      "entropy": 5.343054533004761,
      "epoch": 4.089739989891616,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034298440893127584,
      "loss": 4.9598,
      "mean_token_accuracy": 0.22669940292835236,
      "num_tokens": 81281025.0,
      "step": 36415
    },
    {
      "entropy": 5.319711780548095,
      "epoch": 4.090301566799573,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034294613570273817,
      "loss": 4.9278,
      "mean_token_accuracy": 0.23163323998451232,
      "num_tokens": 81292932.0,
      "step": 36420
    },
    {
      "entropy": 5.2806463718414305,
      "epoch": 4.090863143707531,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034290786031076917,
      "loss": 4.8409,
      "mean_token_accuracy": 0.22883131504058837,
      "num_tokens": 81303007.0,
      "step": 36425
    },
    {
      "entropy": 5.314305305480957,
      "epoch": 4.091424720615488,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003428695827565875,
      "loss": 4.902,
      "mean_token_accuracy": 0.22662333399057388,
      "num_tokens": 81315758.0,
      "step": 36430
    },
    {
      "entropy": 5.287782907485962,
      "epoch": 4.091986297523446,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034283130304141204,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22456922829151155,
      "num_tokens": 81327148.0,
      "step": 36435
    },
    {
      "entropy": 5.415901756286621,
      "epoch": 4.092547874431403,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003427930211664615,
      "loss": 5.0426,
      "mean_token_accuracy": 0.2215466320514679,
      "num_tokens": 81338560.0,
      "step": 36440
    },
    {
      "entropy": 5.314035558700562,
      "epoch": 4.093109451339361,
      "grad_norm": 1.046875,
      "learning_rate": 0.000342754737132955,
      "loss": 4.8824,
      "mean_token_accuracy": 0.2291946679353714,
      "num_tokens": 81349998.0,
      "step": 36445
    },
    {
      "entropy": 5.288207197189331,
      "epoch": 4.093671028247319,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034271645094211125,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22445319443941117,
      "num_tokens": 81360041.0,
      "step": 36450
    },
    {
      "entropy": 5.260379123687744,
      "epoch": 4.094232605155276,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003426781625951495,
      "loss": 4.8664,
      "mean_token_accuracy": 0.2275041773915291,
      "num_tokens": 81370256.0,
      "step": 36455
    },
    {
      "entropy": 5.2805126190185545,
      "epoch": 4.094794182063233,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034263987209328886,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22611627727746964,
      "num_tokens": 81382365.0,
      "step": 36460
    },
    {
      "entropy": 5.293807554244995,
      "epoch": 4.095355758971191,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003426015794377483,
      "loss": 4.8653,
      "mean_token_accuracy": 0.22936853766441345,
      "num_tokens": 81393147.0,
      "step": 36465
    },
    {
      "entropy": 5.306332111358643,
      "epoch": 4.0959173358791485,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003425632846297474,
      "loss": 4.7954,
      "mean_token_accuracy": 0.22976450324058534,
      "num_tokens": 81404464.0,
      "step": 36470
    },
    {
      "entropy": 5.3131749629974365,
      "epoch": 4.096478912787107,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034252498767050525,
      "loss": 4.8645,
      "mean_token_accuracy": 0.22589686065912246,
      "num_tokens": 81415191.0,
      "step": 36475
    },
    {
      "entropy": 5.310141372680664,
      "epoch": 4.097040489695064,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003424866885612413,
      "loss": 4.917,
      "mean_token_accuracy": 0.22786842286586761,
      "num_tokens": 81426258.0,
      "step": 36480
    },
    {
      "entropy": 5.37461838722229,
      "epoch": 4.097602066603021,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000342448387303175,
      "loss": 4.9081,
      "mean_token_accuracy": 0.2314464792609215,
      "num_tokens": 81437763.0,
      "step": 36485
    },
    {
      "entropy": 5.32384614944458,
      "epoch": 4.098163643510979,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034241008389752594,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21918401420116423,
      "num_tokens": 81451012.0,
      "step": 36490
    },
    {
      "entropy": 5.353970670700074,
      "epoch": 4.098725220418936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003423717783455136,
      "loss": 4.9663,
      "mean_token_accuracy": 0.2188495010137558,
      "num_tokens": 81462310.0,
      "step": 36495
    },
    {
      "entropy": 5.293469953536987,
      "epoch": 4.099286797326894,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034233347064835775,
      "loss": 4.892,
      "mean_token_accuracy": 0.22690458595752716,
      "num_tokens": 81473423.0,
      "step": 36500
    },
    {
      "entropy": 5.332803726196289,
      "epoch": 4.099848374234852,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003422951608072781,
      "loss": 4.902,
      "mean_token_accuracy": 0.22693066895008088,
      "num_tokens": 81485019.0,
      "step": 36505
    },
    {
      "entropy": 5.327925777435302,
      "epoch": 4.100409951142809,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003422568488234944,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22172844558954238,
      "num_tokens": 81495724.0,
      "step": 36510
    },
    {
      "entropy": 5.245721769332886,
      "epoch": 4.100971528050766,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034221853469822654,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22440870106220245,
      "num_tokens": 81506192.0,
      "step": 36515
    },
    {
      "entropy": 5.303523683547974,
      "epoch": 4.101533104958724,
      "grad_norm": 1.0,
      "learning_rate": 0.00034218021843269444,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22184186428785324,
      "num_tokens": 81518783.0,
      "step": 36520
    },
    {
      "entropy": 5.303485345840454,
      "epoch": 4.1020946818666815,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003421419000281182,
      "loss": 4.8446,
      "mean_token_accuracy": 0.22213964760303498,
      "num_tokens": 81530426.0,
      "step": 36525
    },
    {
      "entropy": 5.3957287788391115,
      "epoch": 4.10265625877464,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003421035794857178,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21618830561637878,
      "num_tokens": 81540479.0,
      "step": 36530
    },
    {
      "entropy": 5.314896869659424,
      "epoch": 4.103217835682597,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003420652568067134,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22870832830667495,
      "num_tokens": 81550618.0,
      "step": 36535
    },
    {
      "entropy": 5.347701454162598,
      "epoch": 4.103779412590554,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034202693199232516,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2286326467990875,
      "num_tokens": 81561529.0,
      "step": 36540
    },
    {
      "entropy": 5.3070902824401855,
      "epoch": 4.104340989498512,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034198860504377343,
      "loss": 4.8103,
      "mean_token_accuracy": 0.22830644696950914,
      "num_tokens": 81571660.0,
      "step": 36545
    },
    {
      "entropy": 5.363492345809936,
      "epoch": 4.104902566406469,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034195027596227857,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2184915229678154,
      "num_tokens": 81581738.0,
      "step": 36550
    },
    {
      "entropy": 5.324359941482544,
      "epoch": 4.1054641433144266,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003419119447490609,
      "loss": 4.959,
      "mean_token_accuracy": 0.22085279524326323,
      "num_tokens": 81593028.0,
      "step": 36555
    },
    {
      "entropy": 5.328190422058105,
      "epoch": 4.106025720222385,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000341873611405341,
      "loss": 4.8346,
      "mean_token_accuracy": 0.22565679699182511,
      "num_tokens": 81605353.0,
      "step": 36560
    },
    {
      "entropy": 5.29201226234436,
      "epoch": 4.106587297130342,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003418352759323394,
      "loss": 4.8565,
      "mean_token_accuracy": 0.2286763459444046,
      "num_tokens": 81616601.0,
      "step": 36565
    },
    {
      "entropy": 5.324309349060059,
      "epoch": 4.1071488740383,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034179693833127667,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22289418578147888,
      "num_tokens": 81628064.0,
      "step": 36570
    },
    {
      "entropy": 5.3169636726379395,
      "epoch": 4.107710450946257,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034175859860337347,
      "loss": 4.8935,
      "mean_token_accuracy": 0.2169930726289749,
      "num_tokens": 81639844.0,
      "step": 36575
    },
    {
      "entropy": 5.3732442378997805,
      "epoch": 4.108272027854214,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034172025674985076,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22566508799791335,
      "num_tokens": 81650978.0,
      "step": 36580
    },
    {
      "entropy": 5.239879512786866,
      "epoch": 4.1088336047621725,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034168191277192904,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23155894130468369,
      "num_tokens": 81661862.0,
      "step": 36585
    },
    {
      "entropy": 5.299708509445191,
      "epoch": 4.10939518167013,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034164356667082933,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23202944546937943,
      "num_tokens": 81673955.0,
      "step": 36590
    },
    {
      "entropy": 5.234638833999634,
      "epoch": 4.109956758578087,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003416052184477727,
      "loss": 4.7441,
      "mean_token_accuracy": 0.23370224237442017,
      "num_tokens": 81683610.0,
      "step": 36595
    },
    {
      "entropy": 5.334441041946411,
      "epoch": 4.110518335486045,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003415668681039801,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22174402177333832,
      "num_tokens": 81694295.0,
      "step": 36600
    },
    {
      "entropy": 5.327102899551392,
      "epoch": 4.111079912394002,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034152851564067254,
      "loss": 4.9347,
      "mean_token_accuracy": 0.223934568464756,
      "num_tokens": 81705090.0,
      "step": 36605
    },
    {
      "entropy": 5.28317699432373,
      "epoch": 4.1116414893019595,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003414901610590712,
      "loss": 4.8618,
      "mean_token_accuracy": 0.23082437813282014,
      "num_tokens": 81716891.0,
      "step": 36610
    },
    {
      "entropy": 5.360436916351318,
      "epoch": 4.112203066209918,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003414518043603974,
      "loss": 4.9908,
      "mean_token_accuracy": 0.2186895191669464,
      "num_tokens": 81728781.0,
      "step": 36615
    },
    {
      "entropy": 5.427006626129151,
      "epoch": 4.112764643117875,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034141344554587243,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2224336251616478,
      "num_tokens": 81738933.0,
      "step": 36620
    },
    {
      "entropy": 5.314230442047119,
      "epoch": 4.113326220025833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003413750846167175,
      "loss": 4.9138,
      "mean_token_accuracy": 0.224728325009346,
      "num_tokens": 81750241.0,
      "step": 36625
    },
    {
      "entropy": 5.261791753768921,
      "epoch": 4.11388779693379,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003413367215741542,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22759616672992705,
      "num_tokens": 81762100.0,
      "step": 36630
    },
    {
      "entropy": 5.326925325393677,
      "epoch": 4.114449373841747,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034129835641940394,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22471993863582612,
      "num_tokens": 81774628.0,
      "step": 36635
    },
    {
      "entropy": 5.426640510559082,
      "epoch": 4.1150109507497055,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00034125998915368827,
      "loss": 4.9595,
      "mean_token_accuracy": 0.2216418504714966,
      "num_tokens": 81787154.0,
      "step": 36640
    },
    {
      "entropy": 5.265170383453369,
      "epoch": 4.115572527657663,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034122161977822885,
      "loss": 4.885,
      "mean_token_accuracy": 0.22436075806617736,
      "num_tokens": 81797815.0,
      "step": 36645
    },
    {
      "entropy": 5.254416608810425,
      "epoch": 4.11613410456562,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003411832482942474,
      "loss": 4.8959,
      "mean_token_accuracy": 0.2255213662981987,
      "num_tokens": 81809939.0,
      "step": 36650
    },
    {
      "entropy": 5.262873697280884,
      "epoch": 4.116695681473578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034114487470296575,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22544188648462296,
      "num_tokens": 81820775.0,
      "step": 36655
    },
    {
      "entropy": 5.334648370742798,
      "epoch": 4.117257258381535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034110649900560554,
      "loss": 4.8954,
      "mean_token_accuracy": 0.21859526336193086,
      "num_tokens": 81831501.0,
      "step": 36660
    },
    {
      "entropy": 5.322180509567261,
      "epoch": 4.117818835289493,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003410681212033887,
      "loss": 4.9531,
      "mean_token_accuracy": 0.22675986886024474,
      "num_tokens": 81842052.0,
      "step": 36665
    },
    {
      "entropy": 5.376874828338623,
      "epoch": 4.1183804121974505,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034102974129753726,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22156075686216353,
      "num_tokens": 81852730.0,
      "step": 36670
    },
    {
      "entropy": 5.248844575881958,
      "epoch": 4.118941989105408,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003409913592892733,
      "loss": 4.8908,
      "mean_token_accuracy": 0.21988571882247926,
      "num_tokens": 81864436.0,
      "step": 36675
    },
    {
      "entropy": 5.234925365447998,
      "epoch": 4.119503566013366,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003409529751798189,
      "loss": 4.7745,
      "mean_token_accuracy": 0.24416860193014145,
      "num_tokens": 81875284.0,
      "step": 36680
    },
    {
      "entropy": 5.2616828918457035,
      "epoch": 4.120065142921323,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003409145889703961,
      "loss": 4.8229,
      "mean_token_accuracy": 0.2349325954914093,
      "num_tokens": 81885571.0,
      "step": 36685
    },
    {
      "entropy": 5.247549057006836,
      "epoch": 4.12062671982928,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034087620066222733,
      "loss": 4.8491,
      "mean_token_accuracy": 0.22914725691080093,
      "num_tokens": 81897190.0,
      "step": 36690
    },
    {
      "entropy": 5.421279191970825,
      "epoch": 4.121188296737238,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003408378102565347,
      "loss": 5.0376,
      "mean_token_accuracy": 0.21510691791772843,
      "num_tokens": 81909811.0,
      "step": 36695
    },
    {
      "entropy": 5.314928579330444,
      "epoch": 4.121749873645196,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003407994177545407,
      "loss": 4.8585,
      "mean_token_accuracy": 0.22289398312568665,
      "num_tokens": 81920528.0,
      "step": 36700
    },
    {
      "entropy": 5.272383546829223,
      "epoch": 4.122311450553153,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003407610231574677,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2313857287168503,
      "num_tokens": 81931994.0,
      "step": 36705
    },
    {
      "entropy": 5.366690683364868,
      "epoch": 4.122873027461111,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003407226264665382,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2197750225663185,
      "num_tokens": 81943641.0,
      "step": 36710
    },
    {
      "entropy": 5.290005731582641,
      "epoch": 4.123434604369068,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034068422768297486,
      "loss": 4.8706,
      "mean_token_accuracy": 0.23333559483289718,
      "num_tokens": 81955516.0,
      "step": 36715
    },
    {
      "entropy": 5.387358856201172,
      "epoch": 4.123996181277026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003406458268080001,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22269238531589508,
      "num_tokens": 81966019.0,
      "step": 36720
    },
    {
      "entropy": 5.3177131652832035,
      "epoch": 4.1245577581849835,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034060742384283695,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2163916662335396,
      "num_tokens": 81976664.0,
      "step": 36725
    },
    {
      "entropy": 5.294422817230225,
      "epoch": 4.125119335092941,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003405690187887078,
      "loss": 4.8525,
      "mean_token_accuracy": 0.2313166603446007,
      "num_tokens": 81987572.0,
      "step": 36730
    },
    {
      "entropy": 5.380994510650635,
      "epoch": 4.125680912000899,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034053061164683574,
      "loss": 5.0079,
      "mean_token_accuracy": 0.22195477783679962,
      "num_tokens": 81998943.0,
      "step": 36735
    },
    {
      "entropy": 5.32921199798584,
      "epoch": 4.126242488908856,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034049220241844357,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2204698458313942,
      "num_tokens": 82010723.0,
      "step": 36740
    },
    {
      "entropy": 5.277633714675903,
      "epoch": 4.126804065816813,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003404537911047543,
      "loss": 4.9183,
      "mean_token_accuracy": 0.2217444211244583,
      "num_tokens": 82020888.0,
      "step": 36745
    },
    {
      "entropy": 5.330263090133667,
      "epoch": 4.127365642724771,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034041537770699095,
      "loss": 4.9156,
      "mean_token_accuracy": 0.22030480653047563,
      "num_tokens": 82031985.0,
      "step": 36750
    },
    {
      "entropy": 5.370676183700562,
      "epoch": 4.127927219632729,
      "grad_norm": 1.0,
      "learning_rate": 0.0003403769622263765,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22822305113077163,
      "num_tokens": 82042161.0,
      "step": 36755
    },
    {
      "entropy": 5.472585773468017,
      "epoch": 4.128488796540687,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034033854466413425,
      "loss": 5.0559,
      "mean_token_accuracy": 0.2143716648221016,
      "num_tokens": 82054587.0,
      "step": 36760
    },
    {
      "entropy": 5.332214736938477,
      "epoch": 4.129050373448644,
      "grad_norm": 1.0,
      "learning_rate": 0.0003403001250214875,
      "loss": 4.9451,
      "mean_token_accuracy": 0.21777959167957306,
      "num_tokens": 82064850.0,
      "step": 36765
    },
    {
      "entropy": 5.323500251770019,
      "epoch": 4.129611950356601,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003402617032996593,
      "loss": 4.8453,
      "mean_token_accuracy": 0.22687113136053086,
      "num_tokens": 82074963.0,
      "step": 36770
    },
    {
      "entropy": 5.357334280014038,
      "epoch": 4.130173527264559,
      "grad_norm": 1.0,
      "learning_rate": 0.0003402232794998733,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21432416141033173,
      "num_tokens": 82086878.0,
      "step": 36775
    },
    {
      "entropy": 5.262367153167725,
      "epoch": 4.130735104172516,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034018485362335266,
      "loss": 4.8234,
      "mean_token_accuracy": 0.23222053050994873,
      "num_tokens": 82096621.0,
      "step": 36780
    },
    {
      "entropy": 5.2688805103302006,
      "epoch": 4.131296681080474,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003401464256713211,
      "loss": 4.8572,
      "mean_token_accuracy": 0.2253253623843193,
      "num_tokens": 82108565.0,
      "step": 36785
    },
    {
      "entropy": 5.279314708709717,
      "epoch": 4.131858257988432,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034010799564500197,
      "loss": 4.7429,
      "mean_token_accuracy": 0.23947744816541672,
      "num_tokens": 82119434.0,
      "step": 36790
    },
    {
      "entropy": 5.36896505355835,
      "epoch": 4.132419834896389,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003400695635456191,
      "loss": 4.978,
      "mean_token_accuracy": 0.2161254271864891,
      "num_tokens": 82130301.0,
      "step": 36795
    },
    {
      "entropy": 5.175950241088867,
      "epoch": 4.132981411804346,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003400311293743961,
      "loss": 4.7666,
      "mean_token_accuracy": 0.23172130584716796,
      "num_tokens": 82141596.0,
      "step": 36800
    },
    {
      "entropy": 5.35084867477417,
      "epoch": 4.133542988712304,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003399926931325567,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21897535324096679,
      "num_tokens": 82152779.0,
      "step": 36805
    },
    {
      "entropy": 5.28402943611145,
      "epoch": 4.1341045656202615,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00033995425482132464,
      "loss": 4.8349,
      "mean_token_accuracy": 0.22855866104364395,
      "num_tokens": 82164339.0,
      "step": 36810
    },
    {
      "entropy": 5.400484037399292,
      "epoch": 4.13466614252822,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033991581444192405,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21687304377555847,
      "num_tokens": 82176160.0,
      "step": 36815
    },
    {
      "entropy": 5.371975994110107,
      "epoch": 4.135227719436177,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003398773719955787,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2232508987188339,
      "num_tokens": 82188130.0,
      "step": 36820
    },
    {
      "entropy": 5.299878406524658,
      "epoch": 4.135789296344134,
      "grad_norm": 1.0,
      "learning_rate": 0.0003398389274835126,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22035093903541564,
      "num_tokens": 82199585.0,
      "step": 36825
    },
    {
      "entropy": 5.37008318901062,
      "epoch": 4.136350873252092,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033980048090694997,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22746842205524445,
      "num_tokens": 82210319.0,
      "step": 36830
    },
    {
      "entropy": 5.333677816390991,
      "epoch": 4.136912450160049,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003397620322671149,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22761449813842774,
      "num_tokens": 82222597.0,
      "step": 36835
    },
    {
      "entropy": 5.294386625289917,
      "epoch": 4.137474027068007,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033972358156523163,
      "loss": 4.8497,
      "mean_token_accuracy": 0.23704640567302704,
      "num_tokens": 82233651.0,
      "step": 36840
    },
    {
      "entropy": 5.205439567565918,
      "epoch": 4.138035603975965,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003396851288025243,
      "loss": 4.82,
      "mean_token_accuracy": 0.23503964990377427,
      "num_tokens": 82244012.0,
      "step": 36845
    },
    {
      "entropy": 5.338610982894897,
      "epoch": 4.138597180883922,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003396466739802175,
      "loss": 4.896,
      "mean_token_accuracy": 0.2273570030927658,
      "num_tokens": 82255290.0,
      "step": 36850
    },
    {
      "entropy": 5.428255319595337,
      "epoch": 4.13915875779188,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003396082170995355,
      "loss": 4.9095,
      "mean_token_accuracy": 0.2239511176943779,
      "num_tokens": 82266242.0,
      "step": 36855
    },
    {
      "entropy": 5.279091930389404,
      "epoch": 4.139720334699837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003395697581617028,
      "loss": 4.8544,
      "mean_token_accuracy": 0.23276604562997819,
      "num_tokens": 82276937.0,
      "step": 36860
    },
    {
      "entropy": 5.310982275009155,
      "epoch": 4.1402819116077945,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033953129716794396,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2271154284477234,
      "num_tokens": 82287566.0,
      "step": 36865
    },
    {
      "entropy": 5.322726774215698,
      "epoch": 4.140843488515753,
      "grad_norm": 1.0,
      "learning_rate": 0.0003394928341194836,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22682537585496904,
      "num_tokens": 82299464.0,
      "step": 36870
    },
    {
      "entropy": 5.278091478347778,
      "epoch": 4.14140506542371,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003394543690175464,
      "loss": 4.8328,
      "mean_token_accuracy": 0.23418940603733063,
      "num_tokens": 82310205.0,
      "step": 36875
    },
    {
      "entropy": 5.4160013675689695,
      "epoch": 4.141966642331667,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033941590186335703,
      "loss": 4.971,
      "mean_token_accuracy": 0.2292799785733223,
      "num_tokens": 82321591.0,
      "step": 36880
    },
    {
      "entropy": 5.196966028213501,
      "epoch": 4.142528219239625,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003393774326581404,
      "loss": 4.8277,
      "mean_token_accuracy": 0.22784799933433533,
      "num_tokens": 82333263.0,
      "step": 36885
    },
    {
      "entropy": 5.293696689605713,
      "epoch": 4.143089796147582,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033933896140312135,
      "loss": 4.9413,
      "mean_token_accuracy": 0.22517202645540238,
      "num_tokens": 82344627.0,
      "step": 36890
    },
    {
      "entropy": 5.366280269622803,
      "epoch": 4.1436513730555395,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033930048809952484,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21766890287399293,
      "num_tokens": 82355404.0,
      "step": 36895
    },
    {
      "entropy": 5.348947429656983,
      "epoch": 4.144212949963498,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033926201274857575,
      "loss": 4.899,
      "mean_token_accuracy": 0.22629986107349395,
      "num_tokens": 82365383.0,
      "step": 36900
    },
    {
      "entropy": 5.376533889770508,
      "epoch": 4.144774526871455,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033922353535149934,
      "loss": 4.9911,
      "mean_token_accuracy": 0.2205028638243675,
      "num_tokens": 82376535.0,
      "step": 36905
    },
    {
      "entropy": 5.317846155166626,
      "epoch": 4.145336103779413,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003391850559095207,
      "loss": 4.9021,
      "mean_token_accuracy": 0.2183859810233116,
      "num_tokens": 82387732.0,
      "step": 36910
    },
    {
      "entropy": 5.32937331199646,
      "epoch": 4.14589768068737,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003391465744238649,
      "loss": 4.8561,
      "mean_token_accuracy": 0.2312206283211708,
      "num_tokens": 82399486.0,
      "step": 36915
    },
    {
      "entropy": 5.283699893951416,
      "epoch": 4.146459257595327,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003391080908957573,
      "loss": 4.8228,
      "mean_token_accuracy": 0.2293547883629799,
      "num_tokens": 82411254.0,
      "step": 36920
    },
    {
      "entropy": 5.2774817943573,
      "epoch": 4.1470208345032855,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033906960532642327,
      "loss": 4.8736,
      "mean_token_accuracy": 0.21957930326461791,
      "num_tokens": 82421467.0,
      "step": 36925
    },
    {
      "entropy": 5.34038143157959,
      "epoch": 4.147582411411243,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003390311177170881,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2229652374982834,
      "num_tokens": 82432492.0,
      "step": 36930
    },
    {
      "entropy": 5.302399921417236,
      "epoch": 4.1481439883192,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003389926280689773,
      "loss": 4.8442,
      "mean_token_accuracy": 0.22877465188503265,
      "num_tokens": 82441495.0,
      "step": 36935
    },
    {
      "entropy": 5.293322896957397,
      "epoch": 4.148705565227158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033895413638331646,
      "loss": 4.9182,
      "mean_token_accuracy": 0.22011443823575974,
      "num_tokens": 82452811.0,
      "step": 36940
    },
    {
      "entropy": 5.336159658432007,
      "epoch": 4.149267142135115,
      "grad_norm": 1.0,
      "learning_rate": 0.0003389156426613311,
      "loss": 4.9439,
      "mean_token_accuracy": 0.2258983537554741,
      "num_tokens": 82463628.0,
      "step": 36945
    },
    {
      "entropy": 5.339464282989502,
      "epoch": 4.149828719043073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033887714690424687,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21806738525629044,
      "num_tokens": 82473884.0,
      "step": 36950
    },
    {
      "entropy": 5.403109788894653,
      "epoch": 4.150390295951031,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033883864911328954,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22280558794736863,
      "num_tokens": 82485714.0,
      "step": 36955
    },
    {
      "entropy": 5.294149589538574,
      "epoch": 4.150951872858988,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033880014928968483,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22357550412416458,
      "num_tokens": 82497014.0,
      "step": 36960
    },
    {
      "entropy": 5.346518611907959,
      "epoch": 4.151513449766946,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033876164743465865,
      "loss": 4.9596,
      "mean_token_accuracy": 0.22611394971609117,
      "num_tokens": 82508751.0,
      "step": 36965
    },
    {
      "entropy": 5.334374332427979,
      "epoch": 4.152075026674903,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003387231435494368,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22325711250305175,
      "num_tokens": 82519981.0,
      "step": 36970
    },
    {
      "entropy": 5.365139198303223,
      "epoch": 4.15263660358286,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033868463763524545,
      "loss": 4.9971,
      "mean_token_accuracy": 0.2160232409834862,
      "num_tokens": 82531650.0,
      "step": 36975
    },
    {
      "entropy": 5.3403853416442875,
      "epoch": 4.1531981804908185,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00033864612969331046,
      "loss": 4.8307,
      "mean_token_accuracy": 0.22411110997200012,
      "num_tokens": 82542099.0,
      "step": 36980
    },
    {
      "entropy": 5.33960747718811,
      "epoch": 4.153759757398776,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003386076197248581,
      "loss": 4.8209,
      "mean_token_accuracy": 0.24569548517465592,
      "num_tokens": 82552918.0,
      "step": 36985
    },
    {
      "entropy": 5.315575313568115,
      "epoch": 4.154321334306733,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003385691077311144,
      "loss": 4.9725,
      "mean_token_accuracy": 0.219047711789608,
      "num_tokens": 82564432.0,
      "step": 36990
    },
    {
      "entropy": 5.284668445587158,
      "epoch": 4.154882911214691,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033853059371330565,
      "loss": 4.9323,
      "mean_token_accuracy": 0.2173216834664345,
      "num_tokens": 82576100.0,
      "step": 36995
    },
    {
      "entropy": 5.3536614894866945,
      "epoch": 4.155444488122648,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003384920776726582,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21772318482398986,
      "num_tokens": 82588152.0,
      "step": 37000
    },
    {
      "entropy": 5.3722950458526615,
      "epoch": 4.156006065030606,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033845355961039823,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22976850271224974,
      "num_tokens": 82599078.0,
      "step": 37005
    },
    {
      "entropy": 5.293517971038819,
      "epoch": 4.1565676419385635,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033841503952775255,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22267286479473114,
      "num_tokens": 82609262.0,
      "step": 37010
    },
    {
      "entropy": 5.346908044815064,
      "epoch": 4.157129218846521,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00033837651742594734,
      "loss": 4.9202,
      "mean_token_accuracy": 0.2225680783390999,
      "num_tokens": 82620886.0,
      "step": 37015
    },
    {
      "entropy": 5.347197723388672,
      "epoch": 4.157690795754479,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033833799330620917,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21742407530546187,
      "num_tokens": 82632270.0,
      "step": 37020
    },
    {
      "entropy": 5.322649192810059,
      "epoch": 4.158252372662436,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003382994671697647,
      "loss": 4.8935,
      "mean_token_accuracy": 0.2322307914495468,
      "num_tokens": 82643099.0,
      "step": 37025
    },
    {
      "entropy": 5.374436330795288,
      "epoch": 4.158813949570393,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033826093901784065,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22089722156524658,
      "num_tokens": 82653500.0,
      "step": 37030
    },
    {
      "entropy": 5.343902730941773,
      "epoch": 4.159375526478351,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033822240885166384,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22447762191295623,
      "num_tokens": 82664074.0,
      "step": 37035
    },
    {
      "entropy": 5.337005376815796,
      "epoch": 4.159937103386309,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033818387667246095,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2288151815533638,
      "num_tokens": 82674915.0,
      "step": 37040
    },
    {
      "entropy": 5.293570947647095,
      "epoch": 4.160498680294267,
      "grad_norm": 0.9375,
      "learning_rate": 0.000338145342481459,
      "loss": 4.8868,
      "mean_token_accuracy": 0.22743179500102997,
      "num_tokens": 82686452.0,
      "step": 37045
    },
    {
      "entropy": 5.275090169906616,
      "epoch": 4.161060257202224,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003381068062798848,
      "loss": 4.8462,
      "mean_token_accuracy": 0.2292952373623848,
      "num_tokens": 82696448.0,
      "step": 37050
    },
    {
      "entropy": 5.349415302276611,
      "epoch": 4.161621834110181,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003380682680689654,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22297014594078063,
      "num_tokens": 82706637.0,
      "step": 37055
    },
    {
      "entropy": 5.379631090164184,
      "epoch": 4.162183411018139,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033802972784992784,
      "loss": 5.0316,
      "mean_token_accuracy": 0.21922099888324736,
      "num_tokens": 82718454.0,
      "step": 37060
    },
    {
      "entropy": 5.340259981155396,
      "epoch": 4.1627449879260965,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003379911856239994,
      "loss": 4.802,
      "mean_token_accuracy": 0.23701030462980271,
      "num_tokens": 82729405.0,
      "step": 37065
    },
    {
      "entropy": 5.286810350418091,
      "epoch": 4.163306564834054,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003379526413924071,
      "loss": 4.8895,
      "mean_token_accuracy": 0.22214857190847398,
      "num_tokens": 82739342.0,
      "step": 37070
    },
    {
      "entropy": 5.310593366622925,
      "epoch": 4.163868141742012,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003379140951563784,
      "loss": 4.8208,
      "mean_token_accuracy": 0.2256656989455223,
      "num_tokens": 82749460.0,
      "step": 37075
    },
    {
      "entropy": 5.407567310333252,
      "epoch": 4.164429718649969,
      "grad_norm": 1.0,
      "learning_rate": 0.00033787554691714033,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2196066990494728,
      "num_tokens": 82760672.0,
      "step": 37080
    },
    {
      "entropy": 5.354074096679687,
      "epoch": 4.164991295557926,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033783699667592066,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22541877627372742,
      "num_tokens": 82772116.0,
      "step": 37085
    },
    {
      "entropy": 5.295128870010376,
      "epoch": 4.165552872465884,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033779844443394656,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22108893245458602,
      "num_tokens": 82782376.0,
      "step": 37090
    },
    {
      "entropy": 5.264483118057251,
      "epoch": 4.166114449373842,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033775989019244556,
      "loss": 4.8487,
      "mean_token_accuracy": 0.22765025794506072,
      "num_tokens": 82794918.0,
      "step": 37095
    },
    {
      "entropy": 5.393333005905151,
      "epoch": 4.1666760262818,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033772133395264535,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2272569566965103,
      "num_tokens": 82805355.0,
      "step": 37100
    },
    {
      "entropy": 5.249981832504273,
      "epoch": 4.167237603189757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033768277571577364,
      "loss": 4.8644,
      "mean_token_accuracy": 0.23298754096031188,
      "num_tokens": 82815618.0,
      "step": 37105
    },
    {
      "entropy": 5.263239812850952,
      "epoch": 4.167799180097714,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033764421548305795,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2280909925699234,
      "num_tokens": 82826648.0,
      "step": 37110
    },
    {
      "entropy": 5.411647605895996,
      "epoch": 4.168360757005672,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033760565325572614,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22387369871139526,
      "num_tokens": 82838297.0,
      "step": 37115
    },
    {
      "entropy": 5.423958778381348,
      "epoch": 4.168922333913629,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003375670890350061,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21965640038251877,
      "num_tokens": 82849824.0,
      "step": 37120
    },
    {
      "entropy": 5.342590570449829,
      "epoch": 4.169483910821587,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033752852282212557,
      "loss": 4.8798,
      "mean_token_accuracy": 0.23361082524061202,
      "num_tokens": 82861366.0,
      "step": 37125
    },
    {
      "entropy": 5.232738828659057,
      "epoch": 4.170045487729545,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003374899546183128,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22057712078094482,
      "num_tokens": 82874142.0,
      "step": 37130
    },
    {
      "entropy": 5.248940086364746,
      "epoch": 4.170607064637502,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033745138442479544,
      "loss": 4.8748,
      "mean_token_accuracy": 0.228265218436718,
      "num_tokens": 82885786.0,
      "step": 37135
    },
    {
      "entropy": 5.288591480255127,
      "epoch": 4.17116864154546,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003374128122428019,
      "loss": 4.8865,
      "mean_token_accuracy": 0.22655483186244965,
      "num_tokens": 82896876.0,
      "step": 37140
    },
    {
      "entropy": 5.351944398880005,
      "epoch": 4.171730218453417,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003373742380735602,
      "loss": 4.8331,
      "mean_token_accuracy": 0.233024862408638,
      "num_tokens": 82908016.0,
      "step": 37145
    },
    {
      "entropy": 5.339782571792602,
      "epoch": 4.1722917953613745,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003373356619182985,
      "loss": 4.9523,
      "mean_token_accuracy": 0.23154925107955932,
      "num_tokens": 82919969.0,
      "step": 37150
    },
    {
      "entropy": 5.267319631576538,
      "epoch": 4.172853372269333,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003372970837782453,
      "loss": 4.8129,
      "mean_token_accuracy": 0.23327140659093856,
      "num_tokens": 82929729.0,
      "step": 37155
    },
    {
      "entropy": 5.30239052772522,
      "epoch": 4.17341494917729,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003372585036546287,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22484804391860963,
      "num_tokens": 82940888.0,
      "step": 37160
    },
    {
      "entropy": 5.3329582691192625,
      "epoch": 4.173976526085247,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003372199215486772,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22505563050508498,
      "num_tokens": 82953206.0,
      "step": 37165
    },
    {
      "entropy": 5.408367156982422,
      "epoch": 4.174538102993205,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003371813374616192,
      "loss": 4.9639,
      "mean_token_accuracy": 0.22025763243436813,
      "num_tokens": 82963979.0,
      "step": 37170
    },
    {
      "entropy": 5.38442177772522,
      "epoch": 4.175099679901162,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033714275139468336,
      "loss": 4.964,
      "mean_token_accuracy": 0.22308753281831742,
      "num_tokens": 82973673.0,
      "step": 37175
    },
    {
      "entropy": 5.2816386222839355,
      "epoch": 4.17566125680912,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003371041633490982,
      "loss": 4.908,
      "mean_token_accuracy": 0.2241467997431755,
      "num_tokens": 82983653.0,
      "step": 37180
    },
    {
      "entropy": 5.280409145355224,
      "epoch": 4.176222833717078,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003370655733260924,
      "loss": 4.8327,
      "mean_token_accuracy": 0.22640879154205323,
      "num_tokens": 82993631.0,
      "step": 37185
    },
    {
      "entropy": 5.242314338684082,
      "epoch": 4.176784410625035,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033702698132689467,
      "loss": 4.8032,
      "mean_token_accuracy": 0.23242261856794358,
      "num_tokens": 83003872.0,
      "step": 37190
    },
    {
      "entropy": 5.319524049758911,
      "epoch": 4.177345987532993,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003369883873527338,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22850140929222107,
      "num_tokens": 83016779.0,
      "step": 37195
    },
    {
      "entropy": 5.309934186935425,
      "epoch": 4.17790756444095,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033694979140483865,
      "loss": 4.9025,
      "mean_token_accuracy": 0.2268262967467308,
      "num_tokens": 83028544.0,
      "step": 37200
    },
    {
      "entropy": 5.265503454208374,
      "epoch": 4.1784691413489075,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003369111934844381,
      "loss": 4.7939,
      "mean_token_accuracy": 0.23089101165533066,
      "num_tokens": 83039040.0,
      "step": 37205
    },
    {
      "entropy": 5.263065814971924,
      "epoch": 4.179030718256866,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033687259359276115,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22701384425163268,
      "num_tokens": 83050436.0,
      "step": 37210
    },
    {
      "entropy": 5.337719440460205,
      "epoch": 4.179592295164823,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003368339917310368,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22338382452726363,
      "num_tokens": 83062478.0,
      "step": 37215
    },
    {
      "entropy": 5.39286789894104,
      "epoch": 4.18015387207278,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003367953879004942,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22021283358335494,
      "num_tokens": 83073337.0,
      "step": 37220
    },
    {
      "entropy": 5.421606636047363,
      "epoch": 4.180715448980738,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003367567821023625,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21748885810375213,
      "num_tokens": 83087454.0,
      "step": 37225
    },
    {
      "entropy": 5.377406644821167,
      "epoch": 4.181277025888695,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003367181743378708,
      "loss": 4.9023,
      "mean_token_accuracy": 0.23181816935539246,
      "num_tokens": 83099037.0,
      "step": 37230
    },
    {
      "entropy": 5.284109306335449,
      "epoch": 4.181838602796653,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033667956460824866,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22904450744390487,
      "num_tokens": 83109765.0,
      "step": 37235
    },
    {
      "entropy": 5.309484577178955,
      "epoch": 4.182400179704611,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003366409529147251,
      "loss": 4.8893,
      "mean_token_accuracy": 0.2197311669588089,
      "num_tokens": 83120348.0,
      "step": 37240
    },
    {
      "entropy": 5.324617576599121,
      "epoch": 4.182961756612568,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033660233925852977,
      "loss": 4.9087,
      "mean_token_accuracy": 0.2208218827843666,
      "num_tokens": 83130744.0,
      "step": 37245
    },
    {
      "entropy": 5.2930906295776365,
      "epoch": 4.183523333520526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033656372364089206,
      "loss": 4.8382,
      "mean_token_accuracy": 0.2355899214744568,
      "num_tokens": 83140897.0,
      "step": 37250
    },
    {
      "entropy": 5.360458946228027,
      "epoch": 4.184084910428483,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00033652510606304155,
      "loss": 4.9532,
      "mean_token_accuracy": 0.21869450509548188,
      "num_tokens": 83153364.0,
      "step": 37255
    },
    {
      "entropy": 5.210064268112182,
      "epoch": 4.18464648733644,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033648648652620766,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22814815491437912,
      "num_tokens": 83164276.0,
      "step": 37260
    },
    {
      "entropy": 5.33187165260315,
      "epoch": 4.1852080642443985,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033644786503162036,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2224784255027771,
      "num_tokens": 83175585.0,
      "step": 37265
    },
    {
      "entropy": 5.350651645660401,
      "epoch": 4.185769641152356,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003364092415805091,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23852764666080475,
      "num_tokens": 83185893.0,
      "step": 37270
    },
    {
      "entropy": 5.233529758453369,
      "epoch": 4.186331218060313,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033637061617410377,
      "loss": 4.8709,
      "mean_token_accuracy": 0.22415133118629454,
      "num_tokens": 83195990.0,
      "step": 37275
    },
    {
      "entropy": 5.421274328231812,
      "epoch": 4.186892794968271,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003363319888136343,
      "loss": 5.0288,
      "mean_token_accuracy": 0.21949052363634108,
      "num_tokens": 83206154.0,
      "step": 37280
    },
    {
      "entropy": 5.337035369873047,
      "epoch": 4.187454371876228,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003362933595003305,
      "loss": 4.8657,
      "mean_token_accuracy": 0.22797987461090088,
      "num_tokens": 83217984.0,
      "step": 37285
    },
    {
      "entropy": 5.327794551849365,
      "epoch": 4.188015948784186,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003362547282354223,
      "loss": 4.9098,
      "mean_token_accuracy": 0.21761895716190338,
      "num_tokens": 83229693.0,
      "step": 37290
    },
    {
      "entropy": 5.301523733139038,
      "epoch": 4.188577525692144,
      "grad_norm": 1.0,
      "learning_rate": 0.0003362160950201398,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22668897956609727,
      "num_tokens": 83240771.0,
      "step": 37295
    },
    {
      "entropy": 5.256708097457886,
      "epoch": 4.189139102600101,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003361774598557132,
      "loss": 4.8072,
      "mean_token_accuracy": 0.23785578161478044,
      "num_tokens": 83252203.0,
      "step": 37300
    },
    {
      "entropy": 5.439500904083252,
      "epoch": 4.189700679508059,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033613882274337244,
      "loss": 5.0247,
      "mean_token_accuracy": 0.22049438804388047,
      "num_tokens": 83264103.0,
      "step": 37305
    },
    {
      "entropy": 5.377615118026734,
      "epoch": 4.190262256416016,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003361001836843479,
      "loss": 4.958,
      "mean_token_accuracy": 0.2198164865374565,
      "num_tokens": 83275484.0,
      "step": 37310
    },
    {
      "entropy": 5.245244789123535,
      "epoch": 4.190823833323973,
      "grad_norm": 1.0,
      "learning_rate": 0.0003360615426798697,
      "loss": 4.8558,
      "mean_token_accuracy": 0.23056099861860274,
      "num_tokens": 83286150.0,
      "step": 37315
    },
    {
      "entropy": 5.284641170501709,
      "epoch": 4.1913854102319315,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003360228997311685,
      "loss": 4.8604,
      "mean_token_accuracy": 0.22492969483137132,
      "num_tokens": 83297574.0,
      "step": 37320
    },
    {
      "entropy": 5.387963485717774,
      "epoch": 4.191946987139889,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003359842548394744,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22299590557813645,
      "num_tokens": 83309423.0,
      "step": 37325
    },
    {
      "entropy": 5.330846977233887,
      "epoch": 4.192508564047847,
      "grad_norm": 1.0,
      "learning_rate": 0.0003359456080060179,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21651584208011626,
      "num_tokens": 83321003.0,
      "step": 37330
    },
    {
      "entropy": 5.407022094726562,
      "epoch": 4.193070140955804,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003359069592320297,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2208404064178467,
      "num_tokens": 83332513.0,
      "step": 37335
    },
    {
      "entropy": 5.4071447372436525,
      "epoch": 4.193631717863761,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003358683085187403,
      "loss": 4.9545,
      "mean_token_accuracy": 0.22954854369163513,
      "num_tokens": 83344250.0,
      "step": 37340
    },
    {
      "entropy": 5.374446249008178,
      "epoch": 4.194193294771719,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033582965586738037,
      "loss": 5.014,
      "mean_token_accuracy": 0.21904454976320267,
      "num_tokens": 83357359.0,
      "step": 37345
    },
    {
      "entropy": 5.415427875518799,
      "epoch": 4.1947548716796765,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003357910012791804,
      "loss": 5.0709,
      "mean_token_accuracy": 0.2096161127090454,
      "num_tokens": 83369175.0,
      "step": 37350
    },
    {
      "entropy": 5.517972040176391,
      "epoch": 4.195316448587634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003357523447553716,
      "loss": 5.1374,
      "mean_token_accuracy": 0.21501512974500656,
      "num_tokens": 83380977.0,
      "step": 37355
    },
    {
      "entropy": 5.449036455154419,
      "epoch": 4.195878025495592,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00033571368629718457,
      "loss": 4.9887,
      "mean_token_accuracy": 0.2270175576210022,
      "num_tokens": 83393915.0,
      "step": 37360
    },
    {
      "entropy": 5.370309352874756,
      "epoch": 4.196439602403549,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033567502590585007,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22436715066432952,
      "num_tokens": 83405608.0,
      "step": 37365
    },
    {
      "entropy": 5.365134811401367,
      "epoch": 4.197001179311506,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003356363635825994,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2294543504714966,
      "num_tokens": 83417185.0,
      "step": 37370
    },
    {
      "entropy": 5.261742305755615,
      "epoch": 4.197562756219464,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003355976993286633,
      "loss": 4.8666,
      "mean_token_accuracy": 0.2297522872686386,
      "num_tokens": 83428648.0,
      "step": 37375
    },
    {
      "entropy": 5.333370876312256,
      "epoch": 4.198124333127422,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003355590331452729,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22148634791374205,
      "num_tokens": 83439860.0,
      "step": 37380
    },
    {
      "entropy": 5.366733503341675,
      "epoch": 4.19868591003538,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033552036503365945,
      "loss": 4.9397,
      "mean_token_accuracy": 0.21968822181224823,
      "num_tokens": 83451163.0,
      "step": 37385
    },
    {
      "entropy": 5.303592538833618,
      "epoch": 4.199247486943337,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003354816949950541,
      "loss": 4.8215,
      "mean_token_accuracy": 0.22591274082660676,
      "num_tokens": 83461584.0,
      "step": 37390
    },
    {
      "entropy": 5.350347709655762,
      "epoch": 4.199809063851294,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033544302303068814,
      "loss": 4.8953,
      "mean_token_accuracy": 0.229937182366848,
      "num_tokens": 83473243.0,
      "step": 37395
    },
    {
      "entropy": 5.311123991012574,
      "epoch": 4.200370640759252,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003354043491417929,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22909935414791108,
      "num_tokens": 83486143.0,
      "step": 37400
    },
    {
      "entropy": 5.366778707504272,
      "epoch": 4.2009322176672095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033536567332959966,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2172014057636261,
      "num_tokens": 83497809.0,
      "step": 37405
    },
    {
      "entropy": 5.32752571105957,
      "epoch": 4.201493794575167,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00033532699559534,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22884568572044373,
      "num_tokens": 83510558.0,
      "step": 37410
    },
    {
      "entropy": 5.285845756530762,
      "epoch": 4.202055371483125,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033528831594024553,
      "loss": 4.8572,
      "mean_token_accuracy": 0.22786543220281602,
      "num_tokens": 83522272.0,
      "step": 37415
    },
    {
      "entropy": 5.388953304290771,
      "epoch": 4.202616948391082,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00033524963436554753,
      "loss": 5.0421,
      "mean_token_accuracy": 0.21877072751522064,
      "num_tokens": 83535588.0,
      "step": 37420
    },
    {
      "entropy": 5.47525053024292,
      "epoch": 4.20317852529904,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033521095087247794,
      "loss": 5.0249,
      "mean_token_accuracy": 0.21915501952171326,
      "num_tokens": 83548120.0,
      "step": 37425
    },
    {
      "entropy": 5.390908718109131,
      "epoch": 4.203740102206997,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003351722654622683,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2333230808377266,
      "num_tokens": 83559619.0,
      "step": 37430
    },
    {
      "entropy": 5.33630142211914,
      "epoch": 4.204301679114955,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003351335781361503,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2218332827091217,
      "num_tokens": 83571975.0,
      "step": 37435
    },
    {
      "entropy": 5.313950300216675,
      "epoch": 4.204863256022913,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033509488889535595,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21896237134933472,
      "num_tokens": 83583012.0,
      "step": 37440
    },
    {
      "entropy": 5.334907579421997,
      "epoch": 4.20542483293087,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033505619774111696,
      "loss": 4.8657,
      "mean_token_accuracy": 0.23230547308921815,
      "num_tokens": 83592998.0,
      "step": 37445
    },
    {
      "entropy": 5.403669977188111,
      "epoch": 4.205986409838827,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033501750467466535,
      "loss": 4.95,
      "mean_token_accuracy": 0.21858363151550292,
      "num_tokens": 83603669.0,
      "step": 37450
    },
    {
      "entropy": 5.2863945960998535,
      "epoch": 4.206547986746785,
      "grad_norm": 1.0,
      "learning_rate": 0.0003349788096972331,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22917922735214233,
      "num_tokens": 83613243.0,
      "step": 37455
    },
    {
      "entropy": 5.363780117034912,
      "epoch": 4.207109563654742,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003349401128100523,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22426746785640717,
      "num_tokens": 83625131.0,
      "step": 37460
    },
    {
      "entropy": 5.367554378509522,
      "epoch": 4.2076711405627,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033490141401435516,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2277054861187935,
      "num_tokens": 83635282.0,
      "step": 37465
    },
    {
      "entropy": 5.291355323791504,
      "epoch": 4.208232717470658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003348627133113736,
      "loss": 4.8337,
      "mean_token_accuracy": 0.2311304733157158,
      "num_tokens": 83647398.0,
      "step": 37470
    },
    {
      "entropy": 5.318142700195312,
      "epoch": 4.208794294378615,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033482401070234006,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2322290852665901,
      "num_tokens": 83657938.0,
      "step": 37475
    },
    {
      "entropy": 5.276027345657349,
      "epoch": 4.209355871286573,
      "grad_norm": 0.9375,
      "learning_rate": 0.000334785306188487,
      "loss": 4.811,
      "mean_token_accuracy": 0.229682357609272,
      "num_tokens": 83668240.0,
      "step": 37480
    },
    {
      "entropy": 5.221360063552856,
      "epoch": 4.20991744819453,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003347465997710464,
      "loss": 4.8394,
      "mean_token_accuracy": 0.2247369721531868,
      "num_tokens": 83682071.0,
      "step": 37485
    },
    {
      "entropy": 5.256880044937134,
      "epoch": 4.2104790251024875,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000334707891451251,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2289162680506706,
      "num_tokens": 83692895.0,
      "step": 37490
    },
    {
      "entropy": 5.273895406723023,
      "epoch": 4.211040602010446,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003346691812303331,
      "loss": 4.9395,
      "mean_token_accuracy": 0.22451485097408294,
      "num_tokens": 83705577.0,
      "step": 37495
    },
    {
      "entropy": 5.379204034805298,
      "epoch": 4.211602178918403,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003346304691095254,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22185145914554597,
      "num_tokens": 83716694.0,
      "step": 37500
    },
    {
      "entropy": 5.343792867660523,
      "epoch": 4.21216375582636,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003345917550900604,
      "loss": 4.9197,
      "mean_token_accuracy": 0.2169765442609787,
      "num_tokens": 83728744.0,
      "step": 37505
    },
    {
      "entropy": 5.341556406021118,
      "epoch": 4.212725332734318,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033455303917317075,
      "loss": 4.9104,
      "mean_token_accuracy": 0.22701396495103837,
      "num_tokens": 83740731.0,
      "step": 37510
    },
    {
      "entropy": 5.289283227920532,
      "epoch": 4.213286909642275,
      "grad_norm": 0.90625,
      "learning_rate": 0.00033451432136008937,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21846056878566741,
      "num_tokens": 83751742.0,
      "step": 37515
    },
    {
      "entropy": 5.384820413589478,
      "epoch": 4.2138484865502335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003344756016520488,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22772423326969146,
      "num_tokens": 83761771.0,
      "step": 37520
    },
    {
      "entropy": 5.254843616485596,
      "epoch": 4.214410063458191,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00033443688005028213,
      "loss": 4.7988,
      "mean_token_accuracy": 0.22964072078466416,
      "num_tokens": 83772394.0,
      "step": 37525
    },
    {
      "entropy": 5.3295127868652346,
      "epoch": 4.214971640366148,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000334398156556022,
      "loss": 5.0181,
      "mean_token_accuracy": 0.21431985795497893,
      "num_tokens": 83784828.0,
      "step": 37530
    },
    {
      "entropy": 5.316731929779053,
      "epoch": 4.215533217274106,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003343594311705016,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2225389912724495,
      "num_tokens": 83794840.0,
      "step": 37535
    },
    {
      "entropy": 5.382118844985962,
      "epoch": 4.216094794182063,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033432070389495386,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22602129876613616,
      "num_tokens": 83807135.0,
      "step": 37540
    },
    {
      "entropy": 5.444579601287842,
      "epoch": 4.2166563710900204,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003342819747306119,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21936125457286834,
      "num_tokens": 83818979.0,
      "step": 37545
    },
    {
      "entropy": 5.37984037399292,
      "epoch": 4.2172179479979786,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033424324367870896,
      "loss": 4.895,
      "mean_token_accuracy": 0.2246434897184372,
      "num_tokens": 83829568.0,
      "step": 37550
    },
    {
      "entropy": 5.447844648361206,
      "epoch": 4.217779524905936,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033420451074047807,
      "loss": 5.0742,
      "mean_token_accuracy": 0.2126983657479286,
      "num_tokens": 83841311.0,
      "step": 37555
    },
    {
      "entropy": 5.35557746887207,
      "epoch": 4.218341101813893,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003341657759171526,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2299878939986229,
      "num_tokens": 83853648.0,
      "step": 37560
    },
    {
      "entropy": 5.264745807647705,
      "epoch": 4.218902678721851,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003341270392099659,
      "loss": 4.8571,
      "mean_token_accuracy": 0.23345450013875962,
      "num_tokens": 83863848.0,
      "step": 37565
    },
    {
      "entropy": 5.30736198425293,
      "epoch": 4.219464255629808,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033408830062015135,
      "loss": 4.8443,
      "mean_token_accuracy": 0.23347631543874742,
      "num_tokens": 83874446.0,
      "step": 37570
    },
    {
      "entropy": 5.336557197570801,
      "epoch": 4.220025832537766,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033404956014894237,
      "loss": 4.8434,
      "mean_token_accuracy": 0.2346685752272606,
      "num_tokens": 83884681.0,
      "step": 37575
    },
    {
      "entropy": 5.298976564407349,
      "epoch": 4.220587409445724,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033401081779757254,
      "loss": 4.9158,
      "mean_token_accuracy": 0.22727809548377992,
      "num_tokens": 83895026.0,
      "step": 37580
    },
    {
      "entropy": 5.384079027175903,
      "epoch": 4.221148986353681,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003339720735672753,
      "loss": 4.9909,
      "mean_token_accuracy": 0.22236306816339493,
      "num_tokens": 83905855.0,
      "step": 37585
    },
    {
      "entropy": 5.281061410903931,
      "epoch": 4.221710563261639,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00033393332745928453,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22906559705734253,
      "num_tokens": 83918447.0,
      "step": 37590
    },
    {
      "entropy": 5.290679216384888,
      "epoch": 4.222272140169596,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033389457947483357,
      "loss": 4.8805,
      "mean_token_accuracy": 0.23264921605587005,
      "num_tokens": 83929068.0,
      "step": 37595
    },
    {
      "entropy": 5.331558513641357,
      "epoch": 4.222833717077553,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033385582961515645,
      "loss": 4.8951,
      "mean_token_accuracy": 0.2269457682967186,
      "num_tokens": 83939636.0,
      "step": 37600
    },
    {
      "entropy": 5.3921161651611325,
      "epoch": 4.2233952939855115,
      "grad_norm": 1.0,
      "learning_rate": 0.00033381707788148683,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22059089690446854,
      "num_tokens": 83950743.0,
      "step": 37605
    },
    {
      "entropy": 5.476277875900268,
      "epoch": 4.223956870893469,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033377832427505864,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21240339130163194,
      "num_tokens": 83963383.0,
      "step": 37610
    },
    {
      "entropy": 5.352823877334595,
      "epoch": 4.224518447801427,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033373956879710593,
      "loss": 4.8496,
      "mean_token_accuracy": 0.2345030978322029,
      "num_tokens": 83973768.0,
      "step": 37615
    },
    {
      "entropy": 5.293409633636474,
      "epoch": 4.225080024709384,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033370081144886246,
      "loss": 4.9152,
      "mean_token_accuracy": 0.22400543540716172,
      "num_tokens": 83984769.0,
      "step": 37620
    },
    {
      "entropy": 5.270657205581665,
      "epoch": 4.225641601617341,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00033366205223156237,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22660569250583648,
      "num_tokens": 83996894.0,
      "step": 37625
    },
    {
      "entropy": 5.308344030380249,
      "epoch": 4.226203178525299,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033362329114643987,
      "loss": 4.8936,
      "mean_token_accuracy": 0.2207052767276764,
      "num_tokens": 84008536.0,
      "step": 37630
    },
    {
      "entropy": 5.28873839378357,
      "epoch": 4.226764755433257,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033358452819472895,
      "loss": 4.8205,
      "mean_token_accuracy": 0.22345851808786393,
      "num_tokens": 84020239.0,
      "step": 37635
    },
    {
      "entropy": 5.399445819854736,
      "epoch": 4.227326332341214,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033354576337766396,
      "loss": 4.9522,
      "mean_token_accuracy": 0.22376303225755692,
      "num_tokens": 84031580.0,
      "step": 37640
    },
    {
      "entropy": 5.277694892883301,
      "epoch": 4.227887909249172,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003335069966964792,
      "loss": 4.8815,
      "mean_token_accuracy": 0.22409235835075378,
      "num_tokens": 84042111.0,
      "step": 37645
    },
    {
      "entropy": 5.39100079536438,
      "epoch": 4.228449486157129,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000333468228152409,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2189219132065773,
      "num_tokens": 84052048.0,
      "step": 37650
    },
    {
      "entropy": 5.379430818557739,
      "epoch": 4.229011063065086,
      "grad_norm": 1.0,
      "learning_rate": 0.00033342945774668766,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21678949594497682,
      "num_tokens": 84063766.0,
      "step": 37655
    },
    {
      "entropy": 5.319349670410157,
      "epoch": 4.2295726399730444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033339068548054974,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2186967611312866,
      "num_tokens": 84073992.0,
      "step": 37660
    },
    {
      "entropy": 5.284269046783447,
      "epoch": 4.230134216881002,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033335191135522974,
      "loss": 4.7848,
      "mean_token_accuracy": 0.2305704727768898,
      "num_tokens": 84084298.0,
      "step": 37665
    },
    {
      "entropy": 5.269844341278076,
      "epoch": 4.23069579378896,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033331313537196235,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22324786633253096,
      "num_tokens": 84093638.0,
      "step": 37670
    },
    {
      "entropy": 5.271061038970947,
      "epoch": 4.231257370696917,
      "grad_norm": 1.046875,
      "learning_rate": 0.000333274357531982,
      "loss": 4.827,
      "mean_token_accuracy": 0.22800702899694442,
      "num_tokens": 84104627.0,
      "step": 37675
    },
    {
      "entropy": 5.298124837875366,
      "epoch": 4.231818947604874,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003332355778365235,
      "loss": 4.7991,
      "mean_token_accuracy": 0.22745700180530548,
      "num_tokens": 84115474.0,
      "step": 37680
    },
    {
      "entropy": 5.337525272369385,
      "epoch": 4.232380524512832,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003331967962868217,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21558054387569428,
      "num_tokens": 84126101.0,
      "step": 37685
    },
    {
      "entropy": 5.3839293956756595,
      "epoch": 4.2329421014207895,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003331580128841112,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22063490152359008,
      "num_tokens": 84136497.0,
      "step": 37690
    },
    {
      "entropy": 5.347946977615356,
      "epoch": 4.233503678328747,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033311922762962716,
      "loss": 4.899,
      "mean_token_accuracy": 0.23074334859848022,
      "num_tokens": 84149001.0,
      "step": 37695
    },
    {
      "entropy": 5.276639986038208,
      "epoch": 4.234065255236705,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003330804405246043,
      "loss": 4.7743,
      "mean_token_accuracy": 0.2362628310918808,
      "num_tokens": 84159951.0,
      "step": 37700
    },
    {
      "entropy": 5.304737138748169,
      "epoch": 4.234626832144662,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003330416515702777,
      "loss": 4.8345,
      "mean_token_accuracy": 0.23683350682258605,
      "num_tokens": 84170387.0,
      "step": 37705
    },
    {
      "entropy": 5.391392087936401,
      "epoch": 4.23518840905262,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033300286076788237,
      "loss": 5.0153,
      "mean_token_accuracy": 0.22048023492097854,
      "num_tokens": 84182412.0,
      "step": 37710
    },
    {
      "entropy": 5.328062629699707,
      "epoch": 4.235749985960577,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003329640681186535,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23798673897981643,
      "num_tokens": 84193581.0,
      "step": 37715
    },
    {
      "entropy": 5.332671499252319,
      "epoch": 4.236311562868535,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00033292527362382626,
      "loss": 4.8932,
      "mean_token_accuracy": 0.22251082062721253,
      "num_tokens": 84204433.0,
      "step": 37720
    },
    {
      "entropy": 5.293264150619507,
      "epoch": 4.236873139776493,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033288647728463576,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22849951535463334,
      "num_tokens": 84216034.0,
      "step": 37725
    },
    {
      "entropy": 5.313537359237671,
      "epoch": 4.23743471668445,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033284767910231745,
      "loss": 4.8692,
      "mean_token_accuracy": 0.23325546085834503,
      "num_tokens": 84226385.0,
      "step": 37730
    },
    {
      "entropy": 5.3663228988647464,
      "epoch": 4.237996293592407,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003328088790781065,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22583019286394118,
      "num_tokens": 84236742.0,
      "step": 37735
    },
    {
      "entropy": 5.374253606796264,
      "epoch": 4.238557870500365,
      "grad_norm": 1.0,
      "learning_rate": 0.00033277007721323854,
      "loss": 4.9546,
      "mean_token_accuracy": 0.22547658979892732,
      "num_tokens": 84247969.0,
      "step": 37740
    },
    {
      "entropy": 5.3105169296264645,
      "epoch": 4.2391194474083225,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033273127350894884,
      "loss": 4.8667,
      "mean_token_accuracy": 0.2296567901968956,
      "num_tokens": 84260939.0,
      "step": 37745
    },
    {
      "entropy": 5.292934703826904,
      "epoch": 4.23968102431628,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033269246796647296,
      "loss": 4.9083,
      "mean_token_accuracy": 0.22671340852975846,
      "num_tokens": 84272972.0,
      "step": 37750
    },
    {
      "entropy": 5.36196494102478,
      "epoch": 4.240242601224238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003326536605870466,
      "loss": 4.9239,
      "mean_token_accuracy": 0.22532738298177718,
      "num_tokens": 84283920.0,
      "step": 37755
    },
    {
      "entropy": 5.354731130599975,
      "epoch": 4.240804178132195,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033261485137190533,
      "loss": 4.96,
      "mean_token_accuracy": 0.22362091839313508,
      "num_tokens": 84296163.0,
      "step": 37760
    },
    {
      "entropy": 5.2861656665802,
      "epoch": 4.241365755040153,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033257604032228485,
      "loss": 4.8448,
      "mean_token_accuracy": 0.231043803691864,
      "num_tokens": 84306647.0,
      "step": 37765
    },
    {
      "entropy": 5.222252416610718,
      "epoch": 4.24192733194811,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00033253722743942086,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23367975801229476,
      "num_tokens": 84318678.0,
      "step": 37770
    },
    {
      "entropy": 5.2198834896087645,
      "epoch": 4.2424889088560676,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003324984127245492,
      "loss": 4.8268,
      "mean_token_accuracy": 0.2334268197417259,
      "num_tokens": 84329954.0,
      "step": 37775
    },
    {
      "entropy": 5.300224351882934,
      "epoch": 4.243050485764026,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003324595961789058,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22382071912288665,
      "num_tokens": 84341157.0,
      "step": 37780
    },
    {
      "entropy": 5.362799835205078,
      "epoch": 4.243612062671983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033242077780372664,
      "loss": 4.9204,
      "mean_token_accuracy": 0.22149030715227128,
      "num_tokens": 84352893.0,
      "step": 37785
    },
    {
      "entropy": 5.328584861755371,
      "epoch": 4.24417363957994,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033238195760024766,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22386250346899034,
      "num_tokens": 84365019.0,
      "step": 37790
    },
    {
      "entropy": 5.300116682052613,
      "epoch": 4.244735216487898,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00033234313556970485,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22758695781230925,
      "num_tokens": 84376397.0,
      "step": 37795
    },
    {
      "entropy": 5.301068735122681,
      "epoch": 4.245296793395855,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033230431171333435,
      "loss": 4.8953,
      "mean_token_accuracy": 0.22814767807722092,
      "num_tokens": 84387649.0,
      "step": 37800
    },
    {
      "entropy": 5.415524435043335,
      "epoch": 4.2458583703038135,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003322654860323723,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2105608880519867,
      "num_tokens": 84399471.0,
      "step": 37805
    },
    {
      "entropy": 5.413452529907227,
      "epoch": 4.246419947211771,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00033222665852805514,
      "loss": 4.8826,
      "mean_token_accuracy": 0.22914549261331557,
      "num_tokens": 84409975.0,
      "step": 37810
    },
    {
      "entropy": 5.333680486679077,
      "epoch": 4.246981524119728,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033218782920161884,
      "loss": 4.913,
      "mean_token_accuracy": 0.22853086441755294,
      "num_tokens": 84421851.0,
      "step": 37815
    },
    {
      "entropy": 5.237923431396484,
      "epoch": 4.247543101027686,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033214899805429984,
      "loss": 4.822,
      "mean_token_accuracy": 0.23349526524543762,
      "num_tokens": 84433021.0,
      "step": 37820
    },
    {
      "entropy": 5.302494812011719,
      "epoch": 4.248104677935643,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00033211016508733463,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22031232267618178,
      "num_tokens": 84443635.0,
      "step": 37825
    },
    {
      "entropy": 5.368982267379761,
      "epoch": 4.2486662548436005,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033207133030195964,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21550064533948898,
      "num_tokens": 84455259.0,
      "step": 37830
    },
    {
      "entropy": 5.360562562942505,
      "epoch": 4.249227831751559,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003320324936994113,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22678565979003906,
      "num_tokens": 84465080.0,
      "step": 37835
    },
    {
      "entropy": 5.343709135055542,
      "epoch": 4.249789408659516,
      "grad_norm": 1.0,
      "learning_rate": 0.00033199365528092625,
      "loss": 4.933,
      "mean_token_accuracy": 0.2243219718337059,
      "num_tokens": 84476070.0,
      "step": 37840
    },
    {
      "entropy": 5.2715473651885985,
      "epoch": 4.250350985567474,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033195481504774113,
      "loss": 4.8297,
      "mean_token_accuracy": 0.22602640390396117,
      "num_tokens": 84486496.0,
      "step": 37845
    },
    {
      "entropy": 5.425025558471679,
      "epoch": 4.250912562475431,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033191597300109263,
      "loss": 4.931,
      "mean_token_accuracy": 0.2235083594918251,
      "num_tokens": 84497086.0,
      "step": 37850
    },
    {
      "entropy": 5.370567989349365,
      "epoch": 4.251474139383388,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033187712914221744,
      "loss": 4.931,
      "mean_token_accuracy": 0.22409696131944656,
      "num_tokens": 84508498.0,
      "step": 37855
    },
    {
      "entropy": 5.36091799736023,
      "epoch": 4.2520357162913465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003318382834723524,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22141028195619583,
      "num_tokens": 84519117.0,
      "step": 37860
    },
    {
      "entropy": 5.395045328140259,
      "epoch": 4.252597293199304,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033179943599273445,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22491591721773146,
      "num_tokens": 84530752.0,
      "step": 37865
    },
    {
      "entropy": 5.3676458358764645,
      "epoch": 4.253158870107261,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003317605867046003,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21917640268802643,
      "num_tokens": 84542989.0,
      "step": 37870
    },
    {
      "entropy": 5.335285329818726,
      "epoch": 4.253720447015219,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033172173560918716,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22704195231199265,
      "num_tokens": 84554426.0,
      "step": 37875
    },
    {
      "entropy": 5.264306688308716,
      "epoch": 4.254282023923176,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033168288270773197,
      "loss": 4.781,
      "mean_token_accuracy": 0.23185194730758668,
      "num_tokens": 84565669.0,
      "step": 37880
    },
    {
      "entropy": 5.330320501327515,
      "epoch": 4.254843600831133,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033164402800147177,
      "loss": 4.9164,
      "mean_token_accuracy": 0.23119474500417708,
      "num_tokens": 84576967.0,
      "step": 37885
    },
    {
      "entropy": 5.321692609786988,
      "epoch": 4.2554051777390915,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003316051714916436,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22319069355726243,
      "num_tokens": 84586828.0,
      "step": 37890
    },
    {
      "entropy": 5.387816572189331,
      "epoch": 4.255966754647049,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033156631317948505,
      "loss": 4.9746,
      "mean_token_accuracy": 0.22780648320913316,
      "num_tokens": 84597266.0,
      "step": 37895
    },
    {
      "entropy": 5.4791759014129635,
      "epoch": 4.256528331555007,
      "grad_norm": 1.046875,
      "learning_rate": 0.000331527453066233,
      "loss": 5.0183,
      "mean_token_accuracy": 0.2259109601378441,
      "num_tokens": 84609170.0,
      "step": 37900
    },
    {
      "entropy": 5.354374408721924,
      "epoch": 4.257089908462964,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000331488591153125,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22424134165048598,
      "num_tokens": 84620415.0,
      "step": 37905
    },
    {
      "entropy": 5.346922731399536,
      "epoch": 4.257651485370921,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003314497274413983,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22715600579977036,
      "num_tokens": 84631417.0,
      "step": 37910
    },
    {
      "entropy": 5.440149354934692,
      "epoch": 4.258213062278879,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033141086193229043,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21696187406778336,
      "num_tokens": 84642528.0,
      "step": 37915
    },
    {
      "entropy": 5.337555551528931,
      "epoch": 4.258774639186837,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003313719946270388,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23117099404335023,
      "num_tokens": 84653785.0,
      "step": 37920
    },
    {
      "entropy": 5.271041202545166,
      "epoch": 4.259336216094794,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033133312552688094,
      "loss": 4.9175,
      "mean_token_accuracy": 0.21926555633544922,
      "num_tokens": 84664533.0,
      "step": 37925
    },
    {
      "entropy": 5.2576697826385494,
      "epoch": 4.259897793002752,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033129425463305465,
      "loss": 4.8443,
      "mean_token_accuracy": 0.23394346088171006,
      "num_tokens": 84674874.0,
      "step": 37930
    },
    {
      "entropy": 5.337809324264526,
      "epoch": 4.260459369910709,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003312553819467974,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22733468562364578,
      "num_tokens": 84685619.0,
      "step": 37935
    },
    {
      "entropy": 5.370522689819336,
      "epoch": 4.261020946818666,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033121650746934693,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22852530777454377,
      "num_tokens": 84696062.0,
      "step": 37940
    },
    {
      "entropy": 5.393057537078858,
      "epoch": 4.2615825237266245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003311776312019412,
      "loss": 4.9885,
      "mean_token_accuracy": 0.2198973700404167,
      "num_tokens": 84706686.0,
      "step": 37945
    },
    {
      "entropy": 5.329270553588867,
      "epoch": 4.262144100634582,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003311387531458177,
      "loss": 4.89,
      "mean_token_accuracy": 0.22428309470415114,
      "num_tokens": 84717193.0,
      "step": 37950
    },
    {
      "entropy": 5.308779287338257,
      "epoch": 4.26270567754254,
      "grad_norm": 1.0,
      "learning_rate": 0.0003310998733022146,
      "loss": 4.8346,
      "mean_token_accuracy": 0.2235088363289833,
      "num_tokens": 84728906.0,
      "step": 37955
    },
    {
      "entropy": 5.446051692962646,
      "epoch": 4.263267254450497,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003310609916723698,
      "loss": 4.9886,
      "mean_token_accuracy": 0.22709652930498123,
      "num_tokens": 84741417.0,
      "step": 37960
    },
    {
      "entropy": 5.2874627113342285,
      "epoch": 4.263828831358454,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003310221082575213,
      "loss": 4.8187,
      "mean_token_accuracy": 0.23205751776695252,
      "num_tokens": 84752488.0,
      "step": 37965
    },
    {
      "entropy": 5.385778188705444,
      "epoch": 4.264390408266412,
      "grad_norm": 1.0,
      "learning_rate": 0.00033098322305890715,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2266583561897278,
      "num_tokens": 84763573.0,
      "step": 37970
    },
    {
      "entropy": 5.400099658966065,
      "epoch": 4.26495198517437,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033094433607776545,
      "loss": 5.0136,
      "mean_token_accuracy": 0.21979268193244933,
      "num_tokens": 84774803.0,
      "step": 37975
    },
    {
      "entropy": 5.369838762283325,
      "epoch": 4.265513562082327,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033090544731533434,
      "loss": 4.8345,
      "mean_token_accuracy": 0.22289749681949617,
      "num_tokens": 84786261.0,
      "step": 37980
    },
    {
      "entropy": 5.269379758834839,
      "epoch": 4.266075138990285,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003308665567728522,
      "loss": 4.8426,
      "mean_token_accuracy": 0.229385207593441,
      "num_tokens": 84797575.0,
      "step": 37985
    },
    {
      "entropy": 5.311602401733398,
      "epoch": 4.266636715898242,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00033082766445155715,
      "loss": 5.0001,
      "mean_token_accuracy": 0.22648786753416061,
      "num_tokens": 84811127.0,
      "step": 37990
    },
    {
      "entropy": 5.2997321605682375,
      "epoch": 4.2671982928062,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00033078877035268766,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22603930979967118,
      "num_tokens": 84823169.0,
      "step": 37995
    },
    {
      "entropy": 5.33558931350708,
      "epoch": 4.267759869714157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033074987447748205,
      "loss": 4.9431,
      "mean_token_accuracy": 0.2223045989871025,
      "num_tokens": 84834917.0,
      "step": 38000
    },
    {
      "entropy": 5.382665920257568,
      "epoch": 4.268321446622115,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033071097682717883,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2227035790681839,
      "num_tokens": 84845319.0,
      "step": 38005
    },
    {
      "entropy": 5.350390100479126,
      "epoch": 4.268883023530073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003306720774030165,
      "loss": 4.9268,
      "mean_token_accuracy": 0.23035207092761995,
      "num_tokens": 84855438.0,
      "step": 38010
    },
    {
      "entropy": 5.2158020496368405,
      "epoch": 4.26944460043803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003306331762062336,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22204643189907075,
      "num_tokens": 84867177.0,
      "step": 38015
    },
    {
      "entropy": 5.2672257900238035,
      "epoch": 4.270006177345987,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003305942732380689,
      "loss": 4.8679,
      "mean_token_accuracy": 0.23475642800331115,
      "num_tokens": 84878382.0,
      "step": 38020
    },
    {
      "entropy": 5.311164855957031,
      "epoch": 4.270567754253945,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003305553684997609,
      "loss": 4.8271,
      "mean_token_accuracy": 0.22563968896865844,
      "num_tokens": 84889754.0,
      "step": 38025
    },
    {
      "entropy": 5.3620765686035154,
      "epoch": 4.2711293311619025,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033051646199254843,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22493126690387727,
      "num_tokens": 84901638.0,
      "step": 38030
    },
    {
      "entropy": 5.26508412361145,
      "epoch": 4.271690908069861,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003304775537176702,
      "loss": 4.7822,
      "mean_token_accuracy": 0.23419013172388076,
      "num_tokens": 84912552.0,
      "step": 38035
    },
    {
      "entropy": 5.275631332397461,
      "epoch": 4.272252484977818,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003304386436763652,
      "loss": 4.8448,
      "mean_token_accuracy": 0.22976761311292648,
      "num_tokens": 84924920.0,
      "step": 38040
    },
    {
      "entropy": 5.337018918991089,
      "epoch": 4.272814061885775,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003303997318698723,
      "loss": 4.912,
      "mean_token_accuracy": 0.22318752855062485,
      "num_tokens": 84936835.0,
      "step": 38045
    },
    {
      "entropy": 5.367349576950073,
      "epoch": 4.273375638793733,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003303608182994304,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22937808483839034,
      "num_tokens": 84948307.0,
      "step": 38050
    },
    {
      "entropy": 5.406091833114624,
      "epoch": 4.27393721570169,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003303219029662787,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21734505295753478,
      "num_tokens": 84960034.0,
      "step": 38055
    },
    {
      "entropy": 5.333178186416626,
      "epoch": 4.274498792609648,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033028298587165606,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22558943480253218,
      "num_tokens": 84971195.0,
      "step": 38060
    },
    {
      "entropy": 5.367589330673217,
      "epoch": 4.275060369517606,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003302440670168017,
      "loss": 4.9087,
      "mean_token_accuracy": 0.23038232773542405,
      "num_tokens": 84983689.0,
      "step": 38065
    },
    {
      "entropy": 5.357565546035767,
      "epoch": 4.275621946425563,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003302051464029548,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22004874050617218,
      "num_tokens": 84996469.0,
      "step": 38070
    },
    {
      "entropy": 5.262121772766113,
      "epoch": 4.27618352333352,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033016622403135473,
      "loss": 4.7432,
      "mean_token_accuracy": 0.23716548085212708,
      "num_tokens": 85007427.0,
      "step": 38075
    },
    {
      "entropy": 5.228739786148071,
      "epoch": 4.276745100241478,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033012729990324063,
      "loss": 4.8293,
      "mean_token_accuracy": 0.22919696867465972,
      "num_tokens": 85019089.0,
      "step": 38080
    },
    {
      "entropy": 5.3447465896606445,
      "epoch": 4.2773066771494355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003300883740198519,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22521232515573503,
      "num_tokens": 85031032.0,
      "step": 38085
    },
    {
      "entropy": 5.3164464950561525,
      "epoch": 4.277868254057394,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033004944638242805,
      "loss": 4.853,
      "mean_token_accuracy": 0.23376017659902573,
      "num_tokens": 85042002.0,
      "step": 38090
    },
    {
      "entropy": 5.301219749450683,
      "epoch": 4.278429830965351,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003300105169922084,
      "loss": 4.7907,
      "mean_token_accuracy": 0.2315518543124199,
      "num_tokens": 85052844.0,
      "step": 38095
    },
    {
      "entropy": 5.314469671249389,
      "epoch": 4.278991407873308,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003299715858504326,
      "loss": 4.9545,
      "mean_token_accuracy": 0.223158960044384,
      "num_tokens": 85064094.0,
      "step": 38100
    },
    {
      "entropy": 5.396696662902832,
      "epoch": 4.279552984781266,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032993265295834013,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2207372933626175,
      "num_tokens": 85076695.0,
      "step": 38105
    },
    {
      "entropy": 5.326773691177368,
      "epoch": 4.280114561689223,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003298937183171707,
      "loss": 4.8915,
      "mean_token_accuracy": 0.2300471231341362,
      "num_tokens": 85088258.0,
      "step": 38110
    },
    {
      "entropy": 5.322440910339355,
      "epoch": 4.2806761385971805,
      "grad_norm": 1.0,
      "learning_rate": 0.000329854781928164,
      "loss": 4.8362,
      "mean_token_accuracy": 0.22805664390325547,
      "num_tokens": 85098277.0,
      "step": 38115
    },
    {
      "entropy": 5.357497787475586,
      "epoch": 4.281237715505139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003298158437925598,
      "loss": 4.9413,
      "mean_token_accuracy": 0.22389708757400512,
      "num_tokens": 85108242.0,
      "step": 38120
    },
    {
      "entropy": 5.2517787456512455,
      "epoch": 4.281799292413096,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003297769039115978,
      "loss": 4.783,
      "mean_token_accuracy": 0.232975834608078,
      "num_tokens": 85119607.0,
      "step": 38125
    },
    {
      "entropy": 5.356944417953491,
      "epoch": 4.282360869321053,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032973796228651797,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22540232241153718,
      "num_tokens": 85131517.0,
      "step": 38130
    },
    {
      "entropy": 5.422297143936158,
      "epoch": 4.282922446229011,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032969901891856027,
      "loss": 4.9855,
      "mean_token_accuracy": 0.22238709032535553,
      "num_tokens": 85143739.0,
      "step": 38135
    },
    {
      "entropy": 5.444101619720459,
      "epoch": 4.283484023136968,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032966007380896443,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22575369626283645,
      "num_tokens": 85155881.0,
      "step": 38140
    },
    {
      "entropy": 5.297903919219971,
      "epoch": 4.2840456000449265,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003296211269589708,
      "loss": 4.8936,
      "mean_token_accuracy": 0.22601666748523713,
      "num_tokens": 85168709.0,
      "step": 38145
    },
    {
      "entropy": 5.251294994354248,
      "epoch": 4.284607176952884,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003295821783698192,
      "loss": 4.8275,
      "mean_token_accuracy": 0.22873933613300323,
      "num_tokens": 85179877.0,
      "step": 38150
    },
    {
      "entropy": 5.258625221252442,
      "epoch": 4.285168753860841,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032954322804274995,
      "loss": 4.7865,
      "mean_token_accuracy": 0.23179203569889067,
      "num_tokens": 85190425.0,
      "step": 38155
    },
    {
      "entropy": 5.355438947677612,
      "epoch": 4.285730330768799,
      "grad_norm": 1.0,
      "learning_rate": 0.00032950427597900315,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22153533846139908,
      "num_tokens": 85201576.0,
      "step": 38160
    },
    {
      "entropy": 5.293887948989868,
      "epoch": 4.286291907676756,
      "grad_norm": 0.890625,
      "learning_rate": 0.000329465322179819,
      "loss": 4.8474,
      "mean_token_accuracy": 0.22807974815368653,
      "num_tokens": 85212774.0,
      "step": 38165
    },
    {
      "entropy": 5.375597858428955,
      "epoch": 4.2868534845847135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032942636664643794,
      "loss": 4.8647,
      "mean_token_accuracy": 0.22572794556617737,
      "num_tokens": 85224173.0,
      "step": 38170
    },
    {
      "entropy": 5.36904091835022,
      "epoch": 4.287415061492672,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032938740938010024,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2203699216246605,
      "num_tokens": 85234932.0,
      "step": 38175
    },
    {
      "entropy": 5.3150303840637205,
      "epoch": 4.287976638400629,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003293484503820463,
      "loss": 4.929,
      "mean_token_accuracy": 0.22609545290470123,
      "num_tokens": 85245108.0,
      "step": 38180
    },
    {
      "entropy": 5.379029655456543,
      "epoch": 4.288538215308587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032930948965351664,
      "loss": 4.9607,
      "mean_token_accuracy": 0.22703996747732164,
      "num_tokens": 85255782.0,
      "step": 38185
    },
    {
      "entropy": 5.31162166595459,
      "epoch": 4.289099792216544,
      "grad_norm": 1.0,
      "learning_rate": 0.0003292705271957518,
      "loss": 4.823,
      "mean_token_accuracy": 0.22953373640775682,
      "num_tokens": 85266938.0,
      "step": 38190
    },
    {
      "entropy": 5.243238162994385,
      "epoch": 4.289661369124501,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032923156300999223,
      "loss": 4.8131,
      "mean_token_accuracy": 0.22831854224205017,
      "num_tokens": 85278754.0,
      "step": 38195
    },
    {
      "entropy": 5.344998168945312,
      "epoch": 4.2902229460324595,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032919259709747874,
      "loss": 4.9305,
      "mean_token_accuracy": 0.23119351267814636,
      "num_tokens": 85290255.0,
      "step": 38200
    },
    {
      "entropy": 5.314303541183472,
      "epoch": 4.290784522940417,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00032915362945945195,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2188740476965904,
      "num_tokens": 85301390.0,
      "step": 38205
    },
    {
      "entropy": 5.329101467132569,
      "epoch": 4.291346099848374,
      "grad_norm": 1.0,
      "learning_rate": 0.00032911466009715254,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2261006638407707,
      "num_tokens": 85311215.0,
      "step": 38210
    },
    {
      "entropy": 5.38068699836731,
      "epoch": 4.291907676756332,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032907568901182136,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22413080781698227,
      "num_tokens": 85321212.0,
      "step": 38215
    },
    {
      "entropy": 5.426188373565674,
      "epoch": 4.292469253664289,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003290367162046993,
      "loss": 5.01,
      "mean_token_accuracy": 0.2195660650730133,
      "num_tokens": 85331550.0,
      "step": 38220
    },
    {
      "entropy": 5.356154298782348,
      "epoch": 4.293030830572247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032899774167702723,
      "loss": 4.8743,
      "mean_token_accuracy": 0.23053240925073623,
      "num_tokens": 85342659.0,
      "step": 38225
    },
    {
      "entropy": 5.262993383407593,
      "epoch": 4.2935924074802045,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003289587654300461,
      "loss": 4.8188,
      "mean_token_accuracy": 0.2320706367492676,
      "num_tokens": 85353298.0,
      "step": 38230
    },
    {
      "entropy": 5.295417833328247,
      "epoch": 4.294153984388162,
      "grad_norm": 0.890625,
      "learning_rate": 0.000328919787464997,
      "loss": 4.9543,
      "mean_token_accuracy": 0.22089200913906099,
      "num_tokens": 85365274.0,
      "step": 38235
    },
    {
      "entropy": 5.456685018539429,
      "epoch": 4.29471556129612,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000328880807783121,
      "loss": 5.0368,
      "mean_token_accuracy": 0.22182445079088212,
      "num_tokens": 85376572.0,
      "step": 38240
    },
    {
      "entropy": 5.357493925094604,
      "epoch": 4.295277138204077,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032884182638565914,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2247361034154892,
      "num_tokens": 85387940.0,
      "step": 38245
    },
    {
      "entropy": 5.313954496383667,
      "epoch": 4.295838715112034,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032880284327385255,
      "loss": 4.8387,
      "mean_token_accuracy": 0.22735899984836577,
      "num_tokens": 85398048.0,
      "step": 38250
    },
    {
      "entropy": 5.31037278175354,
      "epoch": 4.296400292019992,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003287638584489426,
      "loss": 4.8663,
      "mean_token_accuracy": 0.2277154356241226,
      "num_tokens": 85409535.0,
      "step": 38255
    },
    {
      "entropy": 5.201943397521973,
      "epoch": 4.29696186892795,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003287248719121706,
      "loss": 4.8284,
      "mean_token_accuracy": 0.22540998607873916,
      "num_tokens": 85420720.0,
      "step": 38260
    },
    {
      "entropy": 5.368556022644043,
      "epoch": 4.297523445835907,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032868588366477785,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22406894117593765,
      "num_tokens": 85429943.0,
      "step": 38265
    },
    {
      "entropy": 5.365045738220215,
      "epoch": 4.298085022743865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032864689370800564,
      "loss": 4.9272,
      "mean_token_accuracy": 0.22918977290391923,
      "num_tokens": 85441168.0,
      "step": 38270
    },
    {
      "entropy": 5.392923784255982,
      "epoch": 4.298646599651822,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003286079020430957,
      "loss": 4.9035,
      "mean_token_accuracy": 0.22708001881837844,
      "num_tokens": 85452532.0,
      "step": 38275
    },
    {
      "entropy": 5.371421146392822,
      "epoch": 4.29920817655978,
      "grad_norm": 0.96875,
      "learning_rate": 0.00032856890867128926,
      "loss": 5.0379,
      "mean_token_accuracy": 0.223794025182724,
      "num_tokens": 85466187.0,
      "step": 38280
    },
    {
      "entropy": 5.297076034545898,
      "epoch": 4.2997697534677375,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003285299135938279,
      "loss": 4.8967,
      "mean_token_accuracy": 0.22821110486984253,
      "num_tokens": 85476950.0,
      "step": 38285
    },
    {
      "entropy": 5.339099884033203,
      "epoch": 4.300331330375695,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032849091681195347,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21935181468725204,
      "num_tokens": 85488184.0,
      "step": 38290
    },
    {
      "entropy": 5.340803527832032,
      "epoch": 4.300892907283653,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032845191832690747,
      "loss": 4.8938,
      "mean_token_accuracy": 0.2251333475112915,
      "num_tokens": 85498662.0,
      "step": 38295
    },
    {
      "entropy": 5.342109251022339,
      "epoch": 4.30145448419161,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003284129181399317,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22376884818077086,
      "num_tokens": 85509454.0,
      "step": 38300
    },
    {
      "entropy": 5.405104446411133,
      "epoch": 4.302016061099567,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003283739162522678,
      "loss": 4.979,
      "mean_token_accuracy": 0.22526478618383408,
      "num_tokens": 85520087.0,
      "step": 38305
    },
    {
      "entropy": 5.285982704162597,
      "epoch": 4.302577638007525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003283349126651578,
      "loss": 4.8777,
      "mean_token_accuracy": 0.23185890316963195,
      "num_tokens": 85532765.0,
      "step": 38310
    },
    {
      "entropy": 5.458770895004273,
      "epoch": 4.303139214915483,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032829590737984345,
      "loss": 4.9551,
      "mean_token_accuracy": 0.21847278624773026,
      "num_tokens": 85543863.0,
      "step": 38315
    },
    {
      "entropy": 5.362820100784302,
      "epoch": 4.30370079182344,
      "grad_norm": 1.0,
      "learning_rate": 0.00032825690039756675,
      "loss": 4.913,
      "mean_token_accuracy": 0.22915297746658325,
      "num_tokens": 85554602.0,
      "step": 38320
    },
    {
      "entropy": 5.337579298019409,
      "epoch": 4.304262368731398,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032821789171956966,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22213827073574066,
      "num_tokens": 85566267.0,
      "step": 38325
    },
    {
      "entropy": 5.378640842437744,
      "epoch": 4.304823945639355,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003281788813470943,
      "loss": 4.987,
      "mean_token_accuracy": 0.2185271292924881,
      "num_tokens": 85579042.0,
      "step": 38330
    },
    {
      "entropy": 5.349207019805908,
      "epoch": 4.305385522547313,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003281398692813827,
      "loss": 4.9391,
      "mean_token_accuracy": 0.226966056227684,
      "num_tokens": 85590006.0,
      "step": 38335
    },
    {
      "entropy": 5.373880529403687,
      "epoch": 4.30594709945527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000328100855523677,
      "loss": 4.9322,
      "mean_token_accuracy": 0.2280791461467743,
      "num_tokens": 85600488.0,
      "step": 38340
    },
    {
      "entropy": 5.287370300292968,
      "epoch": 4.306508676363228,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032806184007521955,
      "loss": 4.8451,
      "mean_token_accuracy": 0.2275533676147461,
      "num_tokens": 85613593.0,
      "step": 38345
    },
    {
      "entropy": 5.285519361495972,
      "epoch": 4.307070253271186,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003280228229372525,
      "loss": 4.8372,
      "mean_token_accuracy": 0.23303708881139756,
      "num_tokens": 85623762.0,
      "step": 38350
    },
    {
      "entropy": 5.319044160842895,
      "epoch": 4.307631830179143,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003279838041110182,
      "loss": 4.9245,
      "mean_token_accuracy": 0.225668865442276,
      "num_tokens": 85634747.0,
      "step": 38355
    },
    {
      "entropy": 5.325010538101196,
      "epoch": 4.3081934070871,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032794478359775904,
      "loss": 4.8402,
      "mean_token_accuracy": 0.2271244764328003,
      "num_tokens": 85645268.0,
      "step": 38360
    },
    {
      "entropy": 5.338386821746826,
      "epoch": 4.308754983995058,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032790576139871743,
      "loss": 4.9521,
      "mean_token_accuracy": 0.21890937387943268,
      "num_tokens": 85656839.0,
      "step": 38365
    },
    {
      "entropy": 5.292079257965088,
      "epoch": 4.3093165609030155,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003278667375151359,
      "loss": 4.8487,
      "mean_token_accuracy": 0.23182503283023834,
      "num_tokens": 85667123.0,
      "step": 38370
    },
    {
      "entropy": 5.35593843460083,
      "epoch": 4.309878137810974,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032782771194825685,
      "loss": 4.9693,
      "mean_token_accuracy": 0.21988319009542465,
      "num_tokens": 85679312.0,
      "step": 38375
    },
    {
      "entropy": 5.261866188049316,
      "epoch": 4.310439714718931,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032778868469932304,
      "loss": 4.8432,
      "mean_token_accuracy": 0.22844273298978807,
      "num_tokens": 85689756.0,
      "step": 38380
    },
    {
      "entropy": 5.446440505981445,
      "epoch": 4.311001291626888,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032774965576957705,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21857472360134125,
      "num_tokens": 85701541.0,
      "step": 38385
    },
    {
      "entropy": 5.348806381225586,
      "epoch": 4.311562868534846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003277106251602615,
      "loss": 5.0046,
      "mean_token_accuracy": 0.22378917634487153,
      "num_tokens": 85712902.0,
      "step": 38390
    },
    {
      "entropy": 5.351965475082397,
      "epoch": 4.312124445442803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032767159287261914,
      "loss": 4.8883,
      "mean_token_accuracy": 0.22622557878494262,
      "num_tokens": 85724226.0,
      "step": 38395
    },
    {
      "entropy": 5.299744081497193,
      "epoch": 4.312686022350761,
      "grad_norm": 0.984375,
      "learning_rate": 0.000327632558907893,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22080124616622926,
      "num_tokens": 85734810.0,
      "step": 38400
    },
    {
      "entropy": 5.313776111602783,
      "epoch": 4.313247599258719,
      "grad_norm": 1.0,
      "learning_rate": 0.00032759352326732573,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22348308712244033,
      "num_tokens": 85745548.0,
      "step": 38405
    },
    {
      "entropy": 5.394666290283203,
      "epoch": 4.313809176166676,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032755448595216027,
      "loss": 4.9758,
      "mean_token_accuracy": 0.22308102548122405,
      "num_tokens": 85757803.0,
      "step": 38410
    },
    {
      "entropy": 5.35051908493042,
      "epoch": 4.314370753074634,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00032751544696363945,
      "loss": 4.914,
      "mean_token_accuracy": 0.2257615104317665,
      "num_tokens": 85770252.0,
      "step": 38415
    },
    {
      "entropy": 5.396606349945069,
      "epoch": 4.314932329982591,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032747640630300664,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21962882578372955,
      "num_tokens": 85782605.0,
      "step": 38420
    },
    {
      "entropy": 5.310629987716675,
      "epoch": 4.3154939068905485,
      "grad_norm": 1.0,
      "learning_rate": 0.00032743736397150455,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2299838036298752,
      "num_tokens": 85792921.0,
      "step": 38425
    },
    {
      "entropy": 5.344047737121582,
      "epoch": 4.316055483798507,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003273983199703765,
      "loss": 4.8505,
      "mean_token_accuracy": 0.22579520493745803,
      "num_tokens": 85804364.0,
      "step": 38430
    },
    {
      "entropy": 5.364271926879883,
      "epoch": 4.316617060706464,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003273592743008656,
      "loss": 5.0011,
      "mean_token_accuracy": 0.22640925794839858,
      "num_tokens": 85815299.0,
      "step": 38435
    },
    {
      "entropy": 5.34661602973938,
      "epoch": 4.317178637614421,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032732022696421514,
      "loss": 4.9555,
      "mean_token_accuracy": 0.22503656446933745,
      "num_tokens": 85828273.0,
      "step": 38440
    },
    {
      "entropy": 5.315956878662109,
      "epoch": 4.317740214522379,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003272811779616684,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22146780639886857,
      "num_tokens": 85837813.0,
      "step": 38445
    },
    {
      "entropy": 5.298606586456299,
      "epoch": 4.318301791430336,
      "grad_norm": 1.0,
      "learning_rate": 0.00032724212729446864,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22222209423780442,
      "num_tokens": 85849585.0,
      "step": 38450
    },
    {
      "entropy": 5.249256896972656,
      "epoch": 4.3188633683382935,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003272030749638593,
      "loss": 4.7754,
      "mean_token_accuracy": 0.23343920558691025,
      "num_tokens": 85861519.0,
      "step": 38455
    },
    {
      "entropy": 5.3257821559906,
      "epoch": 4.319424945246252,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003271640209710837,
      "loss": 4.8601,
      "mean_token_accuracy": 0.23442934155464173,
      "num_tokens": 85871936.0,
      "step": 38460
    },
    {
      "entropy": 5.329644632339478,
      "epoch": 4.319986522154209,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003271249653173855,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22641059756278992,
      "num_tokens": 85883433.0,
      "step": 38465
    },
    {
      "entropy": 5.324489879608154,
      "epoch": 4.320548099062167,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032708590800400826,
      "loss": 4.8768,
      "mean_token_accuracy": 0.22415879219770432,
      "num_tokens": 85894852.0,
      "step": 38470
    },
    {
      "entropy": 5.356907320022583,
      "epoch": 4.321109675970124,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003270468490321955,
      "loss": 4.8912,
      "mean_token_accuracy": 0.23001666069030763,
      "num_tokens": 85905619.0,
      "step": 38475
    },
    {
      "entropy": 5.365391063690185,
      "epoch": 4.321671252878081,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032700778840319086,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2249777540564537,
      "num_tokens": 85916064.0,
      "step": 38480
    },
    {
      "entropy": 5.343345308303833,
      "epoch": 4.3222328297860395,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000326968726118238,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22919786870479583,
      "num_tokens": 85927557.0,
      "step": 38485
    },
    {
      "entropy": 5.362963438034058,
      "epoch": 4.322794406693997,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032692966217858074,
      "loss": 4.88,
      "mean_token_accuracy": 0.23261410295963286,
      "num_tokens": 85937628.0,
      "step": 38490
    },
    {
      "entropy": 5.235645294189453,
      "epoch": 4.323355983601954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003268905965854629,
      "loss": 4.8157,
      "mean_token_accuracy": 0.23260591626167298,
      "num_tokens": 85949221.0,
      "step": 38495
    },
    {
      "entropy": 5.356785869598388,
      "epoch": 4.323917560509912,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032685152934012844,
      "loss": 4.923,
      "mean_token_accuracy": 0.22142242640256882,
      "num_tokens": 85960121.0,
      "step": 38500
    },
    {
      "entropy": 5.382949495315552,
      "epoch": 4.324479137417869,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003268124604438211,
      "loss": 4.9679,
      "mean_token_accuracy": 0.21975530087947845,
      "num_tokens": 85971184.0,
      "step": 38505
    },
    {
      "entropy": 5.319850873947144,
      "epoch": 4.3250407143258265,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032677338989778494,
      "loss": 4.8172,
      "mean_token_accuracy": 0.23732099384069444,
      "num_tokens": 85981324.0,
      "step": 38510
    },
    {
      "entropy": 5.278362464904785,
      "epoch": 4.325602291233785,
      "grad_norm": 0.9375,
      "learning_rate": 0.000326734317703264,
      "loss": 4.817,
      "mean_token_accuracy": 0.23807937502861024,
      "num_tokens": 85993723.0,
      "step": 38515
    },
    {
      "entropy": 5.271831226348877,
      "epoch": 4.326163868141742,
      "grad_norm": 1.0,
      "learning_rate": 0.0003266952438615023,
      "loss": 4.8248,
      "mean_token_accuracy": 0.22694857865571977,
      "num_tokens": 86004031.0,
      "step": 38520
    },
    {
      "entropy": 5.388541412353516,
      "epoch": 4.3267254450497,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032665616837374397,
      "loss": 5.0351,
      "mean_token_accuracy": 0.21934666037559508,
      "num_tokens": 86017161.0,
      "step": 38525
    },
    {
      "entropy": 5.316774177551269,
      "epoch": 4.327287021957657,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003266170912412332,
      "loss": 4.8416,
      "mean_token_accuracy": 0.2249375343322754,
      "num_tokens": 86028369.0,
      "step": 38530
    },
    {
      "entropy": 5.3042213916778564,
      "epoch": 4.327848598865614,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003265780124652143,
      "loss": 4.8529,
      "mean_token_accuracy": 0.23284194767475128,
      "num_tokens": 86039709.0,
      "step": 38535
    },
    {
      "entropy": 5.249087762832642,
      "epoch": 4.3284101757735725,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032653893204693136,
      "loss": 4.8095,
      "mean_token_accuracy": 0.24151951670646668,
      "num_tokens": 86050513.0,
      "step": 38540
    },
    {
      "entropy": 5.35543155670166,
      "epoch": 4.32897175268153,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003264998499876289,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22561756521463394,
      "num_tokens": 86062519.0,
      "step": 38545
    },
    {
      "entropy": 5.450119304656982,
      "epoch": 4.329533329589487,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003264607662885512,
      "loss": 5.0526,
      "mean_token_accuracy": 0.2150881379842758,
      "num_tokens": 86073894.0,
      "step": 38550
    },
    {
      "entropy": 5.380965662002564,
      "epoch": 4.330094906497445,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032642168095094274,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21729159653186797,
      "num_tokens": 86084887.0,
      "step": 38555
    },
    {
      "entropy": 5.381901454925537,
      "epoch": 4.330656483405402,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003263825939760481,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22444397509098052,
      "num_tokens": 86094637.0,
      "step": 38560
    },
    {
      "entropy": 5.293575143814087,
      "epoch": 4.33121806031336,
      "grad_norm": 1.0,
      "learning_rate": 0.00032634350536511167,
      "loss": 4.8045,
      "mean_token_accuracy": 0.2303754836320877,
      "num_tokens": 86105044.0,
      "step": 38565
    },
    {
      "entropy": 5.354091835021973,
      "epoch": 4.3317796372213175,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032630441511937813,
      "loss": 4.9444,
      "mean_token_accuracy": 0.22287405729293824,
      "num_tokens": 86116481.0,
      "step": 38570
    },
    {
      "entropy": 5.245355463027954,
      "epoch": 4.332341214129275,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032626532324009204,
      "loss": 4.7903,
      "mean_token_accuracy": 0.2403682827949524,
      "num_tokens": 86126906.0,
      "step": 38575
    },
    {
      "entropy": 5.3704053401947025,
      "epoch": 4.332902791037233,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003262262297284983,
      "loss": 5.0319,
      "mean_token_accuracy": 0.21583458930253982,
      "num_tokens": 86137899.0,
      "step": 38580
    },
    {
      "entropy": 5.449641227722168,
      "epoch": 4.33346436794519,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00032618713458584147,
      "loss": 4.97,
      "mean_token_accuracy": 0.22467987686395646,
      "num_tokens": 86149283.0,
      "step": 38585
    },
    {
      "entropy": 5.285018587112427,
      "epoch": 4.334025944853147,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003261480378133665,
      "loss": 4.7993,
      "mean_token_accuracy": 0.24144284278154374,
      "num_tokens": 86160239.0,
      "step": 38590
    },
    {
      "entropy": 5.301095247268677,
      "epoch": 4.334587521761105,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003261089394123181,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22062250971794128,
      "num_tokens": 86171924.0,
      "step": 38595
    },
    {
      "entropy": 5.373297262191772,
      "epoch": 4.335149098669063,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003260698393839413,
      "loss": 4.873,
      "mean_token_accuracy": 0.22364883422851561,
      "num_tokens": 86182556.0,
      "step": 38600
    },
    {
      "entropy": 5.352274608612061,
      "epoch": 4.335710675577021,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00032603073772948096,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22735241204500198,
      "num_tokens": 86193500.0,
      "step": 38605
    },
    {
      "entropy": 5.234258651733398,
      "epoch": 4.336272252484978,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032599163445018204,
      "loss": 4.8238,
      "mean_token_accuracy": 0.226099456846714,
      "num_tokens": 86205530.0,
      "step": 38610
    },
    {
      "entropy": 5.373866033554077,
      "epoch": 4.336833829392935,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003259525295472899,
      "loss": 4.9424,
      "mean_token_accuracy": 0.21896052211523057,
      "num_tokens": 86215538.0,
      "step": 38615
    },
    {
      "entropy": 5.313181447982788,
      "epoch": 4.337395406300893,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00032591342302204935,
      "loss": 4.796,
      "mean_token_accuracy": 0.22869570702314376,
      "num_tokens": 86226681.0,
      "step": 38620
    },
    {
      "entropy": 5.334412908554077,
      "epoch": 4.3379569832088505,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003258743148757057,
      "loss": 4.8621,
      "mean_token_accuracy": 0.2180955156683922,
      "num_tokens": 86237629.0,
      "step": 38625
    },
    {
      "entropy": 5.36294903755188,
      "epoch": 4.338518560116808,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032583520510950405,
      "loss": 4.9879,
      "mean_token_accuracy": 0.21798202395439148,
      "num_tokens": 86248466.0,
      "step": 38630
    },
    {
      "entropy": 5.329041624069214,
      "epoch": 4.339080137024766,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032579609372468984,
      "loss": 4.9528,
      "mean_token_accuracy": 0.21855527460575103,
      "num_tokens": 86259845.0,
      "step": 38635
    },
    {
      "entropy": 5.436394119262696,
      "epoch": 4.339641713932723,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032575698072250827,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22842836380004883,
      "num_tokens": 86269807.0,
      "step": 38640
    },
    {
      "entropy": 5.3634124279022215,
      "epoch": 4.34020329084068,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003257178661042047,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22558883130550383,
      "num_tokens": 86280095.0,
      "step": 38645
    },
    {
      "entropy": 5.27594780921936,
      "epoch": 4.340764867748638,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032567874987102463,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23466363400220872,
      "num_tokens": 86291842.0,
      "step": 38650
    },
    {
      "entropy": 5.446324586868286,
      "epoch": 4.341326444656596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003256396320242136,
      "loss": 5.0286,
      "mean_token_accuracy": 0.21563541144132614,
      "num_tokens": 86303080.0,
      "step": 38655
    },
    {
      "entropy": 5.3795067310333256,
      "epoch": 4.341888021564554,
      "grad_norm": 0.890625,
      "learning_rate": 0.000325600512565017,
      "loss": 4.8889,
      "mean_token_accuracy": 0.22431754171848298,
      "num_tokens": 86315378.0,
      "step": 38660
    },
    {
      "entropy": 5.384488582611084,
      "epoch": 4.342449598472511,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00032556139149468046,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2233425959944725,
      "num_tokens": 86327823.0,
      "step": 38665
    },
    {
      "entropy": 5.325388097763062,
      "epoch": 4.343011175380468,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032552226881444956,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22577030211687088,
      "num_tokens": 86338312.0,
      "step": 38670
    },
    {
      "entropy": 5.270217323303223,
      "epoch": 4.343572752288426,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032548314452557006,
      "loss": 4.7724,
      "mean_token_accuracy": 0.23336310982704161,
      "num_tokens": 86348157.0,
      "step": 38675
    },
    {
      "entropy": 5.217327451705932,
      "epoch": 4.344134329196383,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032544401862928766,
      "loss": 4.815,
      "mean_token_accuracy": 0.2329142302274704,
      "num_tokens": 86359131.0,
      "step": 38680
    },
    {
      "entropy": 5.314981603622437,
      "epoch": 4.344695906104341,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032540489112684804,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22776159197092055,
      "num_tokens": 86370953.0,
      "step": 38685
    },
    {
      "entropy": 5.307064533233643,
      "epoch": 4.345257483012299,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032536576201949724,
      "loss": 4.8274,
      "mean_token_accuracy": 0.23355590105056762,
      "num_tokens": 86382445.0,
      "step": 38690
    },
    {
      "entropy": 5.311904764175415,
      "epoch": 4.345819059920256,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003253266313084811,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22423868626356125,
      "num_tokens": 86393177.0,
      "step": 38695
    },
    {
      "entropy": 5.2942300796508786,
      "epoch": 4.346380636828213,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003252874989950454,
      "loss": 4.8718,
      "mean_token_accuracy": 0.23282826095819473,
      "num_tokens": 86405535.0,
      "step": 38700
    },
    {
      "entropy": 5.246093606948852,
      "epoch": 4.346942213736171,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032524836508043625,
      "loss": 4.8021,
      "mean_token_accuracy": 0.23416534662246705,
      "num_tokens": 86415980.0,
      "step": 38705
    },
    {
      "entropy": 5.332367563247681,
      "epoch": 4.3475037906441285,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032520922956589973,
      "loss": 4.9469,
      "mean_token_accuracy": 0.23253892362117767,
      "num_tokens": 86427590.0,
      "step": 38710
    },
    {
      "entropy": 5.364048814773559,
      "epoch": 4.348065367552087,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032517009245268175,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2234596401453018,
      "num_tokens": 86438358.0,
      "step": 38715
    },
    {
      "entropy": 5.268425416946411,
      "epoch": 4.348626944460044,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032513095374202863,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23703001588582992,
      "num_tokens": 86448894.0,
      "step": 38720
    },
    {
      "entropy": 5.283785200119018,
      "epoch": 4.349188521368001,
      "grad_norm": 1.0,
      "learning_rate": 0.00032509181343518646,
      "loss": 4.9069,
      "mean_token_accuracy": 0.23675814867019654,
      "num_tokens": 86461617.0,
      "step": 38725
    },
    {
      "entropy": 5.362652397155761,
      "epoch": 4.349750098275959,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00032505267153340163,
      "loss": 4.9612,
      "mean_token_accuracy": 0.22062282115221024,
      "num_tokens": 86473534.0,
      "step": 38730
    },
    {
      "entropy": 5.397263050079346,
      "epoch": 4.350311675183916,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003250135280379202,
      "loss": 5.0018,
      "mean_token_accuracy": 0.22024556398391723,
      "num_tokens": 86486561.0,
      "step": 38735
    },
    {
      "entropy": 5.285743856430054,
      "epoch": 4.350873252091874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003249743829499888,
      "loss": 4.8559,
      "mean_token_accuracy": 0.2369164079427719,
      "num_tokens": 86499368.0,
      "step": 38740
    },
    {
      "entropy": 5.239845895767212,
      "epoch": 4.351434828999832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003249352362708535,
      "loss": 4.8599,
      "mean_token_accuracy": 0.2328908398747444,
      "num_tokens": 86509798.0,
      "step": 38745
    },
    {
      "entropy": 5.374415063858033,
      "epoch": 4.351996405907789,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000324896088001761,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22346590906381608,
      "num_tokens": 86521372.0,
      "step": 38750
    },
    {
      "entropy": 5.385874605178833,
      "epoch": 4.352557982815747,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003248569381439577,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2161380797624588,
      "num_tokens": 86531889.0,
      "step": 38755
    },
    {
      "entropy": 5.2649860858917235,
      "epoch": 4.353119559723704,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003248177866986901,
      "loss": 4.8059,
      "mean_token_accuracy": 0.22731684595346452,
      "num_tokens": 86543456.0,
      "step": 38760
    },
    {
      "entropy": 5.2793745517730715,
      "epoch": 4.3536811366316615,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003247786336672049,
      "loss": 4.845,
      "mean_token_accuracy": 0.2290216490626335,
      "num_tokens": 86554681.0,
      "step": 38765
    },
    {
      "entropy": 5.364985656738281,
      "epoch": 4.35424271353962,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00032473947905074866,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22968198210000992,
      "num_tokens": 86567050.0,
      "step": 38770
    },
    {
      "entropy": 5.333202981948853,
      "epoch": 4.354804290447577,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00032470032285056814,
      "loss": 4.852,
      "mean_token_accuracy": 0.23345038443803787,
      "num_tokens": 86580172.0,
      "step": 38775
    },
    {
      "entropy": 5.363266468048096,
      "epoch": 4.355365867355534,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032466116506791004,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22590737789869308,
      "num_tokens": 86592389.0,
      "step": 38780
    },
    {
      "entropy": 5.281791830062867,
      "epoch": 4.355927444263492,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00032462200570402124,
      "loss": 4.8655,
      "mean_token_accuracy": 0.2263782188296318,
      "num_tokens": 86603668.0,
      "step": 38785
    },
    {
      "entropy": 5.308871221542359,
      "epoch": 4.356489021171449,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003245828447601484,
      "loss": 4.8761,
      "mean_token_accuracy": 0.22213534116744996,
      "num_tokens": 86614032.0,
      "step": 38790
    },
    {
      "entropy": 5.35528793334961,
      "epoch": 4.357050598079407,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003245436822375387,
      "loss": 4.906,
      "mean_token_accuracy": 0.23005196452140808,
      "num_tokens": 86624559.0,
      "step": 38795
    },
    {
      "entropy": 5.293209838867187,
      "epoch": 4.357612174987365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003245045181374389,
      "loss": 4.8239,
      "mean_token_accuracy": 0.22908556014299392,
      "num_tokens": 86635557.0,
      "step": 38800
    },
    {
      "entropy": 5.340503120422364,
      "epoch": 4.358173751895322,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000324465352461096,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22909242063760757,
      "num_tokens": 86647055.0,
      "step": 38805
    },
    {
      "entropy": 5.3156288146972654,
      "epoch": 4.35873532880328,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003244261852097572,
      "loss": 4.8687,
      "mean_token_accuracy": 0.2345954433083534,
      "num_tokens": 86659079.0,
      "step": 38810
    },
    {
      "entropy": 5.323622417449951,
      "epoch": 4.359296905711237,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032438701638466943,
      "loss": 4.8844,
      "mean_token_accuracy": 0.2242441564798355,
      "num_tokens": 86670671.0,
      "step": 38815
    },
    {
      "entropy": 5.346674585342408,
      "epoch": 4.359858482619194,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00032434784598707987,
      "loss": 4.9916,
      "mean_token_accuracy": 0.22346266210079194,
      "num_tokens": 86682299.0,
      "step": 38820
    },
    {
      "entropy": 5.353532695770264,
      "epoch": 4.3604200595271525,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003243086740182358,
      "loss": 4.8922,
      "mean_token_accuracy": 0.22214752584695815,
      "num_tokens": 86692319.0,
      "step": 38825
    },
    {
      "entropy": 5.28872447013855,
      "epoch": 4.36098163643511,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032426950047938436,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22552026510238649,
      "num_tokens": 86702683.0,
      "step": 38830
    },
    {
      "entropy": 5.31775426864624,
      "epoch": 4.361543213343067,
      "grad_norm": 1.015625,
      "learning_rate": 0.000324230325371773,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22883417457342148,
      "num_tokens": 86714135.0,
      "step": 38835
    },
    {
      "entropy": 5.259040594100952,
      "epoch": 4.362104790251025,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003241911486966489,
      "loss": 4.803,
      "mean_token_accuracy": 0.2250555783510208,
      "num_tokens": 86725993.0,
      "step": 38840
    },
    {
      "entropy": 5.3532572269439695,
      "epoch": 4.362666367158982,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003241519704552596,
      "loss": 4.9206,
      "mean_token_accuracy": 0.22259171009063722,
      "num_tokens": 86737049.0,
      "step": 38845
    },
    {
      "entropy": 5.31566162109375,
      "epoch": 4.36322794406694,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032411279064885253,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22582069039344788,
      "num_tokens": 86746550.0,
      "step": 38850
    },
    {
      "entropy": 5.208047389984131,
      "epoch": 4.363789520974898,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032407360927867506,
      "loss": 4.7872,
      "mean_token_accuracy": 0.22906892895698547,
      "num_tokens": 86756701.0,
      "step": 38855
    },
    {
      "entropy": 5.32708854675293,
      "epoch": 4.364351097882855,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003240344263459749,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23313011825084687,
      "num_tokens": 86768667.0,
      "step": 38860
    },
    {
      "entropy": 5.3356125831604,
      "epoch": 4.364912674790813,
      "grad_norm": 1.125,
      "learning_rate": 0.00032399524185199957,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2278653249144554,
      "num_tokens": 86779330.0,
      "step": 38865
    },
    {
      "entropy": 5.37472677230835,
      "epoch": 4.36547425169877,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032395605579799673,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22947632819414138,
      "num_tokens": 86790749.0,
      "step": 38870
    },
    {
      "entropy": 5.32660551071167,
      "epoch": 4.366035828606727,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032391686818521414,
      "loss": 4.8223,
      "mean_token_accuracy": 0.2295699506998062,
      "num_tokens": 86802275.0,
      "step": 38875
    },
    {
      "entropy": 5.306037425994873,
      "epoch": 4.3665974055146854,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032387767901489946,
      "loss": 4.9279,
      "mean_token_accuracy": 0.21835035532712938,
      "num_tokens": 86812965.0,
      "step": 38880
    },
    {
      "entropy": 5.351228904724121,
      "epoch": 4.367158982422643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032383848828830054,
      "loss": 4.9546,
      "mean_token_accuracy": 0.23019843250513078,
      "num_tokens": 86823787.0,
      "step": 38885
    },
    {
      "entropy": 5.335306692123413,
      "epoch": 4.3677205593306,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032379929600666525,
      "loss": 4.944,
      "mean_token_accuracy": 0.22074999213218688,
      "num_tokens": 86835173.0,
      "step": 38890
    },
    {
      "entropy": 5.346673011779785,
      "epoch": 4.368282136238558,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032376010217124133,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22981329560279845,
      "num_tokens": 86845080.0,
      "step": 38895
    },
    {
      "entropy": 5.379323053359985,
      "epoch": 4.368843713146515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032372090678327694,
      "loss": 4.9328,
      "mean_token_accuracy": 0.22010405510663986,
      "num_tokens": 86857074.0,
      "step": 38900
    },
    {
      "entropy": 5.369425249099732,
      "epoch": 4.369405290054473,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032368170984401997,
      "loss": 4.9681,
      "mean_token_accuracy": 0.22165321558713913,
      "num_tokens": 86868050.0,
      "step": 38905
    },
    {
      "entropy": 5.278889417648315,
      "epoch": 4.3699668669624305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032364251135471846,
      "loss": 4.8498,
      "mean_token_accuracy": 0.2303059458732605,
      "num_tokens": 86879259.0,
      "step": 38910
    },
    {
      "entropy": 5.301499557495117,
      "epoch": 4.370528443870388,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003236033113166205,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22018381357192993,
      "num_tokens": 86891659.0,
      "step": 38915
    },
    {
      "entropy": 5.320328855514527,
      "epoch": 4.371090020778346,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00032356410973097423,
      "loss": 4.805,
      "mean_token_accuracy": 0.23199607580900192,
      "num_tokens": 86903633.0,
      "step": 38920
    },
    {
      "entropy": 5.297225475311279,
      "epoch": 4.371651597686303,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032352490659902795,
      "loss": 4.7927,
      "mean_token_accuracy": 0.2333680808544159,
      "num_tokens": 86915157.0,
      "step": 38925
    },
    {
      "entropy": 5.302822780609131,
      "epoch": 4.37221317459426,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003234857019220298,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22782742977142334,
      "num_tokens": 86925970.0,
      "step": 38930
    },
    {
      "entropy": 5.281886625289917,
      "epoch": 4.372774751502218,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000323446495701228,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22982731014490126,
      "num_tokens": 86938440.0,
      "step": 38935
    },
    {
      "entropy": 5.407351446151734,
      "epoch": 4.373336328410176,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003234072879378711,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2316542774438858,
      "num_tokens": 86949864.0,
      "step": 38940
    },
    {
      "entropy": 5.262256717681884,
      "epoch": 4.373897905318134,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003233680786332074,
      "loss": 4.8309,
      "mean_token_accuracy": 0.22920748740434646,
      "num_tokens": 86961372.0,
      "step": 38945
    },
    {
      "entropy": 5.227307653427124,
      "epoch": 4.374459482226091,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003233288677884853,
      "loss": 4.781,
      "mean_token_accuracy": 0.23027345091104506,
      "num_tokens": 86970797.0,
      "step": 38950
    },
    {
      "entropy": 5.2705381393432615,
      "epoch": 4.375021059134048,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003232896554049533,
      "loss": 4.8003,
      "mean_token_accuracy": 0.23287141025066377,
      "num_tokens": 86983022.0,
      "step": 38955
    },
    {
      "entropy": 5.297507047653198,
      "epoch": 4.375582636042006,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032325044148385986,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22562686502933502,
      "num_tokens": 86993340.0,
      "step": 38960
    },
    {
      "entropy": 5.503663539886475,
      "epoch": 4.3761442129499635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032321122602645374,
      "loss": 5.1175,
      "mean_token_accuracy": 0.21916889399290085,
      "num_tokens": 87004421.0,
      "step": 38965
    },
    {
      "entropy": 5.438340950012207,
      "epoch": 4.376705789857921,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003231720090339834,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22032207250595093,
      "num_tokens": 87014775.0,
      "step": 38970
    },
    {
      "entropy": 5.3542396068573,
      "epoch": 4.377267366765879,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032313279050769767,
      "loss": 4.949,
      "mean_token_accuracy": 0.22529728561639786,
      "num_tokens": 87026107.0,
      "step": 38975
    },
    {
      "entropy": 5.333130741119385,
      "epoch": 4.377828943673836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003230935704488452,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21735664904117585,
      "num_tokens": 87036386.0,
      "step": 38980
    },
    {
      "entropy": 5.366417694091797,
      "epoch": 4.378390520581794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032305434885867483,
      "loss": 4.8742,
      "mean_token_accuracy": 0.23247405588626863,
      "num_tokens": 87048134.0,
      "step": 38985
    },
    {
      "entropy": 5.347395038604736,
      "epoch": 4.378952097489751,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032301512573843535,
      "loss": 4.997,
      "mean_token_accuracy": 0.22362251281738282,
      "num_tokens": 87059721.0,
      "step": 38990
    },
    {
      "entropy": 5.335381650924683,
      "epoch": 4.3795136743977086,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003229759010893756,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22603311985731125,
      "num_tokens": 87071017.0,
      "step": 38995
    },
    {
      "entropy": 5.353639554977417,
      "epoch": 4.380075251305667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003229366749127446,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22565826177597045,
      "num_tokens": 87081537.0,
      "step": 39000
    },
    {
      "epoch": 4.380075251305667,
      "eval_entropy": 5.155762018479728,
      "eval_loss": 5.069939613342285,
      "eval_mean_token_accuracy": 0.22495746962881777,
      "eval_num_tokens": 87081537.0,
      "eval_runtime": 24.0751,
      "eval_samples_per_second": 1288.176,
      "eval_steps_per_second": 161.038,
      "step": 39000
    },
    {
      "entropy": 5.361503219604492,
      "epoch": 4.380636828213624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003228974472097913,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22871919870376586,
      "num_tokens": 87093056.0,
      "step": 39005
    },
    {
      "entropy": 5.347721767425537,
      "epoch": 4.381198405121581,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032285821798176456,
      "loss": 4.9753,
      "mean_token_accuracy": 0.21977908313274383,
      "num_tokens": 87104582.0,
      "step": 39010
    },
    {
      "entropy": 5.3754418849945065,
      "epoch": 4.381759982029539,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032281898722991376,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21976435631513597,
      "num_tokens": 87116488.0,
      "step": 39015
    },
    {
      "entropy": 5.376815557479858,
      "epoch": 4.382321558937496,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032277975495548775,
      "loss": 4.9344,
      "mean_token_accuracy": 0.22003254890441895,
      "num_tokens": 87126760.0,
      "step": 39020
    },
    {
      "entropy": 5.426122760772705,
      "epoch": 4.382883135845454,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003227405211597359,
      "loss": 4.9669,
      "mean_token_accuracy": 0.22737135142087936,
      "num_tokens": 87139946.0,
      "step": 39025
    },
    {
      "entropy": 5.377615451812744,
      "epoch": 4.383444712753412,
      "grad_norm": 0.890625,
      "learning_rate": 0.00032270128584390735,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2280343383550644,
      "num_tokens": 87150768.0,
      "step": 39030
    },
    {
      "entropy": 5.183257150650024,
      "epoch": 4.384006289661369,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003226620490092513,
      "loss": 4.7821,
      "mean_token_accuracy": 0.22679193764925004,
      "num_tokens": 87163110.0,
      "step": 39035
    },
    {
      "entropy": 5.255275774002075,
      "epoch": 4.384567866569327,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032262281065701716,
      "loss": 4.8469,
      "mean_token_accuracy": 0.22784461975097656,
      "num_tokens": 87174041.0,
      "step": 39040
    },
    {
      "entropy": 5.328845500946045,
      "epoch": 4.385129443477284,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003225835707884542,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22944344729185104,
      "num_tokens": 87185560.0,
      "step": 39045
    },
    {
      "entropy": 5.352369117736816,
      "epoch": 4.3856910203852415,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003225443294048119,
      "loss": 4.9731,
      "mean_token_accuracy": 0.2203712895512581,
      "num_tokens": 87196222.0,
      "step": 39050
    },
    {
      "entropy": 5.323004722595215,
      "epoch": 4.3862525972932,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00032250508650733976,
      "loss": 4.8236,
      "mean_token_accuracy": 0.2306432083249092,
      "num_tokens": 87206716.0,
      "step": 39055
    },
    {
      "entropy": 5.284487581253051,
      "epoch": 4.386814174201157,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003224658420972871,
      "loss": 4.8088,
      "mean_token_accuracy": 0.2402772292494774,
      "num_tokens": 87217291.0,
      "step": 39060
    },
    {
      "entropy": 5.252010250091553,
      "epoch": 4.387375751109114,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003224265961759038,
      "loss": 4.924,
      "mean_token_accuracy": 0.22360581457614898,
      "num_tokens": 87227058.0,
      "step": 39065
    },
    {
      "entropy": 5.339927816390992,
      "epoch": 4.387937328017072,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003223873487444392,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21583888530731202,
      "num_tokens": 87238426.0,
      "step": 39070
    },
    {
      "entropy": 5.348440027236938,
      "epoch": 4.388498904925029,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003223480998041431,
      "loss": 4.8647,
      "mean_token_accuracy": 0.2268187001347542,
      "num_tokens": 87248801.0,
      "step": 39075
    },
    {
      "entropy": 5.298681354522705,
      "epoch": 4.389060481832987,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032230884935626505,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22873377054929733,
      "num_tokens": 87259391.0,
      "step": 39080
    },
    {
      "entropy": 5.378477716445923,
      "epoch": 4.389622058740945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000322269597402055,
      "loss": 4.8994,
      "mean_token_accuracy": 0.23123930096626283,
      "num_tokens": 87270405.0,
      "step": 39085
    },
    {
      "entropy": 5.380262088775635,
      "epoch": 4.390183635648902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003222303439427625,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22137229442596434,
      "num_tokens": 87281436.0,
      "step": 39090
    },
    {
      "entropy": 5.2122584819793705,
      "epoch": 4.39074521255686,
      "grad_norm": 1.0,
      "learning_rate": 0.00032219108897963767,
      "loss": 4.7607,
      "mean_token_accuracy": 0.22828382551670073,
      "num_tokens": 87292188.0,
      "step": 39095
    },
    {
      "entropy": 5.284840488433838,
      "epoch": 4.391306789464817,
      "grad_norm": 1.0,
      "learning_rate": 0.0003221518325139302,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22439608573913575,
      "num_tokens": 87303059.0,
      "step": 39100
    },
    {
      "entropy": 5.349942445755005,
      "epoch": 4.3918683663727744,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003221125745468901,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23266898095607758,
      "num_tokens": 87314656.0,
      "step": 39105
    },
    {
      "entropy": 5.374396848678589,
      "epoch": 4.3924299432807326,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003220733150797673,
      "loss": 5.0281,
      "mean_token_accuracy": 0.215499247610569,
      "num_tokens": 87326929.0,
      "step": 39110
    },
    {
      "entropy": 5.392226982116699,
      "epoch": 4.39299152018869,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003220340541138119,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22224564701318741,
      "num_tokens": 87338687.0,
      "step": 39115
    },
    {
      "entropy": 5.3941137313842775,
      "epoch": 4.393553097096647,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032199479165027407,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22936271131038666,
      "num_tokens": 87350374.0,
      "step": 39120
    },
    {
      "entropy": 5.335529756546021,
      "epoch": 4.394114674004605,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032195552769040376,
      "loss": 5.0095,
      "mean_token_accuracy": 0.2218741998076439,
      "num_tokens": 87362473.0,
      "step": 39125
    },
    {
      "entropy": 5.208827924728394,
      "epoch": 4.394676250912562,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003219162622354513,
      "loss": 4.7948,
      "mean_token_accuracy": 0.23498421162366867,
      "num_tokens": 87372959.0,
      "step": 39130
    },
    {
      "entropy": 5.349884748458862,
      "epoch": 4.39523782782052,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032187699528666674,
      "loss": 4.8581,
      "mean_token_accuracy": 0.22999487072229385,
      "num_tokens": 87383902.0,
      "step": 39135
    },
    {
      "entropy": 5.350708675384522,
      "epoch": 4.395799404728478,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003218377268453006,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22047436386346816,
      "num_tokens": 87395390.0,
      "step": 39140
    },
    {
      "entropy": 5.339994096755982,
      "epoch": 4.396360981636435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000321798456912603,
      "loss": 4.9104,
      "mean_token_accuracy": 0.22830140292644502,
      "num_tokens": 87406643.0,
      "step": 39145
    },
    {
      "entropy": 5.497971487045288,
      "epoch": 4.396922558544393,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032175918548982437,
      "loss": 5.0468,
      "mean_token_accuracy": 0.21557883620262147,
      "num_tokens": 87417575.0,
      "step": 39150
    },
    {
      "entropy": 5.356855678558349,
      "epoch": 4.39748413545235,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032171991257821513,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22669327557086943,
      "num_tokens": 87428566.0,
      "step": 39155
    },
    {
      "entropy": 5.280314683914185,
      "epoch": 4.398045712360307,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003216806381790258,
      "loss": 4.9105,
      "mean_token_accuracy": 0.23277091681957246,
      "num_tokens": 87440234.0,
      "step": 39160
    },
    {
      "entropy": 5.3343640804290775,
      "epoch": 4.3986072892682655,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003216413622935069,
      "loss": 4.9707,
      "mean_token_accuracy": 0.22954027354717255,
      "num_tokens": 87451546.0,
      "step": 39165
    },
    {
      "entropy": 5.369922828674317,
      "epoch": 4.399168866176223,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00032160208492290886,
      "loss": 4.9153,
      "mean_token_accuracy": 0.22718994617462157,
      "num_tokens": 87462385.0,
      "step": 39170
    },
    {
      "entropy": 5.3728584289550785,
      "epoch": 4.399730443084181,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00032156280606848237,
      "loss": 4.8914,
      "mean_token_accuracy": 0.2203691229224205,
      "num_tokens": 87474378.0,
      "step": 39175
    },
    {
      "entropy": 5.291336917877198,
      "epoch": 4.400292019992138,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003215235257314781,
      "loss": 4.9115,
      "mean_token_accuracy": 0.22477633506059647,
      "num_tokens": 87485363.0,
      "step": 39180
    },
    {
      "entropy": 5.266971397399902,
      "epoch": 4.400853596900095,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032148424391314665,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22167600244283675,
      "num_tokens": 87496026.0,
      "step": 39185
    },
    {
      "entropy": 5.455640506744385,
      "epoch": 4.401415173808053,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032144496061473895,
      "loss": 5.0537,
      "mean_token_accuracy": 0.21666980981826783,
      "num_tokens": 87508017.0,
      "step": 39190
    },
    {
      "entropy": 5.33986759185791,
      "epoch": 4.401976750716011,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003214056758375056,
      "loss": 4.8514,
      "mean_token_accuracy": 0.22708531171083451,
      "num_tokens": 87519195.0,
      "step": 39195
    },
    {
      "entropy": 5.280864000320435,
      "epoch": 4.402538327623968,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003213663895826976,
      "loss": 4.8569,
      "mean_token_accuracy": 0.23063672184944153,
      "num_tokens": 87528662.0,
      "step": 39200
    },
    {
      "entropy": 5.328040552139282,
      "epoch": 4.403099904531926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003213271018515657,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2281997412443161,
      "num_tokens": 87539581.0,
      "step": 39205
    },
    {
      "entropy": 5.4320916652679445,
      "epoch": 4.403661481439883,
      "grad_norm": 1.078125,
      "learning_rate": 0.000321287812645361,
      "loss": 4.9928,
      "mean_token_accuracy": 0.22422082126140594,
      "num_tokens": 87550956.0,
      "step": 39210
    },
    {
      "entropy": 5.307030582427979,
      "epoch": 4.40422305834784,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032124852196533434,
      "loss": 4.8254,
      "mean_token_accuracy": 0.23582567721605302,
      "num_tokens": 87561814.0,
      "step": 39215
    },
    {
      "entropy": 5.313479948043823,
      "epoch": 4.404784635255798,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003212092298127368,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22785967886447905,
      "num_tokens": 87571806.0,
      "step": 39220
    },
    {
      "entropy": 5.292556715011597,
      "epoch": 4.405346212163756,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032116993618881945,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22914689630270005,
      "num_tokens": 87584182.0,
      "step": 39225
    },
    {
      "entropy": 5.35007381439209,
      "epoch": 4.405907789071714,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032113064109483345,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2196941003203392,
      "num_tokens": 87595296.0,
      "step": 39230
    },
    {
      "entropy": 5.323171615600586,
      "epoch": 4.406469365979671,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032109134453202993,
      "loss": 4.8582,
      "mean_token_accuracy": 0.2298856258392334,
      "num_tokens": 87606173.0,
      "step": 39235
    },
    {
      "entropy": 5.389149522781372,
      "epoch": 4.407030942887628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032105204650166016,
      "loss": 4.9743,
      "mean_token_accuracy": 0.22090557366609573,
      "num_tokens": 87615973.0,
      "step": 39240
    },
    {
      "entropy": 5.341455793380737,
      "epoch": 4.407592519795586,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032101274700497537,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22391989827156067,
      "num_tokens": 87626850.0,
      "step": 39245
    },
    {
      "entropy": 5.284790182113648,
      "epoch": 4.4081540967035435,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003209734460432269,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22587104439735411,
      "num_tokens": 87639665.0,
      "step": 39250
    },
    {
      "entropy": 5.4052355766296385,
      "epoch": 4.408715673611501,
      "grad_norm": 1.0,
      "learning_rate": 0.000320934143617666,
      "loss": 4.9512,
      "mean_token_accuracy": 0.2241820365190506,
      "num_tokens": 87650822.0,
      "step": 39255
    },
    {
      "entropy": 5.302371215820313,
      "epoch": 4.409277250519459,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003208948397295441,
      "loss": 4.7561,
      "mean_token_accuracy": 0.24074116498231887,
      "num_tokens": 87661132.0,
      "step": 39260
    },
    {
      "entropy": 5.259728670120239,
      "epoch": 4.409838827427416,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032085553438011285,
      "loss": 4.8461,
      "mean_token_accuracy": 0.2278350129723549,
      "num_tokens": 87672596.0,
      "step": 39265
    },
    {
      "entropy": 5.282335233688355,
      "epoch": 4.410400404335373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003208162275706235,
      "loss": 4.8137,
      "mean_token_accuracy": 0.22991986721754074,
      "num_tokens": 87683394.0,
      "step": 39270
    },
    {
      "entropy": 5.320605087280273,
      "epoch": 4.410961981243331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003207769193023277,
      "loss": 4.8552,
      "mean_token_accuracy": 0.2324218764901161,
      "num_tokens": 87694984.0,
      "step": 39275
    },
    {
      "entropy": 5.330983591079712,
      "epoch": 4.411523558151289,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032073760957647705,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22490293234586717,
      "num_tokens": 87705990.0,
      "step": 39280
    },
    {
      "entropy": 5.309467077255249,
      "epoch": 4.412085135059247,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003206982983943232,
      "loss": 4.8405,
      "mean_token_accuracy": 0.22557423114776612,
      "num_tokens": 87717329.0,
      "step": 39285
    },
    {
      "entropy": 5.207985591888428,
      "epoch": 4.412646711967204,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003206589857571178,
      "loss": 4.8509,
      "mean_token_accuracy": 0.22804830223321915,
      "num_tokens": 87728436.0,
      "step": 39290
    },
    {
      "entropy": 5.285052871704101,
      "epoch": 4.413208288875161,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032061967166611255,
      "loss": 4.8785,
      "mean_token_accuracy": 0.22850220948457717,
      "num_tokens": 87738803.0,
      "step": 39295
    },
    {
      "entropy": 5.38346037864685,
      "epoch": 4.413769865783119,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032058035612255934,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2256330132484436,
      "num_tokens": 87749981.0,
      "step": 39300
    },
    {
      "entropy": 5.312190055847168,
      "epoch": 4.4143314426910765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003205410391277098,
      "loss": 4.8092,
      "mean_token_accuracy": 0.2245026260614395,
      "num_tokens": 87760193.0,
      "step": 39305
    },
    {
      "entropy": 5.421607494354248,
      "epoch": 4.414893019599034,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032050172068281594,
      "loss": 4.9759,
      "mean_token_accuracy": 0.21956278681755065,
      "num_tokens": 87770625.0,
      "step": 39310
    },
    {
      "entropy": 5.308591032028199,
      "epoch": 4.415454596506992,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003204624007891297,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22604877948760987,
      "num_tokens": 87782294.0,
      "step": 39315
    },
    {
      "entropy": 5.315152883529663,
      "epoch": 4.416016173414949,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000320423079447903,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22575534284114837,
      "num_tokens": 87795320.0,
      "step": 39320
    },
    {
      "entropy": 5.166918039321899,
      "epoch": 4.416577750322907,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00032038375666038776,
      "loss": 4.7245,
      "mean_token_accuracy": 0.23472943902015686,
      "num_tokens": 87806749.0,
      "step": 39325
    },
    {
      "entropy": 5.319843816757202,
      "epoch": 4.417139327230864,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003203444324278361,
      "loss": 4.8969,
      "mean_token_accuracy": 0.23146649301052094,
      "num_tokens": 87817023.0,
      "step": 39330
    },
    {
      "entropy": 5.349138402938843,
      "epoch": 4.4177009041388215,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003203051067515002,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22715823948383332,
      "num_tokens": 87829510.0,
      "step": 39335
    },
    {
      "entropy": 5.3002828598022464,
      "epoch": 4.41826248104678,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003202657796326321,
      "loss": 4.826,
      "mean_token_accuracy": 0.23399077355861664,
      "num_tokens": 87839328.0,
      "step": 39340
    },
    {
      "entropy": 5.408784961700439,
      "epoch": 4.418824057954737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000320226451072484,
      "loss": 4.9938,
      "mean_token_accuracy": 0.23135862052440642,
      "num_tokens": 87851370.0,
      "step": 39345
    },
    {
      "entropy": 5.337325668334961,
      "epoch": 4.419385634862694,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032018712107230815,
      "loss": 4.9479,
      "mean_token_accuracy": 0.24026576578617095,
      "num_tokens": 87863336.0,
      "step": 39350
    },
    {
      "entropy": 5.318598461151123,
      "epoch": 4.419947211770652,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003201477896333568,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2276093989610672,
      "num_tokens": 87873835.0,
      "step": 39355
    },
    {
      "entropy": 5.281771945953369,
      "epoch": 4.420508788678609,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003201084567568824,
      "loss": 4.8545,
      "mean_token_accuracy": 0.2339743882417679,
      "num_tokens": 87884711.0,
      "step": 39360
    },
    {
      "entropy": 5.330022287368775,
      "epoch": 4.4210703655865675,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032006912244413727,
      "loss": 4.9096,
      "mean_token_accuracy": 0.21990752071142197,
      "num_tokens": 87895951.0,
      "step": 39365
    },
    {
      "entropy": 5.376646614074707,
      "epoch": 4.421631942494525,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003200297866963737,
      "loss": 4.9885,
      "mean_token_accuracy": 0.2248436138033867,
      "num_tokens": 87908347.0,
      "step": 39370
    },
    {
      "entropy": 5.312908077239991,
      "epoch": 4.422193519402482,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003199904495148443,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23637448400259017,
      "num_tokens": 87919020.0,
      "step": 39375
    },
    {
      "entropy": 5.270682239532471,
      "epoch": 4.42275509631044,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003199511109008016,
      "loss": 4.8922,
      "mean_token_accuracy": 0.22455971986055373,
      "num_tokens": 87930519.0,
      "step": 39380
    },
    {
      "entropy": 5.3403243064880375,
      "epoch": 4.423316673218397,
      "grad_norm": 0.96875,
      "learning_rate": 0.00031991177085549807,
      "loss": 4.9551,
      "mean_token_accuracy": 0.22673356682062148,
      "num_tokens": 87941800.0,
      "step": 39385
    },
    {
      "entropy": 5.360905599594116,
      "epoch": 4.4238782501263545,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00031987242938018626,
      "loss": 4.9916,
      "mean_token_accuracy": 0.23521870523691177,
      "num_tokens": 87954273.0,
      "step": 39390
    },
    {
      "entropy": 5.4632072925567625,
      "epoch": 4.424439827034313,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031983308647611896,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21870106011629104,
      "num_tokens": 87966089.0,
      "step": 39395
    },
    {
      "entropy": 5.308672952651977,
      "epoch": 4.42500140394227,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003197937421445488,
      "loss": 4.9025,
      "mean_token_accuracy": 0.22545765191316605,
      "num_tokens": 87976944.0,
      "step": 39400
    },
    {
      "entropy": 5.348008394241333,
      "epoch": 4.425562980850227,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003197543963867285,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2268843874335289,
      "num_tokens": 87989715.0,
      "step": 39405
    },
    {
      "entropy": 5.348619031906128,
      "epoch": 4.426124557758185,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003197150492039108,
      "loss": 4.891,
      "mean_token_accuracy": 0.22776401042938232,
      "num_tokens": 88000781.0,
      "step": 39410
    },
    {
      "entropy": 5.363515567779541,
      "epoch": 4.426686134666142,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003196757005973487,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21951886713504792,
      "num_tokens": 88013017.0,
      "step": 39415
    },
    {
      "entropy": 5.35766568183899,
      "epoch": 4.4272477115741005,
      "grad_norm": 0.9375,
      "learning_rate": 0.00031963635056829487,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22210378646850587,
      "num_tokens": 88023762.0,
      "step": 39420
    },
    {
      "entropy": 5.336866426467895,
      "epoch": 4.427809288482058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031959699911800237,
      "loss": 4.9903,
      "mean_token_accuracy": 0.22735532522201538,
      "num_tokens": 88035632.0,
      "step": 39425
    },
    {
      "entropy": 5.317085552215576,
      "epoch": 4.428370865390015,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003195576462477241,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22382467240095139,
      "num_tokens": 88046108.0,
      "step": 39430
    },
    {
      "entropy": 5.417238378524781,
      "epoch": 4.428932442297973,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031951829195871315,
      "loss": 4.9654,
      "mean_token_accuracy": 0.22706470489501954,
      "num_tokens": 88057594.0,
      "step": 39435
    },
    {
      "entropy": 5.2410565376281735,
      "epoch": 4.42949401920593,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00031947893625222244,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2294571116566658,
      "num_tokens": 88069596.0,
      "step": 39440
    },
    {
      "entropy": 5.34272198677063,
      "epoch": 4.430055596113887,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031943957912950523,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2264425814151764,
      "num_tokens": 88080802.0,
      "step": 39445
    },
    {
      "entropy": 5.436150026321411,
      "epoch": 4.4306171730218455,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003194002205918145,
      "loss": 5.0063,
      "mean_token_accuracy": 0.22558885663747788,
      "num_tokens": 88092750.0,
      "step": 39450
    },
    {
      "entropy": 5.425251245498657,
      "epoch": 4.431178749929803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003193608606404036,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2209223672747612,
      "num_tokens": 88103643.0,
      "step": 39455
    },
    {
      "entropy": 5.449653291702271,
      "epoch": 4.43174032683776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003193214992765257,
      "loss": 4.954,
      "mean_token_accuracy": 0.22534777373075485,
      "num_tokens": 88114874.0,
      "step": 39460
    },
    {
      "entropy": 5.229110765457153,
      "epoch": 4.432301903745718,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031928213650143407,
      "loss": 4.7951,
      "mean_token_accuracy": 0.2311714246869087,
      "num_tokens": 88126222.0,
      "step": 39465
    },
    {
      "entropy": 5.231367683410644,
      "epoch": 4.432863480653675,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000319242772316382,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2204155907034874,
      "num_tokens": 88137562.0,
      "step": 39470
    },
    {
      "entropy": 5.352523374557495,
      "epoch": 4.433425057561633,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003192034067226228,
      "loss": 4.9104,
      "mean_token_accuracy": 0.22418105602264404,
      "num_tokens": 88148129.0,
      "step": 39475
    },
    {
      "entropy": 5.348887729644775,
      "epoch": 4.433986634469591,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031916403972141017,
      "loss": 4.8247,
      "mean_token_accuracy": 0.23311634212732316,
      "num_tokens": 88159958.0,
      "step": 39480
    },
    {
      "entropy": 5.395280838012695,
      "epoch": 4.434548211377548,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003191246713139973,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22215466499328612,
      "num_tokens": 88171952.0,
      "step": 39485
    },
    {
      "entropy": 5.325673818588257,
      "epoch": 4.435109788285506,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031908530150163777,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22849203199148177,
      "num_tokens": 88183124.0,
      "step": 39490
    },
    {
      "entropy": 5.274728012084961,
      "epoch": 4.435671365193463,
      "grad_norm": 1.0,
      "learning_rate": 0.00031904593028558517,
      "loss": 4.8038,
      "mean_token_accuracy": 0.2319674924015999,
      "num_tokens": 88194262.0,
      "step": 39495
    },
    {
      "entropy": 5.369441366195678,
      "epoch": 4.43623294210142,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031900655766709305,
      "loss": 4.9651,
      "mean_token_accuracy": 0.214591121673584,
      "num_tokens": 88206423.0,
      "step": 39500
    },
    {
      "entropy": 5.343212604522705,
      "epoch": 4.4367945190093785,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003189671836474151,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22116810977458953,
      "num_tokens": 88217299.0,
      "step": 39505
    },
    {
      "entropy": 5.239911937713623,
      "epoch": 4.437356095917336,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003189278082278049,
      "loss": 4.7532,
      "mean_token_accuracy": 0.23532282412052155,
      "num_tokens": 88228042.0,
      "step": 39510
    },
    {
      "entropy": 5.228815317153931,
      "epoch": 4.437917672825294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031888843140951627,
      "loss": 4.779,
      "mean_token_accuracy": 0.24185292571783065,
      "num_tokens": 88239079.0,
      "step": 39515
    },
    {
      "entropy": 5.316821765899658,
      "epoch": 4.438479249733251,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003188490531938031,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2181081622838974,
      "num_tokens": 88250788.0,
      "step": 39520
    },
    {
      "entropy": 5.447083044052124,
      "epoch": 4.439040826641208,
      "grad_norm": 1.0,
      "learning_rate": 0.00031880967358191884,
      "loss": 4.9095,
      "mean_token_accuracy": 0.23072440028190613,
      "num_tokens": 88261763.0,
      "step": 39525
    },
    {
      "entropy": 5.2974930763244625,
      "epoch": 4.439602403549166,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00031877029257511774,
      "loss": 4.8251,
      "mean_token_accuracy": 0.23587096780538558,
      "num_tokens": 88273015.0,
      "step": 39530
    },
    {
      "entropy": 5.345725393295288,
      "epoch": 4.440163980457124,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031873091017465356,
      "loss": 4.891,
      "mean_token_accuracy": 0.22882061451673508,
      "num_tokens": 88283404.0,
      "step": 39535
    },
    {
      "entropy": 5.2896373748779295,
      "epoch": 4.440725557365081,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003186915263817802,
      "loss": 4.863,
      "mean_token_accuracy": 0.22908073365688325,
      "num_tokens": 88294643.0,
      "step": 39540
    },
    {
      "entropy": 5.342891836166382,
      "epoch": 4.441287134273039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031865214119775173,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22388364970684052,
      "num_tokens": 88305421.0,
      "step": 39545
    },
    {
      "entropy": 5.366492748260498,
      "epoch": 4.441848711180996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000318612754623822,
      "loss": 4.9057,
      "mean_token_accuracy": 0.2236397534608841,
      "num_tokens": 88316154.0,
      "step": 39550
    },
    {
      "entropy": 5.369657564163208,
      "epoch": 4.442410288088954,
      "grad_norm": 1.0,
      "learning_rate": 0.00031857336666124537,
      "loss": 4.8992,
      "mean_token_accuracy": 0.23015168756246568,
      "num_tokens": 88327240.0,
      "step": 39555
    },
    {
      "entropy": 5.344067001342774,
      "epoch": 4.442971864996911,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003185339773112758,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23100090771913528,
      "num_tokens": 88338080.0,
      "step": 39560
    },
    {
      "entropy": 5.328589916229248,
      "epoch": 4.443533441904869,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003184945865751675,
      "loss": 4.9371,
      "mean_token_accuracy": 0.22405379712581636,
      "num_tokens": 88349609.0,
      "step": 39565
    },
    {
      "entropy": 5.294997930526733,
      "epoch": 4.444095018812827,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031845519445417475,
      "loss": 4.8279,
      "mean_token_accuracy": 0.22710233628749849,
      "num_tokens": 88360573.0,
      "step": 39570
    },
    {
      "entropy": 5.387909650802612,
      "epoch": 4.444656595720784,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003184158009495518,
      "loss": 4.9311,
      "mean_token_accuracy": 0.23076122552156447,
      "num_tokens": 88371461.0,
      "step": 39575
    },
    {
      "entropy": 5.307783460617065,
      "epoch": 4.445218172628741,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003183764060625528,
      "loss": 4.89,
      "mean_token_accuracy": 0.22906928211450578,
      "num_tokens": 88382372.0,
      "step": 39580
    },
    {
      "entropy": 5.352380466461182,
      "epoch": 4.445779749536699,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031833700979443227,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22698671370744705,
      "num_tokens": 88393166.0,
      "step": 39585
    },
    {
      "entropy": 5.346959066390991,
      "epoch": 4.4463413264446565,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031829761214644453,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22933561652898787,
      "num_tokens": 88403746.0,
      "step": 39590
    },
    {
      "entropy": 5.239298295974732,
      "epoch": 4.446902903352614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000318258213119844,
      "loss": 4.8128,
      "mean_token_accuracy": 0.2387865886092186,
      "num_tokens": 88414770.0,
      "step": 39595
    },
    {
      "entropy": 5.332181787490844,
      "epoch": 4.447464480260572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003182188127158852,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2241550490260124,
      "num_tokens": 88426545.0,
      "step": 39600
    },
    {
      "entropy": 5.382669258117676,
      "epoch": 4.448026057168529,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003181794109358227,
      "loss": 4.8882,
      "mean_token_accuracy": 0.23456624895334244,
      "num_tokens": 88437147.0,
      "step": 39605
    },
    {
      "entropy": 5.312020874023437,
      "epoch": 4.448587634076487,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003181400077809109,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23696399927139283,
      "num_tokens": 88449965.0,
      "step": 39610
    },
    {
      "entropy": 5.3283110618591305,
      "epoch": 4.449149210984444,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031810060325240466,
      "loss": 4.9677,
      "mean_token_accuracy": 0.22498459964990616,
      "num_tokens": 88460713.0,
      "step": 39615
    },
    {
      "entropy": 5.357755470275879,
      "epoch": 4.449710787892402,
      "grad_norm": 1.0,
      "learning_rate": 0.0003180611973515583,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21391791999340057,
      "num_tokens": 88472760.0,
      "step": 39620
    },
    {
      "entropy": 5.368761682510376,
      "epoch": 4.45027236480036,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031802179007962687,
      "loss": 4.967,
      "mean_token_accuracy": 0.22407209426164626,
      "num_tokens": 88482985.0,
      "step": 39625
    },
    {
      "entropy": 5.359822607040405,
      "epoch": 4.450833941708317,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003179823814378649,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2296901524066925,
      "num_tokens": 88493081.0,
      "step": 39630
    },
    {
      "entropy": 5.297291564941406,
      "epoch": 4.451395518616274,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031794297142752727,
      "loss": 4.9052,
      "mean_token_accuracy": 0.2284200146794319,
      "num_tokens": 88503497.0,
      "step": 39635
    },
    {
      "entropy": 5.328436279296875,
      "epoch": 4.451957095524232,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003179035600498687,
      "loss": 4.8986,
      "mean_token_accuracy": 0.23156222701072693,
      "num_tokens": 88514293.0,
      "step": 39640
    },
    {
      "entropy": 5.4137890338897705,
      "epoch": 4.4525186724321895,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031786414730614423,
      "loss": 4.903,
      "mean_token_accuracy": 0.2216462731361389,
      "num_tokens": 88526137.0,
      "step": 39645
    },
    {
      "entropy": 5.305985164642334,
      "epoch": 4.453080249340147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003178247331976086,
      "loss": 4.8776,
      "mean_token_accuracy": 0.23039168864488602,
      "num_tokens": 88536492.0,
      "step": 39650
    },
    {
      "entropy": 5.355554056167603,
      "epoch": 4.453641826248105,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031778531772551696,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2234683632850647,
      "num_tokens": 88548401.0,
      "step": 39655
    },
    {
      "entropy": 5.35480523109436,
      "epoch": 4.454203403156062,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003177459008911241,
      "loss": 4.9107,
      "mean_token_accuracy": 0.23102037608623505,
      "num_tokens": 88560873.0,
      "step": 39660
    },
    {
      "entropy": 5.365744972229004,
      "epoch": 4.45476498006402,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031770648269568524,
      "loss": 4.8357,
      "mean_token_accuracy": 0.23653408885002136,
      "num_tokens": 88572539.0,
      "step": 39665
    },
    {
      "entropy": 5.263909816741943,
      "epoch": 4.455326556971977,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003176670631404554,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22255334705114366,
      "num_tokens": 88585359.0,
      "step": 39670
    },
    {
      "entropy": 5.282906484603882,
      "epoch": 4.4558881338799345,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031762764222668976,
      "loss": 4.8531,
      "mean_token_accuracy": 0.2328341096639633,
      "num_tokens": 88596553.0,
      "step": 39675
    },
    {
      "entropy": 5.479402637481689,
      "epoch": 4.456449710787893,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00031758821995564343,
      "loss": 5.0155,
      "mean_token_accuracy": 0.2182830661535263,
      "num_tokens": 88608955.0,
      "step": 39680
    },
    {
      "entropy": 5.464870738983154,
      "epoch": 4.45701128769585,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003175487963285716,
      "loss": 4.9957,
      "mean_token_accuracy": 0.22368018180131913,
      "num_tokens": 88619505.0,
      "step": 39685
    },
    {
      "entropy": 5.355525112152099,
      "epoch": 4.457572864603807,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003175093713467297,
      "loss": 4.8449,
      "mean_token_accuracy": 0.23348209857940674,
      "num_tokens": 88632168.0,
      "step": 39690
    },
    {
      "entropy": 5.184128999710083,
      "epoch": 4.458134441511765,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00031746994501137283,
      "loss": 4.8001,
      "mean_token_accuracy": 0.2265692487359047,
      "num_tokens": 88643463.0,
      "step": 39695
    },
    {
      "entropy": 5.3032533645629885,
      "epoch": 4.458696018419722,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003174305173237565,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21786656826734543,
      "num_tokens": 88653894.0,
      "step": 39700
    },
    {
      "entropy": 5.32895975112915,
      "epoch": 4.4592575953276805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000317391088285136,
      "loss": 4.8103,
      "mean_token_accuracy": 0.2300771951675415,
      "num_tokens": 88664610.0,
      "step": 39705
    },
    {
      "entropy": 5.3396327018737795,
      "epoch": 4.459819172235638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003173516578967668,
      "loss": 4.8931,
      "mean_token_accuracy": 0.23128711730241774,
      "num_tokens": 88675040.0,
      "step": 39710
    },
    {
      "entropy": 5.4021257877349855,
      "epoch": 4.460380749143595,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031731222615990443,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22582394927740096,
      "num_tokens": 88686344.0,
      "step": 39715
    },
    {
      "entropy": 5.343878126144409,
      "epoch": 4.460942326051553,
      "grad_norm": 1.0,
      "learning_rate": 0.00031727279307580436,
      "loss": 4.8015,
      "mean_token_accuracy": 0.23153689354658127,
      "num_tokens": 88697538.0,
      "step": 39720
    },
    {
      "entropy": 5.343119955062866,
      "epoch": 4.46150390295951,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031723335864572216,
      "loss": 4.9121,
      "mean_token_accuracy": 0.227900193631649,
      "num_tokens": 88709663.0,
      "step": 39725
    },
    {
      "entropy": 5.366219186782837,
      "epoch": 4.4620654798674675,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00031719392287091343,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22820674628019333,
      "num_tokens": 88720822.0,
      "step": 39730
    },
    {
      "entropy": 5.3275978565216064,
      "epoch": 4.462627056775426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003171544857526338,
      "loss": 4.862,
      "mean_token_accuracy": 0.23116662055253984,
      "num_tokens": 88731453.0,
      "step": 39735
    },
    {
      "entropy": 5.34813551902771,
      "epoch": 4.463188633683383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003171150472921391,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22085969746112824,
      "num_tokens": 88742325.0,
      "step": 39740
    },
    {
      "entropy": 5.372896909713745,
      "epoch": 4.463750210591341,
      "grad_norm": 1.0,
      "learning_rate": 0.00031707560749068487,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22508497387170792,
      "num_tokens": 88754165.0,
      "step": 39745
    },
    {
      "entropy": 5.328558301925659,
      "epoch": 4.464311787499298,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000317036166349527,
      "loss": 4.9212,
      "mean_token_accuracy": 0.23438612669706343,
      "num_tokens": 88765469.0,
      "step": 39750
    },
    {
      "entropy": 5.2627912044525145,
      "epoch": 4.464873364407255,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003169967238699213,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22380549609661102,
      "num_tokens": 88775849.0,
      "step": 39755
    },
    {
      "entropy": 5.299552297592163,
      "epoch": 4.4654349413152135,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031695728005312363,
      "loss": 4.8809,
      "mean_token_accuracy": 0.23068566173315047,
      "num_tokens": 88787385.0,
      "step": 39760
    },
    {
      "entropy": 5.2912548065185545,
      "epoch": 4.465996518223171,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003169178349003898,
      "loss": 4.8167,
      "mean_token_accuracy": 0.23500480651855468,
      "num_tokens": 88798264.0,
      "step": 39765
    },
    {
      "entropy": 5.315749025344848,
      "epoch": 4.466558095131128,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000316878388412976,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22320104837417604,
      "num_tokens": 88808833.0,
      "step": 39770
    },
    {
      "entropy": 5.338047361373901,
      "epoch": 4.467119672039086,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000316838940592138,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22150078415870667,
      "num_tokens": 88819757.0,
      "step": 39775
    },
    {
      "entropy": 5.317291116714477,
      "epoch": 4.467681248947043,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031679949143913183,
      "loss": 4.9009,
      "mean_token_accuracy": 0.233622944355011,
      "num_tokens": 88831120.0,
      "step": 39780
    },
    {
      "entropy": 5.366607761383056,
      "epoch": 4.468242825855,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003167600409552137,
      "loss": 4.8902,
      "mean_token_accuracy": 0.225513756275177,
      "num_tokens": 88841674.0,
      "step": 39785
    },
    {
      "entropy": 5.263133525848389,
      "epoch": 4.4688044027629585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003167205891416397,
      "loss": 4.8196,
      "mean_token_accuracy": 0.22906122654676436,
      "num_tokens": 88851844.0,
      "step": 39790
    },
    {
      "entropy": 5.341765832901001,
      "epoch": 4.469365979670916,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000316681135999666,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22438384741544723,
      "num_tokens": 88862638.0,
      "step": 39795
    },
    {
      "entropy": 5.305462741851807,
      "epoch": 4.469927556578874,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003166416815305486,
      "loss": 4.8327,
      "mean_token_accuracy": 0.23317761868238449,
      "num_tokens": 88873427.0,
      "step": 39800
    },
    {
      "entropy": 5.304219388961792,
      "epoch": 4.470489133486831,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031660222573554406,
      "loss": 4.895,
      "mean_token_accuracy": 0.22628800719976425,
      "num_tokens": 88885461.0,
      "step": 39805
    },
    {
      "entropy": 5.352918529510498,
      "epoch": 4.471050710394788,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003165627686159085,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2205628737807274,
      "num_tokens": 88896501.0,
      "step": 39810
    },
    {
      "entropy": 5.385996866226196,
      "epoch": 4.471612287302746,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003165233101728982,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22716703414916992,
      "num_tokens": 88906653.0,
      "step": 39815
    },
    {
      "entropy": 5.257912826538086,
      "epoch": 4.472173864210704,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00031648385040776965,
      "loss": 4.834,
      "mean_token_accuracy": 0.23311492800712585,
      "num_tokens": 88917022.0,
      "step": 39820
    },
    {
      "entropy": 5.277245998382568,
      "epoch": 4.472735441118661,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003164443893217793,
      "loss": 4.7878,
      "mean_token_accuracy": 0.2265198677778244,
      "num_tokens": 88927632.0,
      "step": 39825
    },
    {
      "entropy": 5.290548372268677,
      "epoch": 4.473297018026619,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003164049269161834,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23173531144857407,
      "num_tokens": 88938178.0,
      "step": 39830
    },
    {
      "entropy": 5.27850456237793,
      "epoch": 4.473858594934576,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031636546319223845,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23721863180398942,
      "num_tokens": 88948549.0,
      "step": 39835
    },
    {
      "entropy": 5.247592830657959,
      "epoch": 4.474420171842533,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003163259981512013,
      "loss": 4.8371,
      "mean_token_accuracy": 0.22553878128528596,
      "num_tokens": 88959828.0,
      "step": 39840
    },
    {
      "entropy": 5.38188099861145,
      "epoch": 4.4749817487504915,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0003162865317943283,
      "loss": 4.8955,
      "mean_token_accuracy": 0.23285338282585144,
      "num_tokens": 88971705.0,
      "step": 39845
    },
    {
      "entropy": 5.291318035125732,
      "epoch": 4.475543325658449,
      "grad_norm": 1.0,
      "learning_rate": 0.00031624706412287614,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22545673102140426,
      "num_tokens": 88982853.0,
      "step": 39850
    },
    {
      "entropy": 5.2658953189849855,
      "epoch": 4.476104902566407,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003162075951381014,
      "loss": 4.8089,
      "mean_token_accuracy": 0.23663706332445145,
      "num_tokens": 88993028.0,
      "step": 39855
    },
    {
      "entropy": 5.369491195678711,
      "epoch": 4.476666479474364,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031616812484126084,
      "loss": 4.9743,
      "mean_token_accuracy": 0.2211526960134506,
      "num_tokens": 89004164.0,
      "step": 39860
    },
    {
      "entropy": 5.375478267669678,
      "epoch": 4.477228056382321,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003161286532336113,
      "loss": 4.9093,
      "mean_token_accuracy": 0.23506587743759155,
      "num_tokens": 89015868.0,
      "step": 39865
    },
    {
      "entropy": 5.292371511459351,
      "epoch": 4.477789633290279,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031608918031640946,
      "loss": 4.8307,
      "mean_token_accuracy": 0.2288281112909317,
      "num_tokens": 89026302.0,
      "step": 39870
    },
    {
      "entropy": 5.259103202819825,
      "epoch": 4.478351210198237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003160497060909121,
      "loss": 4.8863,
      "mean_token_accuracy": 0.22497027218341828,
      "num_tokens": 89036480.0,
      "step": 39875
    },
    {
      "entropy": 5.397715711593628,
      "epoch": 4.478912787106194,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031601023055837624,
      "loss": 4.967,
      "mean_token_accuracy": 0.22505880743265153,
      "num_tokens": 89047850.0,
      "step": 39880
    },
    {
      "entropy": 5.3587171077728275,
      "epoch": 4.479474364014152,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031597075372005875,
      "loss": 4.8492,
      "mean_token_accuracy": 0.22981864511966704,
      "num_tokens": 89059298.0,
      "step": 39885
    },
    {
      "entropy": 5.347992467880249,
      "epoch": 4.480035940922109,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031593127557721657,
      "loss": 4.9113,
      "mean_token_accuracy": 0.2297041267156601,
      "num_tokens": 89069966.0,
      "step": 39890
    },
    {
      "entropy": 5.348856639862061,
      "epoch": 4.480597517830067,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003158917961311067,
      "loss": 4.9501,
      "mean_token_accuracy": 0.22232261449098586,
      "num_tokens": 89081277.0,
      "step": 39895
    },
    {
      "entropy": 5.430898237228393,
      "epoch": 4.481159094738024,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003158523153829862,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22373025864362717,
      "num_tokens": 89092221.0,
      "step": 39900
    },
    {
      "entropy": 5.344412469863892,
      "epoch": 4.481720671645982,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003158128333341121,
      "loss": 4.924,
      "mean_token_accuracy": 0.2226986214518547,
      "num_tokens": 89103232.0,
      "step": 39905
    },
    {
      "entropy": 5.240042591094971,
      "epoch": 4.48228224855394,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031577334998574155,
      "loss": 4.876,
      "mean_token_accuracy": 0.22506940364837646,
      "num_tokens": 89113222.0,
      "step": 39910
    },
    {
      "entropy": 5.376622438430786,
      "epoch": 4.482843825461897,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003157338653391317,
      "loss": 4.9936,
      "mean_token_accuracy": 0.216456601023674,
      "num_tokens": 89125396.0,
      "step": 39915
    },
    {
      "entropy": 5.473752212524414,
      "epoch": 4.483405402369854,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003156943793955398,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22036432921886445,
      "num_tokens": 89136199.0,
      "step": 39920
    },
    {
      "entropy": 5.4411139488220215,
      "epoch": 4.483966979277812,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031565489215622304,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21621793806552886,
      "num_tokens": 89146761.0,
      "step": 39925
    },
    {
      "entropy": 5.366348648071289,
      "epoch": 4.4845285561857695,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003156154036224388,
      "loss": 4.9777,
      "mean_token_accuracy": 0.22512317895889283,
      "num_tokens": 89157783.0,
      "step": 39930
    },
    {
      "entropy": 5.237155151367188,
      "epoch": 4.485090133093728,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003155759137954444,
      "loss": 4.8027,
      "mean_token_accuracy": 0.23095589131116867,
      "num_tokens": 89168376.0,
      "step": 39935
    },
    {
      "entropy": 5.276667785644531,
      "epoch": 4.485651710001685,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00031553642267649715,
      "loss": 4.8079,
      "mean_token_accuracy": 0.23364906907081603,
      "num_tokens": 89179113.0,
      "step": 39940
    },
    {
      "entropy": 5.332836627960205,
      "epoch": 4.486213286909642,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003154969302668544,
      "loss": 4.9015,
      "mean_token_accuracy": 0.2293717458844185,
      "num_tokens": 89190233.0,
      "step": 39945
    },
    {
      "entropy": 5.345413541793823,
      "epoch": 4.4867748638176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003154574365677737,
      "loss": 4.8936,
      "mean_token_accuracy": 0.23045855462551118,
      "num_tokens": 89200800.0,
      "step": 39950
    },
    {
      "entropy": 5.478984594345093,
      "epoch": 4.487336440725557,
      "grad_norm": 1.0,
      "learning_rate": 0.0003154179415805126,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2176470100879669,
      "num_tokens": 89211865.0,
      "step": 39955
    },
    {
      "entropy": 5.436945390701294,
      "epoch": 4.487898017633515,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003153784453063285,
      "loss": 4.9708,
      "mean_token_accuracy": 0.2204291731119156,
      "num_tokens": 89224151.0,
      "step": 39960
    },
    {
      "entropy": 5.30970401763916,
      "epoch": 4.488459594541473,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003153389477464791,
      "loss": 4.8481,
      "mean_token_accuracy": 0.22851054668426513,
      "num_tokens": 89235095.0,
      "step": 39965
    },
    {
      "entropy": 5.393049240112305,
      "epoch": 4.48902117144943,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003152994489022219,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21566759496927262,
      "num_tokens": 89247891.0,
      "step": 39970
    },
    {
      "entropy": 5.354640150070191,
      "epoch": 4.489582748357387,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031525994877481465,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22765126526355745,
      "num_tokens": 89259395.0,
      "step": 39975
    },
    {
      "entropy": 5.31620225906372,
      "epoch": 4.490144325265345,
      "grad_norm": 1.015625,
      "learning_rate": 0.000315220447365515,
      "loss": 4.8802,
      "mean_token_accuracy": 0.2293820172548294,
      "num_tokens": 89269715.0,
      "step": 39980
    },
    {
      "entropy": 5.285018110275269,
      "epoch": 4.4907059021733025,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031518094467558074,
      "loss": 4.8624,
      "mean_token_accuracy": 0.23455798029899597,
      "num_tokens": 89280260.0,
      "step": 39985
    },
    {
      "entropy": 5.2565676212310795,
      "epoch": 4.491267479081261,
      "grad_norm": 1.0,
      "learning_rate": 0.00031514144070626956,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23055990636348725,
      "num_tokens": 89289678.0,
      "step": 39990
    },
    {
      "entropy": 5.301475524902344,
      "epoch": 4.491829055989218,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00031510193545883944,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22982937544584275,
      "num_tokens": 89300534.0,
      "step": 39995
    },
    {
      "entropy": 5.402946615219117,
      "epoch": 4.492390632897175,
      "grad_norm": 1.09375,
      "learning_rate": 0.000315062428934548,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22246890068054198,
      "num_tokens": 89311873.0,
      "step": 40000
    },
    {
      "entropy": 5.230341196060181,
      "epoch": 4.492952209805133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031502292113465334,
      "loss": 4.7958,
      "mean_token_accuracy": 0.23731550872325896,
      "num_tokens": 89322941.0,
      "step": 40005
    },
    {
      "entropy": 5.307804727554322,
      "epoch": 4.49351378671309,
      "grad_norm": 0.96875,
      "learning_rate": 0.00031498341206041337,
      "loss": 4.8717,
      "mean_token_accuracy": 0.22978683263063432,
      "num_tokens": 89335028.0,
      "step": 40010
    },
    {
      "entropy": 5.417961645126343,
      "epoch": 4.4940753636210475,
      "grad_norm": 0.953125,
      "learning_rate": 0.000314943901713086,
      "loss": 4.9136,
      "mean_token_accuracy": 0.224075049161911,
      "num_tokens": 89345929.0,
      "step": 40015
    },
    {
      "entropy": 5.356483221054077,
      "epoch": 4.494636940529006,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003149043900939294,
      "loss": 4.937,
      "mean_token_accuracy": 0.22688587754964828,
      "num_tokens": 89357496.0,
      "step": 40020
    },
    {
      "entropy": 5.335915374755859,
      "epoch": 4.495198517436963,
      "grad_norm": 0.921875,
      "learning_rate": 0.00031486487720420146,
      "loss": 4.9664,
      "mean_token_accuracy": 0.21988402158021927,
      "num_tokens": 89370304.0,
      "step": 40025
    },
    {
      "entropy": 5.401082754135132,
      "epoch": 4.49576009434492,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031482536304516047,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22952803373336791,
      "num_tokens": 89382642.0,
      "step": 40030
    },
    {
      "entropy": 5.3780372619628904,
      "epoch": 4.496321671252878,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003147858476180644,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2272280350327492,
      "num_tokens": 89393730.0,
      "step": 40035
    },
    {
      "entropy": 5.395279169082642,
      "epoch": 4.496883248160835,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031474633092417155,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2251102715730667,
      "num_tokens": 89404210.0,
      "step": 40040
    },
    {
      "entropy": 5.392739868164062,
      "epoch": 4.4974448250687935,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031470681296474006,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2275625631213188,
      "num_tokens": 89414204.0,
      "step": 40045
    },
    {
      "entropy": 5.326411724090576,
      "epoch": 4.498006401976751,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003146672937410283,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22785369902849198,
      "num_tokens": 89424933.0,
      "step": 40050
    },
    {
      "entropy": 5.34465594291687,
      "epoch": 4.498567978884708,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00031462777325429447,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22366296797990798,
      "num_tokens": 89437086.0,
      "step": 40055
    },
    {
      "entropy": 5.325714015960694,
      "epoch": 4.499129555792666,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000314588251505797,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2348804384469986,
      "num_tokens": 89449122.0,
      "step": 40060
    },
    {
      "entropy": 5.347835111618042,
      "epoch": 4.499691132700623,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031454872849679425,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22788103371858598,
      "num_tokens": 89459195.0,
      "step": 40065
    },
    {
      "entropy": 5.268680095672607,
      "epoch": 4.500252709608581,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003145092042285446,
      "loss": 4.789,
      "mean_token_accuracy": 0.23398573249578475,
      "num_tokens": 89469967.0,
      "step": 40070
    },
    {
      "entropy": 5.341243076324463,
      "epoch": 4.500814286516539,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003144696787023066,
      "loss": 4.8431,
      "mean_token_accuracy": 0.23212233036756516,
      "num_tokens": 89481950.0,
      "step": 40075
    },
    {
      "entropy": 5.272696590423584,
      "epoch": 4.501375863424496,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031443015191933864,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23381924778223037,
      "num_tokens": 89492915.0,
      "step": 40080
    },
    {
      "entropy": 5.277812194824219,
      "epoch": 4.501937440332454,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031439062388089953,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22860498130321502,
      "num_tokens": 89504384.0,
      "step": 40085
    },
    {
      "entropy": 5.404775810241699,
      "epoch": 4.502499017240411,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003143510945882475,
      "loss": 4.9358,
      "mean_token_accuracy": 0.22490473091602325,
      "num_tokens": 89515518.0,
      "step": 40090
    },
    {
      "entropy": 5.439474821090698,
      "epoch": 4.503060594148368,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031431156404264143,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22579970210790634,
      "num_tokens": 89526538.0,
      "step": 40095
    },
    {
      "entropy": 5.3390813827514645,
      "epoch": 4.5036221710563265,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031427203224533996,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22903421223163606,
      "num_tokens": 89537390.0,
      "step": 40100
    },
    {
      "entropy": 5.334399318695068,
      "epoch": 4.504183747964284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031423249919760167,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22707616537809372,
      "num_tokens": 89548691.0,
      "step": 40105
    },
    {
      "entropy": 5.329623985290527,
      "epoch": 4.504745324872241,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003141929649006854,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22497119009494781,
      "num_tokens": 89559450.0,
      "step": 40110
    },
    {
      "entropy": 5.332333326339722,
      "epoch": 4.505306901780199,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031415342935585,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23746279776096343,
      "num_tokens": 89570589.0,
      "step": 40115
    },
    {
      "entropy": 5.323142576217651,
      "epoch": 4.505868478688156,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003141138925643542,
      "loss": 4.8848,
      "mean_token_accuracy": 0.2259846583008766,
      "num_tokens": 89581469.0,
      "step": 40120
    },
    {
      "entropy": 5.357850885391235,
      "epoch": 4.506430055596114,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003140743545274569,
      "loss": 4.9622,
      "mean_token_accuracy": 0.22668316960334778,
      "num_tokens": 89592023.0,
      "step": 40125
    },
    {
      "entropy": 5.26533203125,
      "epoch": 4.5069916325040715,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00031403481524641696,
      "loss": 4.8291,
      "mean_token_accuracy": 0.235243821144104,
      "num_tokens": 89603270.0,
      "step": 40130
    },
    {
      "entropy": 5.344384098052979,
      "epoch": 4.507553209412029,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00031399527472249336,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22415419965982436,
      "num_tokens": 89614185.0,
      "step": 40135
    },
    {
      "entropy": 5.341907024383545,
      "epoch": 4.508114786319987,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031395573295694507,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22465303242206575,
      "num_tokens": 89627245.0,
      "step": 40140
    },
    {
      "entropy": 5.321564149856568,
      "epoch": 4.508676363227944,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031391618995103116,
      "loss": 4.8264,
      "mean_token_accuracy": 0.2315902516245842,
      "num_tokens": 89638710.0,
      "step": 40145
    },
    {
      "entropy": 5.388982772827148,
      "epoch": 4.509237940135901,
      "grad_norm": 1.0,
      "learning_rate": 0.0003138766457060107,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21901729702949524,
      "num_tokens": 89649658.0,
      "step": 40150
    },
    {
      "entropy": 5.292972230911255,
      "epoch": 4.509799517043859,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003138371002231427,
      "loss": 4.8668,
      "mean_token_accuracy": 0.23209239393472672,
      "num_tokens": 89659933.0,
      "step": 40155
    },
    {
      "entropy": 5.354555082321167,
      "epoch": 4.510361093951817,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003137975535036864,
      "loss": 4.9673,
      "mean_token_accuracy": 0.22156819105148315,
      "num_tokens": 89670790.0,
      "step": 40160
    },
    {
      "entropy": 5.4060704708099365,
      "epoch": 4.510922670859774,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031375800554890094,
      "loss": 4.9362,
      "mean_token_accuracy": 0.2200278088450432,
      "num_tokens": 89682000.0,
      "step": 40165
    },
    {
      "entropy": 5.229625988006592,
      "epoch": 4.511484247767732,
      "grad_norm": 1.0,
      "learning_rate": 0.00031371845636004556,
      "loss": 4.7859,
      "mean_token_accuracy": 0.2404778555035591,
      "num_tokens": 89694058.0,
      "step": 40170
    },
    {
      "entropy": 5.298375177383423,
      "epoch": 4.512045824675689,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003136789059383795,
      "loss": 4.859,
      "mean_token_accuracy": 0.22841891944408416,
      "num_tokens": 89706074.0,
      "step": 40175
    },
    {
      "entropy": 5.395780611038208,
      "epoch": 4.512607401583647,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003136393542851622,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22701324820518493,
      "num_tokens": 89716273.0,
      "step": 40180
    },
    {
      "entropy": 5.381328535079956,
      "epoch": 4.5131689784916045,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031359980140165274,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22792749106884003,
      "num_tokens": 89726354.0,
      "step": 40185
    },
    {
      "entropy": 5.387559843063355,
      "epoch": 4.513730555399562,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003135602472891106,
      "loss": 4.9437,
      "mean_token_accuracy": 0.22242569476366042,
      "num_tokens": 89739554.0,
      "step": 40190
    },
    {
      "entropy": 5.351039552688599,
      "epoch": 4.51429213230752,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031352069194879527,
      "loss": 4.8613,
      "mean_token_accuracy": 0.2265229970216751,
      "num_tokens": 89750717.0,
      "step": 40195
    },
    {
      "entropy": 5.370640087127685,
      "epoch": 4.514853709215477,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003134811353819662,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22477404177188873,
      "num_tokens": 89762649.0,
      "step": 40200
    },
    {
      "entropy": 5.305569839477539,
      "epoch": 4.515415286123434,
      "grad_norm": 1.0,
      "learning_rate": 0.0003134415775898828,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22367028445005416,
      "num_tokens": 89774364.0,
      "step": 40205
    },
    {
      "entropy": 5.421600103378296,
      "epoch": 4.515976863031392,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003134020185738047,
      "loss": 5.0295,
      "mean_token_accuracy": 0.22122501730918884,
      "num_tokens": 89785526.0,
      "step": 40210
    },
    {
      "entropy": 5.31279239654541,
      "epoch": 4.51653843993935,
      "grad_norm": 1.0,
      "learning_rate": 0.00031336245833499146,
      "loss": 4.8332,
      "mean_token_accuracy": 0.23174857050180436,
      "num_tokens": 89797119.0,
      "step": 40215
    },
    {
      "entropy": 5.379166173934936,
      "epoch": 4.517100016847307,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003133228968747026,
      "loss": 5.037,
      "mean_token_accuracy": 0.21919465214014053,
      "num_tokens": 89810802.0,
      "step": 40220
    },
    {
      "entropy": 5.41801700592041,
      "epoch": 4.517661593755265,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003132833341941979,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21967955976724624,
      "num_tokens": 89821879.0,
      "step": 40225
    },
    {
      "entropy": 5.403740119934082,
      "epoch": 4.518223170663222,
      "grad_norm": 1.015625,
      "learning_rate": 0.000313243770294737,
      "loss": 4.9014,
      "mean_token_accuracy": 0.23228334486484528,
      "num_tokens": 89832487.0,
      "step": 40230
    },
    {
      "entropy": 5.399937677383423,
      "epoch": 4.51878474757118,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031320420517757955,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2162322372198105,
      "num_tokens": 89844210.0,
      "step": 40235
    },
    {
      "entropy": 5.367661857604981,
      "epoch": 4.519346324479137,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003131646388439854,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2262765035033226,
      "num_tokens": 89856336.0,
      "step": 40240
    },
    {
      "entropy": 5.382241535186767,
      "epoch": 4.519907901387095,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003131250712952144,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22186087369918822,
      "num_tokens": 89868159.0,
      "step": 40245
    },
    {
      "entropy": 5.284571838378906,
      "epoch": 4.520469478295053,
      "grad_norm": 0.921875,
      "learning_rate": 0.00031308550253252634,
      "loss": 4.7611,
      "mean_token_accuracy": 0.23979784101247786,
      "num_tokens": 89879204.0,
      "step": 40250
    },
    {
      "entropy": 5.290581607818604,
      "epoch": 4.52103105520301,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031304593255718113,
      "loss": 4.8251,
      "mean_token_accuracy": 0.23474963009357452,
      "num_tokens": 89889230.0,
      "step": 40255
    },
    {
      "entropy": 5.262454986572266,
      "epoch": 4.521592632110968,
      "grad_norm": 0.96875,
      "learning_rate": 0.00031300636137043857,
      "loss": 4.8696,
      "mean_token_accuracy": 0.2279553771018982,
      "num_tokens": 89900318.0,
      "step": 40260
    },
    {
      "entropy": 5.2942774295806885,
      "epoch": 4.522154209018925,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00031296678897355877,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22452975660562516,
      "num_tokens": 89911754.0,
      "step": 40265
    },
    {
      "entropy": 5.366271686553955,
      "epoch": 4.5227157859268825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003129272153678017,
      "loss": 4.8512,
      "mean_token_accuracy": 0.2315245598554611,
      "num_tokens": 89922423.0,
      "step": 40270
    },
    {
      "entropy": 5.348597192764283,
      "epoch": 4.523277362834841,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003128876405544274,
      "loss": 4.8548,
      "mean_token_accuracy": 0.23298238664865495,
      "num_tokens": 89934275.0,
      "step": 40275
    },
    {
      "entropy": 5.313104200363159,
      "epoch": 4.523838939742798,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031284806453469596,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22207435518503188,
      "num_tokens": 89944636.0,
      "step": 40280
    },
    {
      "entropy": 5.4078912258148195,
      "epoch": 4.524400516650755,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031280848730986735,
      "loss": 4.9168,
      "mean_token_accuracy": 0.23145437985658646,
      "num_tokens": 89955659.0,
      "step": 40285
    },
    {
      "entropy": 5.289071416854858,
      "epoch": 4.524962093558713,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003127689088812019,
      "loss": 4.8783,
      "mean_token_accuracy": 0.2295724034309387,
      "num_tokens": 89967454.0,
      "step": 40290
    },
    {
      "entropy": 5.412450933456421,
      "epoch": 4.52552367046667,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003127293292499598,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21798314154148102,
      "num_tokens": 89979080.0,
      "step": 40295
    },
    {
      "entropy": 5.312991809844971,
      "epoch": 4.526085247374628,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003126897484174012,
      "loss": 4.8492,
      "mean_token_accuracy": 0.22314361482858658,
      "num_tokens": 89989849.0,
      "step": 40300
    },
    {
      "entropy": 5.37908296585083,
      "epoch": 4.526646824282586,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003126501663847863,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22751522660255433,
      "num_tokens": 90000352.0,
      "step": 40305
    },
    {
      "entropy": 5.462992000579834,
      "epoch": 4.527208401190543,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003126105831533756,
      "loss": 5.0095,
      "mean_token_accuracy": 0.2169117733836174,
      "num_tokens": 90011958.0,
      "step": 40310
    },
    {
      "entropy": 5.391449737548828,
      "epoch": 4.527769978098501,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003125709987244293,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2298033729195595,
      "num_tokens": 90022951.0,
      "step": 40315
    },
    {
      "entropy": 5.321111059188842,
      "epoch": 4.528331555006458,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00031253141309920784,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22936664521694183,
      "num_tokens": 90034295.0,
      "step": 40320
    },
    {
      "entropy": 5.3699744701385494,
      "epoch": 4.5288931319144154,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031249182627897165,
      "loss": 4.9816,
      "mean_token_accuracy": 0.2253014475107193,
      "num_tokens": 90046552.0,
      "step": 40325
    },
    {
      "entropy": 5.3779387950897215,
      "epoch": 4.5294547088223736,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00031245223826498116,
      "loss": 4.9502,
      "mean_token_accuracy": 0.21852296590805054,
      "num_tokens": 90057791.0,
      "step": 40330
    },
    {
      "entropy": 5.264748430252075,
      "epoch": 4.530016285730331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031241264905849687,
      "loss": 4.818,
      "mean_token_accuracy": 0.2281045436859131,
      "num_tokens": 90069569.0,
      "step": 40335
    },
    {
      "entropy": 5.265216398239136,
      "epoch": 4.530577862638288,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031237305866077926,
      "loss": 4.8169,
      "mean_token_accuracy": 0.2273932456970215,
      "num_tokens": 90080641.0,
      "step": 40340
    },
    {
      "entropy": 5.382593011856079,
      "epoch": 4.531139439546246,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031233346707308906,
      "loss": 4.91,
      "mean_token_accuracy": 0.224020154774189,
      "num_tokens": 90091755.0,
      "step": 40345
    },
    {
      "entropy": 5.403290939331055,
      "epoch": 4.531701016454203,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031229387429668675,
      "loss": 4.9792,
      "mean_token_accuracy": 0.22800492495298386,
      "num_tokens": 90102731.0,
      "step": 40350
    },
    {
      "entropy": 5.21077151298523,
      "epoch": 4.5322625933621605,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031225428033283305,
      "loss": 4.8059,
      "mean_token_accuracy": 0.23273332715034484,
      "num_tokens": 90113660.0,
      "step": 40355
    },
    {
      "entropy": 5.346395969390869,
      "epoch": 4.532824170270119,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003122146851827886,
      "loss": 4.8593,
      "mean_token_accuracy": 0.2289905548095703,
      "num_tokens": 90124634.0,
      "step": 40360
    },
    {
      "entropy": 5.429896450042724,
      "epoch": 4.533385747178076,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031217508884781413,
      "loss": 4.9815,
      "mean_token_accuracy": 0.21123291701078414,
      "num_tokens": 90135550.0,
      "step": 40365
    },
    {
      "entropy": 5.284555578231812,
      "epoch": 4.533947324086034,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003121354913291704,
      "loss": 4.7934,
      "mean_token_accuracy": 0.23622881323099137,
      "num_tokens": 90146151.0,
      "step": 40370
    },
    {
      "entropy": 5.22879467010498,
      "epoch": 4.534508900993991,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003120958926281182,
      "loss": 4.817,
      "mean_token_accuracy": 0.22975816428661347,
      "num_tokens": 90157680.0,
      "step": 40375
    },
    {
      "entropy": 5.280187702178955,
      "epoch": 4.535070477901948,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031205629274591844,
      "loss": 4.7982,
      "mean_token_accuracy": 0.23245138972997664,
      "num_tokens": 90168051.0,
      "step": 40380
    },
    {
      "entropy": 5.277214527130127,
      "epoch": 4.5356320548099065,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031201669168383193,
      "loss": 4.787,
      "mean_token_accuracy": 0.23422672748565673,
      "num_tokens": 90179995.0,
      "step": 40385
    },
    {
      "entropy": 5.305891227722168,
      "epoch": 4.536193631717864,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00031197708944311964,
      "loss": 4.8998,
      "mean_token_accuracy": 0.2319309875369072,
      "num_tokens": 90189714.0,
      "step": 40390
    },
    {
      "entropy": 5.38742036819458,
      "epoch": 4.536755208625821,
      "grad_norm": 1.125,
      "learning_rate": 0.0003119374860250425,
      "loss": 4.9339,
      "mean_token_accuracy": 0.22491195648908616,
      "num_tokens": 90200830.0,
      "step": 40395
    },
    {
      "entropy": 5.324993896484375,
      "epoch": 4.537316785533779,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003118978814308614,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2188737392425537,
      "num_tokens": 90213032.0,
      "step": 40400
    },
    {
      "entropy": 5.340184926986694,
      "epoch": 4.537878362441736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003118582756618375,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2289588302373886,
      "num_tokens": 90224788.0,
      "step": 40405
    },
    {
      "entropy": 5.287873840332031,
      "epoch": 4.5384399393496935,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003118186687192318,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22336844801902772,
      "num_tokens": 90237142.0,
      "step": 40410
    },
    {
      "entropy": 5.35399899482727,
      "epoch": 4.539001516257652,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031177906060430536,
      "loss": 4.8413,
      "mean_token_accuracy": 0.23093292117118835,
      "num_tokens": 90247327.0,
      "step": 40415
    },
    {
      "entropy": 5.234691047668457,
      "epoch": 4.539563093165609,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031173945131831936,
      "loss": 4.7822,
      "mean_token_accuracy": 0.2365085706114769,
      "num_tokens": 90258416.0,
      "step": 40420
    },
    {
      "entropy": 5.303056335449218,
      "epoch": 4.540124670073567,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000311699840862535,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22568028569221496,
      "num_tokens": 90269700.0,
      "step": 40425
    },
    {
      "entropy": 5.319293546676636,
      "epoch": 4.540686246981524,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003116602292382135,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23040542602539063,
      "num_tokens": 90280181.0,
      "step": 40430
    },
    {
      "entropy": 5.417269039154053,
      "epoch": 4.541247823889481,
      "grad_norm": 1.0,
      "learning_rate": 0.00031162061644661606,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22350662350654601,
      "num_tokens": 90291317.0,
      "step": 40435
    },
    {
      "entropy": 5.3964598178863525,
      "epoch": 4.5418094007974394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000311581002489004,
      "loss": 4.9388,
      "mean_token_accuracy": 0.2275594025850296,
      "num_tokens": 90302462.0,
      "step": 40440
    },
    {
      "entropy": 5.313172626495361,
      "epoch": 4.542370977705397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003115413873666387,
      "loss": 4.8323,
      "mean_token_accuracy": 0.22701680660247803,
      "num_tokens": 90312432.0,
      "step": 40445
    },
    {
      "entropy": 5.261250972747803,
      "epoch": 4.542932554613355,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003115017710807813,
      "loss": 4.9177,
      "mean_token_accuracy": 0.22395511567592621,
      "num_tokens": 90323445.0,
      "step": 40450
    },
    {
      "entropy": 5.370176076889038,
      "epoch": 4.543494131521312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003114621536326934,
      "loss": 4.9225,
      "mean_token_accuracy": 0.22511665225028993,
      "num_tokens": 90334445.0,
      "step": 40455
    },
    {
      "entropy": 5.425312042236328,
      "epoch": 4.544055708429269,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003114225350236365,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22874351292848588,
      "num_tokens": 90346679.0,
      "step": 40460
    },
    {
      "entropy": 5.310408639907837,
      "epoch": 4.544617285337227,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003113829152548719,
      "loss": 4.8225,
      "mean_token_accuracy": 0.23328294605016708,
      "num_tokens": 90358643.0,
      "step": 40465
    },
    {
      "entropy": 5.292751359939575,
      "epoch": 4.5451788622451845,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003113432943276611,
      "loss": 4.9076,
      "mean_token_accuracy": 0.22950988560914992,
      "num_tokens": 90369401.0,
      "step": 40470
    },
    {
      "entropy": 5.318414306640625,
      "epoch": 4.545740439153142,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00031130367224326585,
      "loss": 4.8328,
      "mean_token_accuracy": 0.23486417829990386,
      "num_tokens": 90380671.0,
      "step": 40475
    },
    {
      "entropy": 5.328016233444214,
      "epoch": 4.5463020160611,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003112640490029475,
      "loss": 4.9005,
      "mean_token_accuracy": 0.23132936507463456,
      "num_tokens": 90392123.0,
      "step": 40480
    },
    {
      "entropy": 5.32335352897644,
      "epoch": 4.546863592969057,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00031122442460796774,
      "loss": 4.862,
      "mean_token_accuracy": 0.22665992081165315,
      "num_tokens": 90404441.0,
      "step": 40485
    },
    {
      "entropy": 5.398189926147461,
      "epoch": 4.547425169877014,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003111847990595883,
      "loss": 5.0414,
      "mean_token_accuracy": 0.22558159083127977,
      "num_tokens": 90415281.0,
      "step": 40490
    },
    {
      "entropy": 5.295166778564453,
      "epoch": 4.547986746784972,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003111451723590708,
      "loss": 4.8067,
      "mean_token_accuracy": 0.23319427520036698,
      "num_tokens": 90424719.0,
      "step": 40495
    },
    {
      "entropy": 5.39511079788208,
      "epoch": 4.54854832369293,
      "grad_norm": 1.0,
      "learning_rate": 0.000311105544507677,
      "loss": 4.9834,
      "mean_token_accuracy": 0.22110475897789,
      "num_tokens": 90436335.0,
      "step": 40500
    },
    {
      "entropy": 5.329921913146973,
      "epoch": 4.549109900600888,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00031106591550666865,
      "loss": 4.8887,
      "mean_token_accuracy": 0.22687437683343886,
      "num_tokens": 90449268.0,
      "step": 40505
    },
    {
      "entropy": 5.451831960678101,
      "epoch": 4.549671477508845,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003110262853573076,
      "loss": 5.0151,
      "mean_token_accuracy": 0.22151652723550797,
      "num_tokens": 90460974.0,
      "step": 40510
    },
    {
      "entropy": 5.400880479812622,
      "epoch": 4.550233054416802,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003109866540608556,
      "loss": 4.987,
      "mean_token_accuracy": 0.2195681020617485,
      "num_tokens": 90472988.0,
      "step": 40515
    },
    {
      "entropy": 5.422921991348266,
      "epoch": 4.55079463132476,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031094702161857455,
      "loss": 5.0262,
      "mean_token_accuracy": 0.22243448048830033,
      "num_tokens": 90484737.0,
      "step": 40520
    },
    {
      "entropy": 5.2891241073608395,
      "epoch": 4.5513562082327175,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031090738803172646,
      "loss": 4.7905,
      "mean_token_accuracy": 0.23159411400556565,
      "num_tokens": 90496015.0,
      "step": 40525
    },
    {
      "entropy": 5.314245319366455,
      "epoch": 4.551917785140675,
      "grad_norm": 0.96875,
      "learning_rate": 0.00031086775330157324,
      "loss": 4.8422,
      "mean_token_accuracy": 0.23718664795160294,
      "num_tokens": 90507709.0,
      "step": 40530
    },
    {
      "entropy": 5.309314727783203,
      "epoch": 4.552479362048633,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003108281174293768,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2331332355737686,
      "num_tokens": 90520167.0,
      "step": 40535
    },
    {
      "entropy": 5.279068565368652,
      "epoch": 4.55304093895659,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031078848041639926,
      "loss": 4.8861,
      "mean_token_accuracy": 0.23162511140108108,
      "num_tokens": 90531797.0,
      "step": 40540
    },
    {
      "entropy": 5.392660665512085,
      "epoch": 4.553602515864547,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031074884226390246,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22987184375524522,
      "num_tokens": 90542201.0,
      "step": 40545
    },
    {
      "entropy": 5.359649515151977,
      "epoch": 4.554164092772505,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003107092029731488,
      "loss": 4.8886,
      "mean_token_accuracy": 0.23862247914075851,
      "num_tokens": 90553906.0,
      "step": 40550
    },
    {
      "entropy": 5.227881956100464,
      "epoch": 4.5547256696804626,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031066956254540025,
      "loss": 4.7347,
      "mean_token_accuracy": 0.23676453828811644,
      "num_tokens": 90564839.0,
      "step": 40555
    },
    {
      "entropy": 5.382263612747193,
      "epoch": 4.555287246588421,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031062992098191897,
      "loss": 4.9635,
      "mean_token_accuracy": 0.2199525222182274,
      "num_tokens": 90575252.0,
      "step": 40560
    },
    {
      "entropy": 5.2524830341339115,
      "epoch": 4.555848823496378,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003105902782839672,
      "loss": 4.802,
      "mean_token_accuracy": 0.23877596259117126,
      "num_tokens": 90585676.0,
      "step": 40565
    },
    {
      "entropy": 5.338509941101075,
      "epoch": 4.556410400404335,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003105506344528071,
      "loss": 4.9928,
      "mean_token_accuracy": 0.2235603451728821,
      "num_tokens": 90597453.0,
      "step": 40570
    },
    {
      "entropy": 5.314253330230713,
      "epoch": 4.556971977312293,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000310510989489701,
      "loss": 4.8579,
      "mean_token_accuracy": 0.23660089522600175,
      "num_tokens": 90608318.0,
      "step": 40575
    },
    {
      "entropy": 5.2612395763397215,
      "epoch": 4.55753355422025,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031047134339591136,
      "loss": 4.8328,
      "mean_token_accuracy": 0.23319352567195892,
      "num_tokens": 90618627.0,
      "step": 40580
    },
    {
      "entropy": 5.291187429428101,
      "epoch": 4.558095131128208,
      "grad_norm": 1.0,
      "learning_rate": 0.00031043169617270033,
      "loss": 4.8885,
      "mean_token_accuracy": 0.22365162074565886,
      "num_tokens": 90630150.0,
      "step": 40585
    },
    {
      "entropy": 5.297217941284179,
      "epoch": 4.558656708036166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031039204782133033,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23420062363147737,
      "num_tokens": 90640882.0,
      "step": 40590
    },
    {
      "entropy": 5.3987876892089846,
      "epoch": 4.559218284944123,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003103523983430638,
      "loss": 4.974,
      "mean_token_accuracy": 0.22469801008701323,
      "num_tokens": 90652473.0,
      "step": 40595
    },
    {
      "entropy": 5.304210233688354,
      "epoch": 4.55977986185208,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031031274773916326,
      "loss": 4.902,
      "mean_token_accuracy": 0.231027489900589,
      "num_tokens": 90663547.0,
      "step": 40600
    },
    {
      "entropy": 5.325047206878662,
      "epoch": 4.560341438760038,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031027309601089104,
      "loss": 4.8957,
      "mean_token_accuracy": 0.2282556712627411,
      "num_tokens": 90675093.0,
      "step": 40605
    },
    {
      "entropy": 5.340421915054321,
      "epoch": 4.5609030156679955,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003102334431595098,
      "loss": 4.8321,
      "mean_token_accuracy": 0.22638730257749556,
      "num_tokens": 90687648.0,
      "step": 40610
    },
    {
      "entropy": 5.365055513381958,
      "epoch": 4.561464592575954,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031019378918628203,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2288220167160034,
      "num_tokens": 90698075.0,
      "step": 40615
    },
    {
      "entropy": 5.262012147903443,
      "epoch": 4.562026169483911,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031015413409247046,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22785797715187073,
      "num_tokens": 90708864.0,
      "step": 40620
    },
    {
      "entropy": 5.378125715255737,
      "epoch": 4.562587746391868,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031011447787933756,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2218271091580391,
      "num_tokens": 90719789.0,
      "step": 40625
    },
    {
      "entropy": 5.350632429122925,
      "epoch": 4.563149323299826,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003100748205481461,
      "loss": 4.9655,
      "mean_token_accuracy": 0.21992011666297911,
      "num_tokens": 90732643.0,
      "step": 40630
    },
    {
      "entropy": 5.300983238220215,
      "epoch": 4.563710900207783,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003100351621001587,
      "loss": 4.7888,
      "mean_token_accuracy": 0.24004290103912354,
      "num_tokens": 90742412.0,
      "step": 40635
    },
    {
      "entropy": 5.336408042907715,
      "epoch": 4.5642724771157415,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003099955025366382,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22885123044252395,
      "num_tokens": 90753372.0,
      "step": 40640
    },
    {
      "entropy": 5.24726414680481,
      "epoch": 4.564834054023699,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00030995584185884733,
      "loss": 4.7906,
      "mean_token_accuracy": 0.23425825387239457,
      "num_tokens": 90763517.0,
      "step": 40645
    },
    {
      "entropy": 5.294766807556153,
      "epoch": 4.565395630931656,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003099161800680488,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22073613107204437,
      "num_tokens": 90773870.0,
      "step": 40650
    },
    {
      "entropy": 5.253192043304443,
      "epoch": 4.565957207839614,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003098765171655056,
      "loss": 4.7924,
      "mean_token_accuracy": 0.2357379451394081,
      "num_tokens": 90784423.0,
      "step": 40655
    },
    {
      "entropy": 5.358261966705323,
      "epoch": 4.566518784747571,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003098368531524806,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22183917313814164,
      "num_tokens": 90795332.0,
      "step": 40660
    },
    {
      "entropy": 5.292052745819092,
      "epoch": 4.567080361655528,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003097971880302367,
      "loss": 4.8101,
      "mean_token_accuracy": 0.2326342284679413,
      "num_tokens": 90805995.0,
      "step": 40665
    },
    {
      "entropy": 5.220079612731934,
      "epoch": 4.5676419385634865,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00030975752180003674,
      "loss": 4.8226,
      "mean_token_accuracy": 0.23068981617689133,
      "num_tokens": 90817925.0,
      "step": 40670
    },
    {
      "entropy": 5.2120285987854,
      "epoch": 4.568203515471444,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000309717854463144,
      "loss": 4.8221,
      "mean_token_accuracy": 0.23476998507976532,
      "num_tokens": 90829042.0,
      "step": 40675
    },
    {
      "entropy": 5.400878286361694,
      "epoch": 4.568765092379401,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003096781860208211,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23612628728151322,
      "num_tokens": 90840559.0,
      "step": 40680
    },
    {
      "entropy": 5.379316902160644,
      "epoch": 4.569326669287359,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003096385164743314,
      "loss": 4.8379,
      "mean_token_accuracy": 0.24086399227380753,
      "num_tokens": 90851009.0,
      "step": 40685
    },
    {
      "entropy": 5.3228983879089355,
      "epoch": 4.569888246195316,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030959884582493793,
      "loss": 4.8992,
      "mean_token_accuracy": 0.23333416134119034,
      "num_tokens": 90861698.0,
      "step": 40690
    },
    {
      "entropy": 5.184554624557495,
      "epoch": 4.570449823103274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003095591740739038,
      "loss": 4.8339,
      "mean_token_accuracy": 0.23177196383476256,
      "num_tokens": 90871745.0,
      "step": 40695
    },
    {
      "entropy": 5.343463516235351,
      "epoch": 4.571011400011232,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003095195012224921,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2241009756922722,
      "num_tokens": 90882401.0,
      "step": 40700
    },
    {
      "entropy": 5.326072359085083,
      "epoch": 4.571572976919189,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003094798272719661,
      "loss": 4.7055,
      "mean_token_accuracy": 0.258037306368351,
      "num_tokens": 90895024.0,
      "step": 40705
    },
    {
      "entropy": 5.185869550704956,
      "epoch": 4.572134553827147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030944015222358906,
      "loss": 4.7491,
      "mean_token_accuracy": 0.2403952047228813,
      "num_tokens": 90906414.0,
      "step": 40710
    },
    {
      "entropy": 5.260491752624512,
      "epoch": 4.572696130735104,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030940047607862415,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23044947683811187,
      "num_tokens": 90916950.0,
      "step": 40715
    },
    {
      "entropy": 5.361950874328613,
      "epoch": 4.573257707643061,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003093607988383348,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22891452312469482,
      "num_tokens": 90928721.0,
      "step": 40720
    },
    {
      "entropy": 5.387753868103028,
      "epoch": 4.5738192845510195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003093211205039842,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2205940991640091,
      "num_tokens": 90940768.0,
      "step": 40725
    },
    {
      "entropy": 5.366475200653076,
      "epoch": 4.574380861458977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003092814410768359,
      "loss": 4.8285,
      "mean_token_accuracy": 0.23287745714187622,
      "num_tokens": 90952822.0,
      "step": 40730
    },
    {
      "entropy": 5.391913270950317,
      "epoch": 4.574942438366934,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030924176055815327,
      "loss": 4.9342,
      "mean_token_accuracy": 0.23330748975276946,
      "num_tokens": 90964380.0,
      "step": 40735
    },
    {
      "entropy": 5.3014203071594235,
      "epoch": 4.575504015274892,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030920207894919957,
      "loss": 4.8333,
      "mean_token_accuracy": 0.22261991053819657,
      "num_tokens": 90975597.0,
      "step": 40740
    },
    {
      "entropy": 5.262994909286499,
      "epoch": 4.576065592182849,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003091623962512385,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22532882690429687,
      "num_tokens": 90989451.0,
      "step": 40745
    },
    {
      "entropy": 5.304627418518066,
      "epoch": 4.576627169090807,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00030912271246553346,
      "loss": 4.8262,
      "mean_token_accuracy": 0.23263555616140366,
      "num_tokens": 91001367.0,
      "step": 40750
    },
    {
      "entropy": 5.285611629486084,
      "epoch": 4.577188745998765,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000309083027593348,
      "loss": 4.8536,
      "mean_token_accuracy": 0.22785632610321044,
      "num_tokens": 91012951.0,
      "step": 40755
    },
    {
      "entropy": 5.312336874008179,
      "epoch": 4.577750322906722,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030904334163594577,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2307036966085434,
      "num_tokens": 91024400.0,
      "step": 40760
    },
    {
      "entropy": 5.408748435974121,
      "epoch": 4.57831189981468,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003090036545945904,
      "loss": 4.9505,
      "mean_token_accuracy": 0.22507032752037048,
      "num_tokens": 91035802.0,
      "step": 40765
    },
    {
      "entropy": 5.324979972839356,
      "epoch": 4.578873476722637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030896396647054535,
      "loss": 4.8842,
      "mean_token_accuracy": 0.23229270577430725,
      "num_tokens": 91046546.0,
      "step": 40770
    },
    {
      "entropy": 5.271066331863404,
      "epoch": 4.579435053630594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003089242772650744,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22811909168958663,
      "num_tokens": 91057000.0,
      "step": 40775
    },
    {
      "entropy": 5.441847038269043,
      "epoch": 4.579996630538552,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030888458697944144,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21965378671884536,
      "num_tokens": 91067910.0,
      "step": 40780
    },
    {
      "entropy": 5.487842416763305,
      "epoch": 4.58055820744651,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030884489561491005,
      "loss": 4.9852,
      "mean_token_accuracy": 0.22289651334285737,
      "num_tokens": 91078914.0,
      "step": 40785
    },
    {
      "entropy": 5.313452100753784,
      "epoch": 4.581119784354467,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000308805203172744,
      "loss": 4.8844,
      "mean_token_accuracy": 0.23024605661630632,
      "num_tokens": 91089448.0,
      "step": 40790
    },
    {
      "entropy": 5.273844957351685,
      "epoch": 4.581681361262425,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030876550965420725,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22996511906385422,
      "num_tokens": 91099533.0,
      "step": 40795
    },
    {
      "entropy": 5.386033630371093,
      "epoch": 4.582242938170382,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003087258150605635,
      "loss": 4.913,
      "mean_token_accuracy": 0.22366792261600493,
      "num_tokens": 91111926.0,
      "step": 40800
    },
    {
      "entropy": 5.405502557754517,
      "epoch": 4.58280451507834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003086861193930767,
      "loss": 4.8721,
      "mean_token_accuracy": 0.22700327336788179,
      "num_tokens": 91122782.0,
      "step": 40805
    },
    {
      "entropy": 5.332604122161865,
      "epoch": 4.5833660919862975,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003086464226530109,
      "loss": 4.8946,
      "mean_token_accuracy": 0.22845852822065355,
      "num_tokens": 91133626.0,
      "step": 40810
    },
    {
      "entropy": 5.2289458274841305,
      "epoch": 4.583927668894255,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030860672484162987,
      "loss": 4.77,
      "mean_token_accuracy": 0.23419603556394578,
      "num_tokens": 91143362.0,
      "step": 40815
    },
    {
      "entropy": 5.272774934768677,
      "epoch": 4.584489245802213,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003085670259601977,
      "loss": 4.79,
      "mean_token_accuracy": 0.23499622493982314,
      "num_tokens": 91154603.0,
      "step": 40820
    },
    {
      "entropy": 5.385566091537475,
      "epoch": 4.58505082271017,
      "grad_norm": 0.953125,
      "learning_rate": 0.00030852732600997846,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2247173622250557,
      "num_tokens": 91166750.0,
      "step": 40825
    },
    {
      "entropy": 5.35115795135498,
      "epoch": 4.585612399618128,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030848762499223616,
      "loss": 4.8027,
      "mean_token_accuracy": 0.2329890936613083,
      "num_tokens": 91176042.0,
      "step": 40830
    },
    {
      "entropy": 5.321157169342041,
      "epoch": 4.586173976526085,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003084479229082348,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22412883788347243,
      "num_tokens": 91187895.0,
      "step": 40835
    },
    {
      "entropy": 5.291553211212158,
      "epoch": 4.586735553434043,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003084082197592387,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22413255870342255,
      "num_tokens": 91198740.0,
      "step": 40840
    },
    {
      "entropy": 5.277764701843262,
      "epoch": 4.587297130342001,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030836851554651184,
      "loss": 4.844,
      "mean_token_accuracy": 0.22957988530397416,
      "num_tokens": 91209663.0,
      "step": 40845
    },
    {
      "entropy": 5.363742065429688,
      "epoch": 4.587858707249958,
      "grad_norm": 1.0,
      "learning_rate": 0.00030832881027131854,
      "loss": 4.8936,
      "mean_token_accuracy": 0.23161114752292633,
      "num_tokens": 91221277.0,
      "step": 40850
    },
    {
      "entropy": 5.306375646591187,
      "epoch": 4.588420284157915,
      "grad_norm": 1.0,
      "learning_rate": 0.00030828910393492306,
      "loss": 4.8602,
      "mean_token_accuracy": 0.2298295110464096,
      "num_tokens": 91232265.0,
      "step": 40855
    },
    {
      "entropy": 5.290942907333374,
      "epoch": 4.588981861065873,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030824939653858953,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22888291776180267,
      "num_tokens": 91243935.0,
      "step": 40860
    },
    {
      "entropy": 5.398237037658691,
      "epoch": 4.5895434379738305,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030820968808358223,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2222931370139122,
      "num_tokens": 91253760.0,
      "step": 40865
    },
    {
      "entropy": 5.336049747467041,
      "epoch": 4.590105014881788,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030816997857116573,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2296506568789482,
      "num_tokens": 91265777.0,
      "step": 40870
    },
    {
      "entropy": 5.354352760314941,
      "epoch": 4.590666591789746,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003081302680026042,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2320181354880333,
      "num_tokens": 91276439.0,
      "step": 40875
    },
    {
      "entropy": 5.376702070236206,
      "epoch": 4.591228168697703,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000308090556379162,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22121851593255998,
      "num_tokens": 91286330.0,
      "step": 40880
    },
    {
      "entropy": 5.290316724777222,
      "epoch": 4.591789745605661,
      "grad_norm": 1.0,
      "learning_rate": 0.00030805084370210377,
      "loss": 4.7859,
      "mean_token_accuracy": 0.22543356269598008,
      "num_tokens": 91296547.0,
      "step": 40885
    },
    {
      "entropy": 5.279999113082885,
      "epoch": 4.592351322513618,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00030801112997269386,
      "loss": 4.9268,
      "mean_token_accuracy": 0.22081861793994903,
      "num_tokens": 91309412.0,
      "step": 40890
    },
    {
      "entropy": 5.2870221614837645,
      "epoch": 4.5929128994215755,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003079714151921967,
      "loss": 4.789,
      "mean_token_accuracy": 0.23933058828115464,
      "num_tokens": 91319939.0,
      "step": 40895
    },
    {
      "entropy": 5.351127576828003,
      "epoch": 4.593474476329534,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00030793169936187695,
      "loss": 4.953,
      "mean_token_accuracy": 0.2269701838493347,
      "num_tokens": 91331579.0,
      "step": 40900
    },
    {
      "entropy": 5.338293409347534,
      "epoch": 4.594036053237491,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030789198248299906,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22708901464939119,
      "num_tokens": 91344365.0,
      "step": 40905
    },
    {
      "entropy": 5.330491781234741,
      "epoch": 4.594597630145448,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030785226455682773,
      "loss": 4.8527,
      "mean_token_accuracy": 0.2328613057732582,
      "num_tokens": 91354701.0,
      "step": 40910
    },
    {
      "entropy": 5.266106414794922,
      "epoch": 4.595159207053406,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003078125455846276,
      "loss": 4.7734,
      "mean_token_accuracy": 0.23115283697843553,
      "num_tokens": 91365052.0,
      "step": 40915
    },
    {
      "entropy": 5.287390232086182,
      "epoch": 4.595720783961363,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030777282556766316,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22954680472612382,
      "num_tokens": 91376399.0,
      "step": 40920
    },
    {
      "entropy": 5.284762620925903,
      "epoch": 4.596282360869321,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00030773310450719937,
      "loss": 4.7835,
      "mean_token_accuracy": 0.2330758184194565,
      "num_tokens": 91388042.0,
      "step": 40925
    },
    {
      "entropy": 5.305308866500854,
      "epoch": 4.596843937777279,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003076933824045008,
      "loss": 4.8654,
      "mean_token_accuracy": 0.23301891535520552,
      "num_tokens": 91400167.0,
      "step": 40930
    },
    {
      "entropy": 5.3915486335754395,
      "epoch": 4.597405514685236,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030765365926083214,
      "loss": 5.0179,
      "mean_token_accuracy": 0.22003554701805114,
      "num_tokens": 91412335.0,
      "step": 40935
    },
    {
      "entropy": 5.307770299911499,
      "epoch": 4.597967091593194,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003076139350774584,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22711893320083618,
      "num_tokens": 91423879.0,
      "step": 40940
    },
    {
      "entropy": 5.330508375167847,
      "epoch": 4.598528668501151,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003075742098556444,
      "loss": 4.878,
      "mean_token_accuracy": 0.23118707090616225,
      "num_tokens": 91434104.0,
      "step": 40945
    },
    {
      "entropy": 5.348813819885254,
      "epoch": 4.5990902454091085,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003075344835966548,
      "loss": 4.9018,
      "mean_token_accuracy": 0.2258077308535576,
      "num_tokens": 91445019.0,
      "step": 40950
    },
    {
      "entropy": 5.38912992477417,
      "epoch": 4.599651822317067,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003074947563017547,
      "loss": 4.916,
      "mean_token_accuracy": 0.23124136477708818,
      "num_tokens": 91457100.0,
      "step": 40955
    },
    {
      "entropy": 5.333905029296875,
      "epoch": 4.600213399225024,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003074550279722089,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22788846641778945,
      "num_tokens": 91468354.0,
      "step": 40960
    },
    {
      "entropy": 5.368706130981446,
      "epoch": 4.600774976132981,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003074152986092825,
      "loss": 4.904,
      "mean_token_accuracy": 0.2223096787929535,
      "num_tokens": 91480017.0,
      "step": 40965
    },
    {
      "entropy": 5.396568965911865,
      "epoch": 4.601336553040939,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003073755682142404,
      "loss": 4.9691,
      "mean_token_accuracy": 0.22397394478321075,
      "num_tokens": 91492387.0,
      "step": 40970
    },
    {
      "entropy": 5.2795093059539795,
      "epoch": 4.601898129948896,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030733583678834765,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22602591514587403,
      "num_tokens": 91504095.0,
      "step": 40975
    },
    {
      "entropy": 5.353981018066406,
      "epoch": 4.602459706856854,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030729610433286926,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22745669931173323,
      "num_tokens": 91515298.0,
      "step": 40980
    },
    {
      "entropy": 5.3391400337219235,
      "epoch": 4.603021283764812,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030725637084907037,
      "loss": 4.8146,
      "mean_token_accuracy": 0.2257561355829239,
      "num_tokens": 91525337.0,
      "step": 40985
    },
    {
      "entropy": 5.336915397644043,
      "epoch": 4.603582860672769,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003072166363382162,
      "loss": 4.9205,
      "mean_token_accuracy": 0.23640895783901214,
      "num_tokens": 91536555.0,
      "step": 40990
    },
    {
      "entropy": 5.362166452407837,
      "epoch": 4.604144437580727,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003071769008015717,
      "loss": 4.974,
      "mean_token_accuracy": 0.221169313788414,
      "num_tokens": 91546726.0,
      "step": 40995
    },
    {
      "entropy": 5.352299976348877,
      "epoch": 4.604706014488684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003071371642404024,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22462376058101655,
      "num_tokens": 91556509.0,
      "step": 41000
    },
    {
      "entropy": 5.353421449661255,
      "epoch": 4.605267591396641,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003070974266559732,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2244385302066803,
      "num_tokens": 91567144.0,
      "step": 41005
    },
    {
      "entropy": 5.325935792922974,
      "epoch": 4.6058291683045995,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003070576880495495,
      "loss": 4.8477,
      "mean_token_accuracy": 0.2386883243918419,
      "num_tokens": 91577901.0,
      "step": 41010
    },
    {
      "entropy": 5.374489164352417,
      "epoch": 4.606390745212557,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003070179484223966,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22217340767383575,
      "num_tokens": 91588071.0,
      "step": 41015
    },
    {
      "entropy": 5.343785095214844,
      "epoch": 4.606952322120515,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003069782077757797,
      "loss": 4.9874,
      "mean_token_accuracy": 0.22150601893663407,
      "num_tokens": 91599378.0,
      "step": 41020
    },
    {
      "entropy": 5.336264562606812,
      "epoch": 4.607513899028472,
      "grad_norm": 1.0,
      "learning_rate": 0.00030693846611096437,
      "loss": 4.8386,
      "mean_token_accuracy": 0.23400854468345642,
      "num_tokens": 91609551.0,
      "step": 41025
    },
    {
      "entropy": 5.399582624435425,
      "epoch": 4.608075475936429,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030689872342921595,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22892725467681885,
      "num_tokens": 91619440.0,
      "step": 41030
    },
    {
      "entropy": 5.301528549194336,
      "epoch": 4.608637052844387,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030685897973179967,
      "loss": 4.8723,
      "mean_token_accuracy": 0.23370989859104158,
      "num_tokens": 91629992.0,
      "step": 41035
    },
    {
      "entropy": 5.365511465072632,
      "epoch": 4.609198629752345,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030681923501998124,
      "loss": 4.9344,
      "mean_token_accuracy": 0.23325403183698654,
      "num_tokens": 91640874.0,
      "step": 41040
    },
    {
      "entropy": 5.3338416576385494,
      "epoch": 4.609760206660302,
      "grad_norm": 1.0,
      "learning_rate": 0.00030677948929502596,
      "loss": 4.8342,
      "mean_token_accuracy": 0.24024593383073806,
      "num_tokens": 91651427.0,
      "step": 41045
    },
    {
      "entropy": 5.2986249923706055,
      "epoch": 4.61032178356826,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00030673974255819944,
      "loss": 4.8458,
      "mean_token_accuracy": 0.22677867710590363,
      "num_tokens": 91662799.0,
      "step": 41050
    },
    {
      "entropy": 5.414012241363525,
      "epoch": 4.610883360476217,
      "grad_norm": 1.0,
      "learning_rate": 0.0003066999948107672,
      "loss": 5.0032,
      "mean_token_accuracy": 0.22121190577745437,
      "num_tokens": 91673334.0,
      "step": 41055
    },
    {
      "entropy": 5.288450908660889,
      "epoch": 4.611444937384174,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003066602460539948,
      "loss": 4.8713,
      "mean_token_accuracy": 0.22822053134441375,
      "num_tokens": 91684098.0,
      "step": 41060
    },
    {
      "entropy": 5.289797401428222,
      "epoch": 4.6120065142921325,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030662049628914787,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22806766927242278,
      "num_tokens": 91696100.0,
      "step": 41065
    },
    {
      "entropy": 5.267695426940918,
      "epoch": 4.61256809120009,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003065807455174921,
      "loss": 4.8,
      "mean_token_accuracy": 0.22593198716640472,
      "num_tokens": 91707890.0,
      "step": 41070
    },
    {
      "entropy": 5.356400966644287,
      "epoch": 4.613129668108048,
      "grad_norm": 1.0,
      "learning_rate": 0.0003065409937402932,
      "loss": 4.9534,
      "mean_token_accuracy": 0.22129171192646027,
      "num_tokens": 91718311.0,
      "step": 41075
    },
    {
      "entropy": 5.304121732711792,
      "epoch": 4.613691245016005,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00030650124095881674,
      "loss": 4.8808,
      "mean_token_accuracy": 0.231160768866539,
      "num_tokens": 91729859.0,
      "step": 41080
    },
    {
      "entropy": 5.322818565368652,
      "epoch": 4.614252821923962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030646148717432855,
      "loss": 4.8719,
      "mean_token_accuracy": 0.23038736283779143,
      "num_tokens": 91741174.0,
      "step": 41085
    },
    {
      "entropy": 5.367798900604248,
      "epoch": 4.61481439883192,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030642173238809453,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22266789078712462,
      "num_tokens": 91752858.0,
      "step": 41090
    },
    {
      "entropy": 5.321347570419311,
      "epoch": 4.615375975739878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003063819766013802,
      "loss": 4.8427,
      "mean_token_accuracy": 0.2348729968070984,
      "num_tokens": 91762819.0,
      "step": 41095
    },
    {
      "entropy": 5.334008502960205,
      "epoch": 4.615937552647835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003063422198154516,
      "loss": 4.8768,
      "mean_token_accuracy": 0.23042566925287247,
      "num_tokens": 91773838.0,
      "step": 41100
    },
    {
      "entropy": 5.277377796173096,
      "epoch": 4.616499129555793,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003063024620315746,
      "loss": 4.8235,
      "mean_token_accuracy": 0.23049089908599854,
      "num_tokens": 91784660.0,
      "step": 41105
    },
    {
      "entropy": 5.341361331939697,
      "epoch": 4.61706070646375,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003062627032510151,
      "loss": 4.9454,
      "mean_token_accuracy": 0.21505187600851058,
      "num_tokens": 91795368.0,
      "step": 41110
    },
    {
      "entropy": 5.417652225494384,
      "epoch": 4.617622283371707,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000306222943475039,
      "loss": 4.9624,
      "mean_token_accuracy": 0.23059315234422684,
      "num_tokens": 91805280.0,
      "step": 41115
    },
    {
      "entropy": 5.399887037277222,
      "epoch": 4.618183860279665,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003061831827049123,
      "loss": 5.0135,
      "mean_token_accuracy": 0.22373663187026976,
      "num_tokens": 91816130.0,
      "step": 41120
    },
    {
      "entropy": 5.4240926742553714,
      "epoch": 4.618745437187623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030614342094190095,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2261948063969612,
      "num_tokens": 91827513.0,
      "step": 41125
    },
    {
      "entropy": 5.3655694961547855,
      "epoch": 4.619307014095581,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030610365818727097,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22804270088672637,
      "num_tokens": 91838284.0,
      "step": 41130
    },
    {
      "entropy": 5.316544675827027,
      "epoch": 4.619868591003538,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003060638944422884,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22981217205524446,
      "num_tokens": 91848801.0,
      "step": 41135
    },
    {
      "entropy": 5.292419815063477,
      "epoch": 4.620430167911495,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00030602412970821944,
      "loss": 4.9014,
      "mean_token_accuracy": 0.2280547708272934,
      "num_tokens": 91860740.0,
      "step": 41140
    },
    {
      "entropy": 5.375370216369629,
      "epoch": 4.620991744819453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003059843639863301,
      "loss": 4.9566,
      "mean_token_accuracy": 0.222332464158535,
      "num_tokens": 91872020.0,
      "step": 41145
    },
    {
      "entropy": 5.3890430450439455,
      "epoch": 4.6215533217274105,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030594459727788665,
      "loss": 4.901,
      "mean_token_accuracy": 0.22961317747831345,
      "num_tokens": 91882100.0,
      "step": 41150
    },
    {
      "entropy": 5.363692569732666,
      "epoch": 4.622114898635368,
      "grad_norm": 0.90625,
      "learning_rate": 0.00030590482958415524,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22560274600982666,
      "num_tokens": 91894014.0,
      "step": 41155
    },
    {
      "entropy": 5.34701623916626,
      "epoch": 4.622676475543326,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030586506090640205,
      "loss": 4.8628,
      "mean_token_accuracy": 0.2359864130616188,
      "num_tokens": 91905759.0,
      "step": 41160
    },
    {
      "entropy": 5.411273765563965,
      "epoch": 4.623238052451283,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003058252912458933,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2238863155245781,
      "num_tokens": 91916833.0,
      "step": 41165
    },
    {
      "entropy": 5.285590600967407,
      "epoch": 4.62379962935924,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003057855206038954,
      "loss": 4.7766,
      "mean_token_accuracy": 0.2355724960565567,
      "num_tokens": 91927672.0,
      "step": 41170
    },
    {
      "entropy": 5.1173614978790285,
      "epoch": 4.624361206267198,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003057457489816745,
      "loss": 4.7266,
      "mean_token_accuracy": 0.2390070453286171,
      "num_tokens": 91938283.0,
      "step": 41175
    },
    {
      "entropy": 5.258405876159668,
      "epoch": 4.624922783175156,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003057059763804971,
      "loss": 4.8582,
      "mean_token_accuracy": 0.2263607934117317,
      "num_tokens": 91948764.0,
      "step": 41180
    },
    {
      "entropy": 5.344851398468018,
      "epoch": 4.625484360083114,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003056662028016295,
      "loss": 4.9874,
      "mean_token_accuracy": 0.22238798588514327,
      "num_tokens": 91960110.0,
      "step": 41185
    },
    {
      "entropy": 5.2690558433532715,
      "epoch": 4.626045936991071,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00030562642824633805,
      "loss": 4.7058,
      "mean_token_accuracy": 0.2409820333123207,
      "num_tokens": 91971084.0,
      "step": 41190
    },
    {
      "entropy": 5.332602167129517,
      "epoch": 4.626607513899028,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003055866527158893,
      "loss": 4.8789,
      "mean_token_accuracy": 0.23012007772922516,
      "num_tokens": 91983373.0,
      "step": 41195
    },
    {
      "entropy": 5.2721282005310055,
      "epoch": 4.627169090806986,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003055468762115497,
      "loss": 4.8905,
      "mean_token_accuracy": 0.23763176649808884,
      "num_tokens": 91995000.0,
      "step": 41200
    },
    {
      "entropy": 5.26688323020935,
      "epoch": 4.6277306677149435,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00030550709873458567,
      "loss": 4.8143,
      "mean_token_accuracy": 0.23237926810979842,
      "num_tokens": 92007486.0,
      "step": 41205
    },
    {
      "entropy": 5.425273418426514,
      "epoch": 4.628292244622902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030546732028626373,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22674815207719803,
      "num_tokens": 92018415.0,
      "step": 41210
    },
    {
      "entropy": 5.357337379455567,
      "epoch": 4.628853821530859,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030542754086785055,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2338629424571991,
      "num_tokens": 92030463.0,
      "step": 41215
    },
    {
      "entropy": 5.3753925323486325,
      "epoch": 4.629415398438816,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003053877604806127,
      "loss": 4.9898,
      "mean_token_accuracy": 0.21690915077924727,
      "num_tokens": 92042284.0,
      "step": 41220
    },
    {
      "entropy": 5.4021471500396725,
      "epoch": 4.629976975346774,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030534797912581664,
      "loss": 4.9351,
      "mean_token_accuracy": 0.23072191178798676,
      "num_tokens": 92052719.0,
      "step": 41225
    },
    {
      "entropy": 5.251800012588501,
      "epoch": 4.630538552254731,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030530819680472926,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22958330512046815,
      "num_tokens": 92062986.0,
      "step": 41230
    },
    {
      "entropy": 5.373140621185303,
      "epoch": 4.6311001291626885,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030526841351861717,
      "loss": 4.8384,
      "mean_token_accuracy": 0.2295661225914955,
      "num_tokens": 92073192.0,
      "step": 41235
    },
    {
      "entropy": 5.326973962783813,
      "epoch": 4.631661706070647,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030522862926874697,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23133132606744766,
      "num_tokens": 92084304.0,
      "step": 41240
    },
    {
      "entropy": 5.375598669052124,
      "epoch": 4.632223282978604,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003051888440563855,
      "loss": 4.9113,
      "mean_token_accuracy": 0.22796494513750076,
      "num_tokens": 92095660.0,
      "step": 41245
    },
    {
      "entropy": 5.3735638618469235,
      "epoch": 4.632784859886561,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003051490578827996,
      "loss": 4.9798,
      "mean_token_accuracy": 0.2177684262394905,
      "num_tokens": 92107822.0,
      "step": 41250
    },
    {
      "entropy": 5.436110973358154,
      "epoch": 4.633346436794519,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00030510927074925594,
      "loss": 5.0038,
      "mean_token_accuracy": 0.21619817912578582,
      "num_tokens": 92119054.0,
      "step": 41255
    },
    {
      "entropy": 5.2532689571380615,
      "epoch": 4.633908013702476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030506948265702144,
      "loss": 4.7733,
      "mean_token_accuracy": 0.23836730122566224,
      "num_tokens": 92130007.0,
      "step": 41260
    },
    {
      "entropy": 5.285567045211792,
      "epoch": 4.6344695906104345,
      "grad_norm": 1.015625,
      "learning_rate": 0.000305029693607363,
      "loss": 4.871,
      "mean_token_accuracy": 0.2373387858271599,
      "num_tokens": 92140059.0,
      "step": 41265
    },
    {
      "entropy": 5.27090744972229,
      "epoch": 4.635031167518392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030498990360154746,
      "loss": 4.8625,
      "mean_token_accuracy": 0.2302174985408783,
      "num_tokens": 92151024.0,
      "step": 41270
    },
    {
      "entropy": 5.303188896179199,
      "epoch": 4.635592744426349,
      "grad_norm": 0.953125,
      "learning_rate": 0.00030495011264084165,
      "loss": 4.8778,
      "mean_token_accuracy": 0.233488130569458,
      "num_tokens": 92162799.0,
      "step": 41275
    },
    {
      "entropy": 5.2864104270935055,
      "epoch": 4.636154321334307,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030491032072651266,
      "loss": 4.8333,
      "mean_token_accuracy": 0.2376761704683304,
      "num_tokens": 92173772.0,
      "step": 41280
    },
    {
      "entropy": 5.328063106536865,
      "epoch": 4.636715898242264,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030487052785982757,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23393264561891555,
      "num_tokens": 92184365.0,
      "step": 41285
    },
    {
      "entropy": 5.360870695114135,
      "epoch": 4.6372774751502215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003048307340420532,
      "loss": 4.9841,
      "mean_token_accuracy": 0.2175539642572403,
      "num_tokens": 92196457.0,
      "step": 41290
    },
    {
      "entropy": 5.282446622848511,
      "epoch": 4.63783905205818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003047909392744566,
      "loss": 4.8165,
      "mean_token_accuracy": 0.23035632818937302,
      "num_tokens": 92206885.0,
      "step": 41295
    },
    {
      "entropy": 5.367913389205933,
      "epoch": 4.638400628966137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030475114355830513,
      "loss": 4.8733,
      "mean_token_accuracy": 0.2326648160815239,
      "num_tokens": 92218488.0,
      "step": 41300
    },
    {
      "entropy": 5.279191637039185,
      "epoch": 4.638962205874094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030471134689486563,
      "loss": 4.7892,
      "mean_token_accuracy": 0.23354291319847106,
      "num_tokens": 92230439.0,
      "step": 41305
    },
    {
      "entropy": 5.32328987121582,
      "epoch": 4.639523782782052,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003046715492854053,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2314096674323082,
      "num_tokens": 92240299.0,
      "step": 41310
    },
    {
      "entropy": 5.338251447677612,
      "epoch": 4.640085359690009,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030463175073119133,
      "loss": 4.9715,
      "mean_token_accuracy": 0.22575788348913192,
      "num_tokens": 92252867.0,
      "step": 41315
    },
    {
      "entropy": 5.311400127410889,
      "epoch": 4.6406469365979675,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030459195123349095,
      "loss": 4.8708,
      "mean_token_accuracy": 0.22691351622343064,
      "num_tokens": 92263960.0,
      "step": 41320
    },
    {
      "entropy": 5.281176424026489,
      "epoch": 4.641208513505925,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003045521507935714,
      "loss": 4.845,
      "mean_token_accuracy": 0.23200464397668838,
      "num_tokens": 92276179.0,
      "step": 41325
    },
    {
      "entropy": 5.3044219493865965,
      "epoch": 4.641770090413882,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003045123494126999,
      "loss": 4.798,
      "mean_token_accuracy": 0.23357588797807693,
      "num_tokens": 92286474.0,
      "step": 41330
    },
    {
      "entropy": 5.295082378387451,
      "epoch": 4.64233166732184,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003044725470921437,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22580238580703735,
      "num_tokens": 92297611.0,
      "step": 41335
    },
    {
      "entropy": 5.293924999237061,
      "epoch": 4.642893244229797,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003044327438331702,
      "loss": 4.8159,
      "mean_token_accuracy": 0.2321597844362259,
      "num_tokens": 92308882.0,
      "step": 41340
    },
    {
      "entropy": 5.290621137619018,
      "epoch": 4.643454821137754,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003043929396370467,
      "loss": 4.8014,
      "mean_token_accuracy": 0.23470281511545182,
      "num_tokens": 92320439.0,
      "step": 41345
    },
    {
      "entropy": 5.3831390857696535,
      "epoch": 4.6440163980457125,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00030435313450504056,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22893356680870056,
      "num_tokens": 92330550.0,
      "step": 41350
    },
    {
      "entropy": 5.265456962585449,
      "epoch": 4.64457797495367,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003043133284384193,
      "loss": 4.8254,
      "mean_token_accuracy": 0.23077475279569626,
      "num_tokens": 92341079.0,
      "step": 41355
    },
    {
      "entropy": 5.331912612915039,
      "epoch": 4.645139551861627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003042735214384503,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2233578383922577,
      "num_tokens": 92351995.0,
      "step": 41360
    },
    {
      "entropy": 5.284488344192505,
      "epoch": 4.645701128769585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000304233713506401,
      "loss": 4.855,
      "mean_token_accuracy": 0.2372998610138893,
      "num_tokens": 92362329.0,
      "step": 41365
    },
    {
      "entropy": 5.267970371246338,
      "epoch": 4.646262705677542,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003041939046435389,
      "loss": 4.8076,
      "mean_token_accuracy": 0.2321912929415703,
      "num_tokens": 92372662.0,
      "step": 41370
    },
    {
      "entropy": 5.414710283279419,
      "epoch": 4.6468242825855,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030415409485113156,
      "loss": 5.0801,
      "mean_token_accuracy": 0.21449503302574158,
      "num_tokens": 92382727.0,
      "step": 41375
    },
    {
      "entropy": 5.406077146530151,
      "epoch": 4.647385859493458,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030411428413044657,
      "loss": 4.88,
      "mean_token_accuracy": 0.22908592075109482,
      "num_tokens": 92394034.0,
      "step": 41380
    },
    {
      "entropy": 5.31969633102417,
      "epoch": 4.647947436401415,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00030407447248275137,
      "loss": 4.8812,
      "mean_token_accuracy": 0.2302614524960518,
      "num_tokens": 92404652.0,
      "step": 41385
    },
    {
      "entropy": 5.266722106933594,
      "epoch": 4.648509013309373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003040346599093138,
      "loss": 4.8002,
      "mean_token_accuracy": 0.23170799911022186,
      "num_tokens": 92415220.0,
      "step": 41390
    },
    {
      "entropy": 5.1992593765258786,
      "epoch": 4.64907059021733,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003039948464114014,
      "loss": 4.7576,
      "mean_token_accuracy": 0.23910584747791291,
      "num_tokens": 92425584.0,
      "step": 41395
    },
    {
      "entropy": 5.372399234771729,
      "epoch": 4.649632167125288,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030395503199028186,
      "loss": 4.918,
      "mean_token_accuracy": 0.23162589073181153,
      "num_tokens": 92436298.0,
      "step": 41400
    },
    {
      "entropy": 5.2505110740661625,
      "epoch": 4.6501937440332455,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003039152166472228,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22725290507078172,
      "num_tokens": 92446778.0,
      "step": 41405
    },
    {
      "entropy": 5.314688396453858,
      "epoch": 4.650755320941203,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003038754003834921,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22880032062530517,
      "num_tokens": 92457922.0,
      "step": 41410
    },
    {
      "entropy": 5.348233461380005,
      "epoch": 4.651316897849161,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030383558320035744,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22798738330602647,
      "num_tokens": 92470155.0,
      "step": 41415
    },
    {
      "entropy": 5.369081640243531,
      "epoch": 4.651878474757118,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00030379576509908656,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22573762238025666,
      "num_tokens": 92480862.0,
      "step": 41420
    },
    {
      "entropy": 5.342126369476318,
      "epoch": 4.652440051665075,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030375594608094737,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22417790740728377,
      "num_tokens": 92491867.0,
      "step": 41425
    },
    {
      "entropy": 5.328060436248779,
      "epoch": 4.653001628573033,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003037161261472078,
      "loss": 4.9169,
      "mean_token_accuracy": 0.23198655545711516,
      "num_tokens": 92503104.0,
      "step": 41430
    },
    {
      "entropy": 5.243779850006104,
      "epoch": 4.653563205480991,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00030367630529913554,
      "loss": 4.834,
      "mean_token_accuracy": 0.22172652333974838,
      "num_tokens": 92514825.0,
      "step": 41435
    },
    {
      "entropy": 5.329059886932373,
      "epoch": 4.654124782388948,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030363648353799857,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22627724558115006,
      "num_tokens": 92525764.0,
      "step": 41440
    },
    {
      "entropy": 5.382903814315796,
      "epoch": 4.654686359296906,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030359666086506493,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2252148285508156,
      "num_tokens": 92536025.0,
      "step": 41445
    },
    {
      "entropy": 5.473212862014771,
      "epoch": 4.655247936204863,
      "grad_norm": 0.9375,
      "learning_rate": 0.00030355683728160247,
      "loss": 5.0373,
      "mean_token_accuracy": 0.21357335448265075,
      "num_tokens": 92547917.0,
      "step": 41450
    },
    {
      "entropy": 5.347277116775513,
      "epoch": 4.655809513112821,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003035170127888793,
      "loss": 4.8632,
      "mean_token_accuracy": 0.23422294110059738,
      "num_tokens": 92558637.0,
      "step": 41455
    },
    {
      "entropy": 5.366068506240845,
      "epoch": 4.656371090020778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030347718738816333,
      "loss": 4.887,
      "mean_token_accuracy": 0.2264869973063469,
      "num_tokens": 92570110.0,
      "step": 41460
    },
    {
      "entropy": 5.3393982410430905,
      "epoch": 4.656932666928736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030343736108072266,
      "loss": 4.9377,
      "mean_token_accuracy": 0.229929219186306,
      "num_tokens": 92579974.0,
      "step": 41465
    },
    {
      "entropy": 5.312402963638306,
      "epoch": 4.657494243836694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003033975338678254,
      "loss": 4.9021,
      "mean_token_accuracy": 0.2270752504467964,
      "num_tokens": 92590202.0,
      "step": 41470
    },
    {
      "entropy": 5.315021419525147,
      "epoch": 4.658055820744651,
      "grad_norm": 0.9375,
      "learning_rate": 0.00030335770575073967,
      "loss": 4.8415,
      "mean_token_accuracy": 0.23187289983034134,
      "num_tokens": 92601525.0,
      "step": 41475
    },
    {
      "entropy": 5.3786262512207035,
      "epoch": 4.658617397652608,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003033178767307336,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22999364733695984,
      "num_tokens": 92611055.0,
      "step": 41480
    },
    {
      "entropy": 5.343531465530395,
      "epoch": 4.659178974560566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003032780468090753,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22033920139074326,
      "num_tokens": 92621337.0,
      "step": 41485
    },
    {
      "entropy": 5.230831623077393,
      "epoch": 4.6597405514685235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003032382159870331,
      "loss": 4.7953,
      "mean_token_accuracy": 0.23890607208013534,
      "num_tokens": 92632174.0,
      "step": 41490
    },
    {
      "entropy": 5.323486042022705,
      "epoch": 4.660302128376481,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00030319838426587505,
      "loss": 4.877,
      "mean_token_accuracy": 0.22932358533143998,
      "num_tokens": 92643667.0,
      "step": 41495
    },
    {
      "entropy": 5.353532075881958,
      "epoch": 4.660863705284439,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030315855164686956,
      "loss": 4.8935,
      "mean_token_accuracy": 0.23287750035524368,
      "num_tokens": 92654945.0,
      "step": 41500
    },
    {
      "entropy": 5.32953462600708,
      "epoch": 4.661425282192396,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003031187181312848,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2186361402273178,
      "num_tokens": 92666136.0,
      "step": 41505
    },
    {
      "entropy": 5.372505855560303,
      "epoch": 4.661986859100354,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030307888372038927,
      "loss": 4.8085,
      "mean_token_accuracy": 0.23596761822700502,
      "num_tokens": 92676683.0,
      "step": 41510
    },
    {
      "entropy": 5.360294771194458,
      "epoch": 4.662548436008311,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030303904841545113,
      "loss": 4.9098,
      "mean_token_accuracy": 0.22208456844091415,
      "num_tokens": 92687313.0,
      "step": 41515
    },
    {
      "entropy": 5.292513227462768,
      "epoch": 4.663110012916269,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003029992122177388,
      "loss": 4.8253,
      "mean_token_accuracy": 0.2298249512910843,
      "num_tokens": 92699760.0,
      "step": 41520
    },
    {
      "entropy": 5.360175466537475,
      "epoch": 4.663671589824227,
      "grad_norm": 1.0,
      "learning_rate": 0.00030295937512852075,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22484273314476014,
      "num_tokens": 92710893.0,
      "step": 41525
    },
    {
      "entropy": 5.280935907363892,
      "epoch": 4.664233166732184,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003029195371490652,
      "loss": 4.8445,
      "mean_token_accuracy": 0.23224566578865052,
      "num_tokens": 92722992.0,
      "step": 41530
    },
    {
      "entropy": 5.26893048286438,
      "epoch": 4.664794743640142,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00030287969828064087,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2277648627758026,
      "num_tokens": 92734084.0,
      "step": 41535
    },
    {
      "entropy": 5.309473371505737,
      "epoch": 4.665356320548099,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00030283985852451624,
      "loss": 4.851,
      "mean_token_accuracy": 0.23623912632465363,
      "num_tokens": 92745666.0,
      "step": 41540
    },
    {
      "entropy": 5.324749994277954,
      "epoch": 4.6659178974560565,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003028000178819596,
      "loss": 4.8436,
      "mean_token_accuracy": 0.23884017765522003,
      "num_tokens": 92756403.0,
      "step": 41545
    },
    {
      "entropy": 5.295525646209716,
      "epoch": 4.666479474364014,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030276017635423965,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22796893417835234,
      "num_tokens": 92767838.0,
      "step": 41550
    },
    {
      "entropy": 5.3345067501068115,
      "epoch": 4.667041051271972,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030272033394262483,
      "loss": 4.834,
      "mean_token_accuracy": 0.23367543667554855,
      "num_tokens": 92779515.0,
      "step": 41555
    },
    {
      "entropy": 5.286647129058838,
      "epoch": 4.667602628179929,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030268049064838395,
      "loss": 4.8136,
      "mean_token_accuracy": 0.23452157378196717,
      "num_tokens": 92790965.0,
      "step": 41560
    },
    {
      "entropy": 5.340078210830688,
      "epoch": 4.668164205087887,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030264064647278537,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22167731523513795,
      "num_tokens": 92802059.0,
      "step": 41565
    },
    {
      "entropy": 5.350535154342651,
      "epoch": 4.668725781995844,
      "grad_norm": 1.03125,
      "learning_rate": 0.000302600801417098,
      "loss": 4.9167,
      "mean_token_accuracy": 0.226625494658947,
      "num_tokens": 92814190.0,
      "step": 41570
    },
    {
      "entropy": 5.338301801681519,
      "epoch": 4.6692873589038015,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00030256095548259035,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2266266882419586,
      "num_tokens": 92825247.0,
      "step": 41575
    },
    {
      "entropy": 5.411154270172119,
      "epoch": 4.66984893581176,
      "grad_norm": 1.125,
      "learning_rate": 0.0003025211086705312,
      "loss": 4.9026,
      "mean_token_accuracy": 0.22971309274435042,
      "num_tokens": 92835155.0,
      "step": 41580
    },
    {
      "entropy": 5.3078932762146,
      "epoch": 4.670410512719717,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003024812609821893,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22828188836574553,
      "num_tokens": 92846357.0,
      "step": 41585
    },
    {
      "entropy": 5.304434871673584,
      "epoch": 4.670972089627675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003024414124188333,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23285400122404099,
      "num_tokens": 92856438.0,
      "step": 41590
    },
    {
      "entropy": 5.409440422058106,
      "epoch": 4.671533666535632,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003024015629817322,
      "loss": 4.9675,
      "mean_token_accuracy": 0.2270009458065033,
      "num_tokens": 92867748.0,
      "step": 41595
    },
    {
      "entropy": 5.276484870910645,
      "epoch": 4.672095243443589,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003023617126721547,
      "loss": 4.7571,
      "mean_token_accuracy": 0.23525290340185165,
      "num_tokens": 92879566.0,
      "step": 41600
    },
    {
      "entropy": 5.331803178787231,
      "epoch": 4.6726568203515475,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003023218614913696,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23209280520677567,
      "num_tokens": 92891659.0,
      "step": 41605
    },
    {
      "entropy": 5.319676065444947,
      "epoch": 4.673218397259505,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003022820094406458,
      "loss": 4.8791,
      "mean_token_accuracy": 0.23009195327758789,
      "num_tokens": 92903076.0,
      "step": 41610
    },
    {
      "entropy": 5.264736366271973,
      "epoch": 4.673779974167462,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003022421565212524,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2326295256614685,
      "num_tokens": 92913087.0,
      "step": 41615
    },
    {
      "entropy": 5.306989049911499,
      "epoch": 4.67434155107542,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00030220230273445806,
      "loss": 4.8349,
      "mean_token_accuracy": 0.23474291414022447,
      "num_tokens": 92924063.0,
      "step": 41620
    },
    {
      "entropy": 5.400181293487549,
      "epoch": 4.674903127983377,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030216244808153186,
      "loss": 5.063,
      "mean_token_accuracy": 0.21812464892864228,
      "num_tokens": 92934795.0,
      "step": 41625
    },
    {
      "entropy": 5.317264747619629,
      "epoch": 4.6754647048913345,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003021225925637428,
      "loss": 4.8503,
      "mean_token_accuracy": 0.2366244837641716,
      "num_tokens": 92946171.0,
      "step": 41630
    },
    {
      "entropy": 5.398207187652588,
      "epoch": 4.676026281799293,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003020827361823598,
      "loss": 4.955,
      "mean_token_accuracy": 0.22650111168622972,
      "num_tokens": 92957131.0,
      "step": 41635
    },
    {
      "entropy": 5.35333743095398,
      "epoch": 4.67658785870725,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030204287893865204,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2263099581003189,
      "num_tokens": 92968982.0,
      "step": 41640
    },
    {
      "entropy": 5.325126266479492,
      "epoch": 4.677149435615208,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030200302083388856,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22723960727453232,
      "num_tokens": 92980378.0,
      "step": 41645
    },
    {
      "entropy": 5.380585956573486,
      "epoch": 4.677711012523165,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00030196316186933834,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22924090921878815,
      "num_tokens": 92991820.0,
      "step": 41650
    },
    {
      "entropy": 5.4138611316680905,
      "epoch": 4.678272589431122,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030192330204627054,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22421641647815704,
      "num_tokens": 93003832.0,
      "step": 41655
    },
    {
      "entropy": 5.354798793792725,
      "epoch": 4.6788341663390804,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003018834413659545,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22609520256519317,
      "num_tokens": 93015750.0,
      "step": 41660
    },
    {
      "entropy": 5.351752996444702,
      "epoch": 4.679395743247038,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003018435798296592,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2255645677447319,
      "num_tokens": 93025355.0,
      "step": 41665
    },
    {
      "entropy": 5.325830698013306,
      "epoch": 4.679957320154995,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003018037174386539,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22123912125825881,
      "num_tokens": 93037138.0,
      "step": 41670
    },
    {
      "entropy": 5.327441263198852,
      "epoch": 4.680518897062953,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003017638541942078,
      "loss": 4.8878,
      "mean_token_accuracy": 0.23112013936042786,
      "num_tokens": 93048766.0,
      "step": 41675
    },
    {
      "entropy": 5.28716082572937,
      "epoch": 4.68108047397091,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030172399009759034,
      "loss": 4.8252,
      "mean_token_accuracy": 0.22574735432863235,
      "num_tokens": 93059439.0,
      "step": 41680
    },
    {
      "entropy": 5.304604578018188,
      "epoch": 4.681642050878867,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003016841251500705,
      "loss": 4.8387,
      "mean_token_accuracy": 0.2357617884874344,
      "num_tokens": 93070039.0,
      "step": 41685
    },
    {
      "entropy": 5.285853242874145,
      "epoch": 4.6822036277868255,
      "grad_norm": 1.0,
      "learning_rate": 0.00030164425935291794,
      "loss": 4.9417,
      "mean_token_accuracy": 0.22601652294397354,
      "num_tokens": 93082001.0,
      "step": 41690
    },
    {
      "entropy": 5.42135181427002,
      "epoch": 4.682765204694783,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003016043927074018,
      "loss": 4.9895,
      "mean_token_accuracy": 0.225245001912117,
      "num_tokens": 93093360.0,
      "step": 41695
    },
    {
      "entropy": 5.35796046257019,
      "epoch": 4.683326781602741,
      "grad_norm": 1.0,
      "learning_rate": 0.0003015645252147914,
      "loss": 4.7945,
      "mean_token_accuracy": 0.2374746784567833,
      "num_tokens": 93103377.0,
      "step": 41700
    },
    {
      "entropy": 5.295522546768188,
      "epoch": 4.683888358510698,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003015246568763562,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23456366807222367,
      "num_tokens": 93114289.0,
      "step": 41705
    },
    {
      "entropy": 5.3129456520080565,
      "epoch": 4.684449935418655,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030148478769336564,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23312475979328157,
      "num_tokens": 93125846.0,
      "step": 41710
    },
    {
      "entropy": 5.270769691467285,
      "epoch": 4.685011512326613,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003014449176670892,
      "loss": 4.8852,
      "mean_token_accuracy": 0.22353301644325257,
      "num_tokens": 93135552.0,
      "step": 41715
    },
    {
      "entropy": 5.454571676254273,
      "epoch": 4.685573089234571,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030140504679879637,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2142038345336914,
      "num_tokens": 93146797.0,
      "step": 41720
    },
    {
      "entropy": 5.237526512145996,
      "epoch": 4.686134666142529,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003013651750897566,
      "loss": 4.7761,
      "mean_token_accuracy": 0.23331716507673264,
      "num_tokens": 93159153.0,
      "step": 41725
    },
    {
      "entropy": 5.312166452407837,
      "epoch": 4.686696243050486,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003013253025412393,
      "loss": 4.833,
      "mean_token_accuracy": 0.23308446556329726,
      "num_tokens": 93170015.0,
      "step": 41730
    },
    {
      "entropy": 5.340100860595703,
      "epoch": 4.687257819958443,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030128542915451427,
      "loss": 4.913,
      "mean_token_accuracy": 0.23216505944728852,
      "num_tokens": 93181869.0,
      "step": 41735
    },
    {
      "entropy": 5.3710884094238285,
      "epoch": 4.6878193968664,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030124555493085095,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22505066990852357,
      "num_tokens": 93193771.0,
      "step": 41740
    },
    {
      "entropy": 5.432128667831421,
      "epoch": 4.6883809737743585,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003012056798715189,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21850201189517976,
      "num_tokens": 93205554.0,
      "step": 41745
    },
    {
      "entropy": 5.346857976913452,
      "epoch": 4.688942550682316,
      "grad_norm": 1.015625,
      "learning_rate": 0.000301165803977788,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2319184809923172,
      "num_tokens": 93217168.0,
      "step": 41750
    },
    {
      "entropy": 5.374003505706787,
      "epoch": 4.689504127590274,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003011259272509277,
      "loss": 4.9202,
      "mean_token_accuracy": 0.229892098903656,
      "num_tokens": 93227668.0,
      "step": 41755
    },
    {
      "entropy": 5.3652167320251465,
      "epoch": 4.690065704498231,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003010860496922077,
      "loss": 4.8713,
      "mean_token_accuracy": 0.22822520285844802,
      "num_tokens": 93239310.0,
      "step": 41760
    },
    {
      "entropy": 5.323997545242309,
      "epoch": 4.690627281406188,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030104617130289783,
      "loss": 4.858,
      "mean_token_accuracy": 0.22342328876256942,
      "num_tokens": 93249887.0,
      "step": 41765
    },
    {
      "entropy": 5.2810447216033936,
      "epoch": 4.691188858314146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003010062920842677,
      "loss": 4.9197,
      "mean_token_accuracy": 0.2253631681203842,
      "num_tokens": 93260129.0,
      "step": 41770
    },
    {
      "entropy": 5.352622127532959,
      "epoch": 4.6917504352221036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030096641203758716,
      "loss": 4.9143,
      "mean_token_accuracy": 0.23084412813186644,
      "num_tokens": 93271171.0,
      "step": 41775
    },
    {
      "entropy": 5.283004665374756,
      "epoch": 4.692312012130062,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00030092653116412605,
      "loss": 4.8017,
      "mean_token_accuracy": 0.23713499754667283,
      "num_tokens": 93282535.0,
      "step": 41780
    },
    {
      "entropy": 5.302482604980469,
      "epoch": 4.692873589038019,
      "grad_norm": 1.109375,
      "learning_rate": 0.000300886649465154,
      "loss": 4.8497,
      "mean_token_accuracy": 0.2360524892807007,
      "num_tokens": 93293317.0,
      "step": 41785
    },
    {
      "entropy": 5.34490704536438,
      "epoch": 4.693435165945976,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003008467669419411,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2291066452860832,
      "num_tokens": 93304180.0,
      "step": 41790
    },
    {
      "entropy": 5.327944326400757,
      "epoch": 4.693996742853934,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003008068835957571,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2264651983976364,
      "num_tokens": 93315146.0,
      "step": 41795
    },
    {
      "entropy": 5.425363397598266,
      "epoch": 4.694558319761891,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003007669994278719,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22421457320451738,
      "num_tokens": 93325756.0,
      "step": 41800
    },
    {
      "entropy": 5.316203022003174,
      "epoch": 4.695119896669849,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003007271144395554,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22993295639753342,
      "num_tokens": 93337653.0,
      "step": 41805
    },
    {
      "entropy": 5.353049802780151,
      "epoch": 4.695681473577807,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003006872286320777,
      "loss": 4.9492,
      "mean_token_accuracy": 0.23363907784223556,
      "num_tokens": 93348751.0,
      "step": 41810
    },
    {
      "entropy": 5.25953254699707,
      "epoch": 4.696243050485764,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003006473420067086,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22484659999608994,
      "num_tokens": 93359704.0,
      "step": 41815
    },
    {
      "entropy": 5.316268968582153,
      "epoch": 4.696804627393721,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003006074545647182,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23668383806943893,
      "num_tokens": 93369252.0,
      "step": 41820
    },
    {
      "entropy": 5.3241295337677,
      "epoch": 4.697366204301679,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030056756630737655,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2251626282930374,
      "num_tokens": 93380384.0,
      "step": 41825
    },
    {
      "entropy": 5.3698451042175295,
      "epoch": 4.6979277812096365,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00030052767723595365,
      "loss": 4.887,
      "mean_token_accuracy": 0.2305729791522026,
      "num_tokens": 93390994.0,
      "step": 41830
    },
    {
      "entropy": 5.342101049423218,
      "epoch": 4.698489358117595,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003004877873517196,
      "loss": 4.8499,
      "mean_token_accuracy": 0.23364374041557312,
      "num_tokens": 93401828.0,
      "step": 41835
    },
    {
      "entropy": 5.402885866165161,
      "epoch": 4.699050935025552,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030044789665594444,
      "loss": 5.0001,
      "mean_token_accuracy": 0.22452054619789125,
      "num_tokens": 93414653.0,
      "step": 41840
    },
    {
      "entropy": 5.2787576675415036,
      "epoch": 4.699612511933509,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003004080051498985,
      "loss": 4.8345,
      "mean_token_accuracy": 0.23452488481998443,
      "num_tokens": 93426020.0,
      "step": 41845
    },
    {
      "entropy": 5.299846601486206,
      "epoch": 4.700174088841467,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030036811283485184,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2297765865921974,
      "num_tokens": 93436992.0,
      "step": 41850
    },
    {
      "entropy": 5.317082262039184,
      "epoch": 4.700735665749424,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00030032821971207453,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21899321526288987,
      "num_tokens": 93448747.0,
      "step": 41855
    },
    {
      "entropy": 5.381482839584351,
      "epoch": 4.701297242657382,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003002883257828369,
      "loss": 4.9389,
      "mean_token_accuracy": 0.23237981051206588,
      "num_tokens": 93460349.0,
      "step": 41860
    },
    {
      "entropy": 5.407089948654175,
      "epoch": 4.70185881956534,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030024843104840924,
      "loss": 4.989,
      "mean_token_accuracy": 0.22065868228673935,
      "num_tokens": 93470465.0,
      "step": 41865
    },
    {
      "entropy": 5.394988584518432,
      "epoch": 4.702420396473297,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030020853551006164,
      "loss": 4.9026,
      "mean_token_accuracy": 0.23135693967342377,
      "num_tokens": 93482409.0,
      "step": 41870
    },
    {
      "entropy": 5.284492874145508,
      "epoch": 4.702981973381254,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003001686391690645,
      "loss": 4.818,
      "mean_token_accuracy": 0.24273510426282882,
      "num_tokens": 93492212.0,
      "step": 41875
    },
    {
      "entropy": 5.298301553726196,
      "epoch": 4.703543550289212,
      "grad_norm": 1.0,
      "learning_rate": 0.00030012874202668823,
      "loss": 4.8962,
      "mean_token_accuracy": 0.23534810096025466,
      "num_tokens": 93503343.0,
      "step": 41880
    },
    {
      "entropy": 5.3711999416351315,
      "epoch": 4.7041051271971694,
      "grad_norm": 1.0,
      "learning_rate": 0.00030008884408420293,
      "loss": 4.9583,
      "mean_token_accuracy": 0.22147030681371688,
      "num_tokens": 93514401.0,
      "step": 41885
    },
    {
      "entropy": 5.452882242202759,
      "epoch": 4.7046667041051276,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003000489453428792,
      "loss": 4.9597,
      "mean_token_accuracy": 0.23541709631681443,
      "num_tokens": 93527096.0,
      "step": 41890
    },
    {
      "entropy": 5.371601676940918,
      "epoch": 4.705228281013085,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003000090458039873,
      "loss": 4.8317,
      "mean_token_accuracy": 0.23409150242805482,
      "num_tokens": 93538349.0,
      "step": 41895
    },
    {
      "entropy": 5.183589696884155,
      "epoch": 4.705789857921042,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029996914546879773,
      "loss": 4.768,
      "mean_token_accuracy": 0.23753306418657302,
      "num_tokens": 93548883.0,
      "step": 41900
    },
    {
      "entropy": 5.258791399002075,
      "epoch": 4.706351434829,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002999292443385809,
      "loss": 4.852,
      "mean_token_accuracy": 0.22709605544805528,
      "num_tokens": 93559796.0,
      "step": 41905
    },
    {
      "entropy": 5.397894620895386,
      "epoch": 4.706913011736957,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002998893424146072,
      "loss": 4.8575,
      "mean_token_accuracy": 0.23156777918338775,
      "num_tokens": 93571083.0,
      "step": 41910
    },
    {
      "entropy": 5.28754997253418,
      "epoch": 4.707474588644915,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002998494396981472,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22676147371530533,
      "num_tokens": 93583559.0,
      "step": 41915
    },
    {
      "entropy": 5.32647819519043,
      "epoch": 4.708036165552873,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002998095361904714,
      "loss": 4.8658,
      "mean_token_accuracy": 0.2334922432899475,
      "num_tokens": 93594419.0,
      "step": 41920
    },
    {
      "entropy": 5.334461879730225,
      "epoch": 4.70859774246083,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002997696318928503,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2304815411567688,
      "num_tokens": 93605140.0,
      "step": 41925
    },
    {
      "entropy": 5.318619060516357,
      "epoch": 4.709159319368787,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002997297268065546,
      "loss": 4.877,
      "mean_token_accuracy": 0.23193345516920089,
      "num_tokens": 93617572.0,
      "step": 41930
    },
    {
      "entropy": 5.303957176208496,
      "epoch": 4.709720896276745,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029968982093285475,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22686406522989272,
      "num_tokens": 93627511.0,
      "step": 41935
    },
    {
      "entropy": 5.286347246170044,
      "epoch": 4.710282473184702,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029964991427302147,
      "loss": 4.8851,
      "mean_token_accuracy": 0.23438214063644408,
      "num_tokens": 93637651.0,
      "step": 41940
    },
    {
      "entropy": 5.295897436141968,
      "epoch": 4.7108440500926605,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002996100068283253,
      "loss": 4.8505,
      "mean_token_accuracy": 0.23737146556377411,
      "num_tokens": 93648832.0,
      "step": 41945
    },
    {
      "entropy": 5.262290954589844,
      "epoch": 4.711405627000618,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029957009860003697,
      "loss": 4.8141,
      "mean_token_accuracy": 0.2357766166329384,
      "num_tokens": 93660639.0,
      "step": 41950
    },
    {
      "entropy": 5.35911283493042,
      "epoch": 4.711967203908575,
      "grad_norm": 1.15625,
      "learning_rate": 0.00029953018958942717,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22842550575733184,
      "num_tokens": 93671296.0,
      "step": 41955
    },
    {
      "entropy": 5.291916036605835,
      "epoch": 4.712528780816533,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002994902797977667,
      "loss": 4.8187,
      "mean_token_accuracy": 0.23807838559150696,
      "num_tokens": 93682428.0,
      "step": 41960
    },
    {
      "entropy": 5.259958410263062,
      "epoch": 4.71309035772449,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029945036922632615,
      "loss": 4.842,
      "mean_token_accuracy": 0.232601660490036,
      "num_tokens": 93693543.0,
      "step": 41965
    },
    {
      "entropy": 5.3224828243255615,
      "epoch": 4.713651934632448,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029941045787637636,
      "loss": 4.8239,
      "mean_token_accuracy": 0.23139526844024658,
      "num_tokens": 93704834.0,
      "step": 41970
    },
    {
      "entropy": 5.442697763442993,
      "epoch": 4.714213511540406,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002993705457491882,
      "loss": 5.0471,
      "mean_token_accuracy": 0.2174109324812889,
      "num_tokens": 93717291.0,
      "step": 41975
    },
    {
      "entropy": 5.415569543838501,
      "epoch": 4.714775088448363,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029933063284603235,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2260613813996315,
      "num_tokens": 93729412.0,
      "step": 41980
    },
    {
      "entropy": 5.366035318374633,
      "epoch": 4.715336665356321,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002992907191681797,
      "loss": 4.8344,
      "mean_token_accuracy": 0.23486850261688233,
      "num_tokens": 93740197.0,
      "step": 41985
    },
    {
      "entropy": 5.301580572128296,
      "epoch": 4.715898242264278,
      "grad_norm": 0.953125,
      "learning_rate": 0.00029925080471690117,
      "loss": 4.902,
      "mean_token_accuracy": 0.22497205734252929,
      "num_tokens": 93751915.0,
      "step": 41990
    },
    {
      "entropy": 5.38932614326477,
      "epoch": 4.716459819172235,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029921088949346765,
      "loss": 4.9796,
      "mean_token_accuracy": 0.22377128154039383,
      "num_tokens": 93763392.0,
      "step": 41995
    },
    {
      "entropy": 5.340775060653686,
      "epoch": 4.717021396080193,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002991709734991499,
      "loss": 4.7913,
      "mean_token_accuracy": 0.23970773369073867,
      "num_tokens": 93774207.0,
      "step": 42000
    },
    {
      "epoch": 4.717021396080193,
      "eval_entropy": 5.141952825324388,
      "eval_loss": 5.041710376739502,
      "eval_mean_token_accuracy": 0.22828397447873183,
      "eval_num_tokens": 93774207.0,
      "eval_runtime": 23.8024,
      "eval_samples_per_second": 1302.938,
      "eval_steps_per_second": 162.883,
      "step": 42000
    },
    {
      "entropy": 5.39470477104187,
      "epoch": 4.717582972988151,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029913105673521907,
      "loss": 4.9703,
      "mean_token_accuracy": 0.22448041290044785,
      "num_tokens": 93785599.0,
      "step": 42005
    },
    {
      "entropy": 5.2145013332366945,
      "epoch": 4.718144549896108,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000299091139202946,
      "loss": 4.7363,
      "mean_token_accuracy": 0.24043385088443756,
      "num_tokens": 93796942.0,
      "step": 42010
    },
    {
      "entropy": 5.3697656154632565,
      "epoch": 4.718706126804066,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029905122090360167,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2190815970301628,
      "num_tokens": 93808858.0,
      "step": 42015
    },
    {
      "entropy": 5.314112758636474,
      "epoch": 4.719267703712023,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002990113018384572,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2302725210785866,
      "num_tokens": 93820817.0,
      "step": 42020
    },
    {
      "entropy": 5.281701755523682,
      "epoch": 4.719829280619981,
      "grad_norm": 1.0,
      "learning_rate": 0.0002989713820087836,
      "loss": 4.7886,
      "mean_token_accuracy": 0.23238065242767333,
      "num_tokens": 93831615.0,
      "step": 42025
    },
    {
      "entropy": 5.341869258880616,
      "epoch": 4.7203908575279385,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029893146141585187,
      "loss": 4.8675,
      "mean_token_accuracy": 0.22723120152950288,
      "num_tokens": 93842622.0,
      "step": 42030
    },
    {
      "entropy": 5.262009954452514,
      "epoch": 4.720952434435896,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029889154006093306,
      "loss": 4.8536,
      "mean_token_accuracy": 0.2291353315114975,
      "num_tokens": 93852409.0,
      "step": 42035
    },
    {
      "entropy": 5.301942205429077,
      "epoch": 4.721514011343854,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002988516179452984,
      "loss": 4.9363,
      "mean_token_accuracy": 0.23218819499015808,
      "num_tokens": 93863428.0,
      "step": 42040
    },
    {
      "entropy": 5.314897441864014,
      "epoch": 4.722075588251811,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029881169507021895,
      "loss": 4.9018,
      "mean_token_accuracy": 0.23433074057102204,
      "num_tokens": 93874561.0,
      "step": 42045
    },
    {
      "entropy": 5.251282453536987,
      "epoch": 4.722637165159768,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000298771771436966,
      "loss": 4.802,
      "mean_token_accuracy": 0.23917344361543655,
      "num_tokens": 93885923.0,
      "step": 42050
    },
    {
      "entropy": 5.181094694137573,
      "epoch": 4.723198742067726,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002987318470468105,
      "loss": 4.6631,
      "mean_token_accuracy": 0.2396415427327156,
      "num_tokens": 93897329.0,
      "step": 42055
    },
    {
      "entropy": 5.202418422698974,
      "epoch": 4.723760318975684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029869192190102385,
      "loss": 4.7645,
      "mean_token_accuracy": 0.23552075326442717,
      "num_tokens": 93908198.0,
      "step": 42060
    },
    {
      "entropy": 5.319151878356934,
      "epoch": 4.724321895883641,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029865199600087724,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2210560604929924,
      "num_tokens": 93919368.0,
      "step": 42065
    },
    {
      "entropy": 5.34576416015625,
      "epoch": 4.724883472791599,
      "grad_norm": 1.125,
      "learning_rate": 0.0002986120693476418,
      "loss": 4.9211,
      "mean_token_accuracy": 0.22232408970594406,
      "num_tokens": 93930585.0,
      "step": 42070
    },
    {
      "entropy": 5.345205307006836,
      "epoch": 4.725445049699556,
      "grad_norm": 1.0,
      "learning_rate": 0.00029857214194258896,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22531274408102037,
      "num_tokens": 93941990.0,
      "step": 42075
    },
    {
      "entropy": 5.329469394683838,
      "epoch": 4.726006626607514,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029853221378699,
      "loss": 4.9047,
      "mean_token_accuracy": 0.22552771866321564,
      "num_tokens": 93953219.0,
      "step": 42080
    },
    {
      "entropy": 5.274214076995849,
      "epoch": 4.7265682035154715,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002984922848821162,
      "loss": 4.823,
      "mean_token_accuracy": 0.2345086842775345,
      "num_tokens": 93964858.0,
      "step": 42085
    },
    {
      "entropy": 5.187658500671387,
      "epoch": 4.727129780423429,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029845235522923903,
      "loss": 4.7565,
      "mean_token_accuracy": 0.23073508143424987,
      "num_tokens": 93975120.0,
      "step": 42090
    },
    {
      "entropy": 5.237909746170044,
      "epoch": 4.727691357331387,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029841242482962966,
      "loss": 4.8265,
      "mean_token_accuracy": 0.23450278341770173,
      "num_tokens": 93985546.0,
      "step": 42095
    },
    {
      "entropy": 5.323454189300537,
      "epoch": 4.728252934239344,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00029837249368455963,
      "loss": 4.8678,
      "mean_token_accuracy": 0.23242430239915848,
      "num_tokens": 93996116.0,
      "step": 42100
    },
    {
      "entropy": 5.320510053634644,
      "epoch": 4.728814511147302,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002983325617953004,
      "loss": 4.8626,
      "mean_token_accuracy": 0.23077965825796126,
      "num_tokens": 94006531.0,
      "step": 42105
    },
    {
      "entropy": 5.259982442855835,
      "epoch": 4.729376088055259,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029829262916312326,
      "loss": 4.8003,
      "mean_token_accuracy": 0.24292090088129042,
      "num_tokens": 94017924.0,
      "step": 42110
    },
    {
      "entropy": 5.321058988571167,
      "epoch": 4.7299376649632165,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00029825269578929986,
      "loss": 4.87,
      "mean_token_accuracy": 0.23641814291477203,
      "num_tokens": 94029115.0,
      "step": 42115
    },
    {
      "entropy": 5.3869048118591305,
      "epoch": 4.730499241871174,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029821276167510153,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22516782879829406,
      "num_tokens": 94041223.0,
      "step": 42120
    },
    {
      "entropy": 5.414372730255127,
      "epoch": 4.731060818779132,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002981728268217999,
      "loss": 4.9759,
      "mean_token_accuracy": 0.21641747206449508,
      "num_tokens": 94054091.0,
      "step": 42125
    },
    {
      "entropy": 5.326960849761963,
      "epoch": 4.731622395687089,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002981328912306665,
      "loss": 4.879,
      "mean_token_accuracy": 0.23538580685853958,
      "num_tokens": 94064988.0,
      "step": 42130
    },
    {
      "entropy": 5.325798892974854,
      "epoch": 4.732183972595047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029809295490297285,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22631106227636338,
      "num_tokens": 94075553.0,
      "step": 42135
    },
    {
      "entropy": 5.282143783569336,
      "epoch": 4.732745549503004,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002980530178399906,
      "loss": 4.8269,
      "mean_token_accuracy": 0.23100461214780807,
      "num_tokens": 94086245.0,
      "step": 42140
    },
    {
      "entropy": 5.406793260574341,
      "epoch": 4.733307126410962,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002980130800429912,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22933005839586257,
      "num_tokens": 94097146.0,
      "step": 42145
    },
    {
      "entropy": 5.287923526763916,
      "epoch": 4.73386870331892,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029797314151324653,
      "loss": 4.8178,
      "mean_token_accuracy": 0.22874056547880173,
      "num_tokens": 94108046.0,
      "step": 42150
    },
    {
      "entropy": 5.317944669723511,
      "epoch": 4.734430280226877,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00029793320225202807,
      "loss": 4.8556,
      "mean_token_accuracy": 0.2271113336086273,
      "num_tokens": 94118859.0,
      "step": 42155
    },
    {
      "entropy": 5.321755456924438,
      "epoch": 4.734991857134835,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002978932622606075,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2293991133570671,
      "num_tokens": 94130593.0,
      "step": 42160
    },
    {
      "entropy": 5.346603870391846,
      "epoch": 4.735553434042792,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00029785332154025676,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2274392992258072,
      "num_tokens": 94142684.0,
      "step": 42165
    },
    {
      "entropy": 5.306607866287232,
      "epoch": 4.7361150109507495,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002978133800922473,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22872566282749177,
      "num_tokens": 94152520.0,
      "step": 42170
    },
    {
      "entropy": 5.314824199676513,
      "epoch": 4.736676587858708,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0002977734379178509,
      "loss": 4.8809,
      "mean_token_accuracy": 0.23492157012224196,
      "num_tokens": 94164608.0,
      "step": 42175
    },
    {
      "entropy": 5.282477951049804,
      "epoch": 4.737238164766665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002977334950183394,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23734786063432695,
      "num_tokens": 94175343.0,
      "step": 42180
    },
    {
      "entropy": 5.281869173049927,
      "epoch": 4.737799741674622,
      "grad_norm": 0.9375,
      "learning_rate": 0.00029769355139498465,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22460848689079285,
      "num_tokens": 94186832.0,
      "step": 42185
    },
    {
      "entropy": 5.253571319580078,
      "epoch": 4.73836131858258,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002976536070490585,
      "loss": 4.859,
      "mean_token_accuracy": 0.22547847032546997,
      "num_tokens": 94198268.0,
      "step": 42190
    },
    {
      "entropy": 5.315924978256225,
      "epoch": 4.738922895490537,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029761366198183264,
      "loss": 4.8707,
      "mean_token_accuracy": 0.231480173766613,
      "num_tokens": 94209286.0,
      "step": 42195
    },
    {
      "entropy": 5.439875268936158,
      "epoch": 4.739484472398495,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029757371619457904,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2321220964193344,
      "num_tokens": 94220831.0,
      "step": 42200
    },
    {
      "entropy": 5.29284439086914,
      "epoch": 4.740046049306453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029753376968856955,
      "loss": 4.786,
      "mean_token_accuracy": 0.23891403675079345,
      "num_tokens": 94232055.0,
      "step": 42205
    },
    {
      "entropy": 5.234249019622803,
      "epoch": 4.74060762621441,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029749382246507604,
      "loss": 4.8258,
      "mean_token_accuracy": 0.2223077416419983,
      "num_tokens": 94242261.0,
      "step": 42210
    },
    {
      "entropy": 5.355115795135498,
      "epoch": 4.741169203122368,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029745387452537056,
      "loss": 4.9688,
      "mean_token_accuracy": 0.22774934470653535,
      "num_tokens": 94253756.0,
      "step": 42215
    },
    {
      "entropy": 5.2920389652252195,
      "epoch": 4.741730780030325,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029741392587072504,
      "loss": 4.7742,
      "mean_token_accuracy": 0.23443887382745743,
      "num_tokens": 94265713.0,
      "step": 42220
    },
    {
      "entropy": 5.412127590179443,
      "epoch": 4.742292356938282,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00029737397650241135,
      "loss": 5.0415,
      "mean_token_accuracy": 0.21866599023342131,
      "num_tokens": 94276913.0,
      "step": 42225
    },
    {
      "entropy": 5.346949911117553,
      "epoch": 4.7428539338462405,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002973340264217016,
      "loss": 4.9666,
      "mean_token_accuracy": 0.22231330424547197,
      "num_tokens": 94288869.0,
      "step": 42230
    },
    {
      "entropy": 5.335604476928711,
      "epoch": 4.743415510754198,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029729407562986774,
      "loss": 4.9073,
      "mean_token_accuracy": 0.23433059006929396,
      "num_tokens": 94300262.0,
      "step": 42235
    },
    {
      "entropy": 5.440021753311157,
      "epoch": 4.743977087662155,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029725412412818196,
      "loss": 4.9835,
      "mean_token_accuracy": 0.22318292856216432,
      "num_tokens": 94312045.0,
      "step": 42240
    },
    {
      "entropy": 5.34285798072815,
      "epoch": 4.744538664570113,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029721417191791617,
      "loss": 4.8631,
      "mean_token_accuracy": 0.2324235886335373,
      "num_tokens": 94322750.0,
      "step": 42245
    },
    {
      "entropy": 5.257214546203613,
      "epoch": 4.74510024147807,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00029717421900034253,
      "loss": 4.7196,
      "mean_token_accuracy": 0.24010470807552337,
      "num_tokens": 94333945.0,
      "step": 42250
    },
    {
      "entropy": 5.245966100692749,
      "epoch": 4.7456618183860275,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002971342653767332,
      "loss": 4.8193,
      "mean_token_accuracy": 0.22996306419372559,
      "num_tokens": 94344021.0,
      "step": 42255
    },
    {
      "entropy": 5.3238081455230715,
      "epoch": 4.746223395293986,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029709431104836027,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22468641102313996,
      "num_tokens": 94354863.0,
      "step": 42260
    },
    {
      "entropy": 5.308480167388916,
      "epoch": 4.746784972201943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029705435601649584,
      "loss": 4.8403,
      "mean_token_accuracy": 0.23219407945871354,
      "num_tokens": 94365729.0,
      "step": 42265
    },
    {
      "entropy": 5.4176257133483885,
      "epoch": 4.747346549109901,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002970144002824122,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22682932764291763,
      "num_tokens": 94377467.0,
      "step": 42270
    },
    {
      "entropy": 5.352931261062622,
      "epoch": 4.747908126017858,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002969744438473815,
      "loss": 4.9311,
      "mean_token_accuracy": 0.23567714542150497,
      "num_tokens": 94389154.0,
      "step": 42275
    },
    {
      "entropy": 5.40204381942749,
      "epoch": 4.748469702925815,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002969344867126761,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22720986008644103,
      "num_tokens": 94401292.0,
      "step": 42280
    },
    {
      "entropy": 5.351775121688843,
      "epoch": 4.7490312798337735,
      "grad_norm": 1.125,
      "learning_rate": 0.00029689452887956804,
      "loss": 4.9147,
      "mean_token_accuracy": 0.2251904234290123,
      "num_tokens": 94412707.0,
      "step": 42285
    },
    {
      "entropy": 5.337867879867554,
      "epoch": 4.749592856741731,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029685457034932966,
      "loss": 4.8798,
      "mean_token_accuracy": 0.2245976909995079,
      "num_tokens": 94423887.0,
      "step": 42290
    },
    {
      "entropy": 5.377493143081665,
      "epoch": 4.750154433649689,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002968146111232333,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2251439571380615,
      "num_tokens": 94435046.0,
      "step": 42295
    },
    {
      "entropy": 5.336307144165039,
      "epoch": 4.750716010557646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029677465120255133,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23388715386390685,
      "num_tokens": 94446160.0,
      "step": 42300
    },
    {
      "entropy": 5.3447037696838375,
      "epoch": 4.751277587465603,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002967346905885559,
      "loss": 4.8934,
      "mean_token_accuracy": 0.23582927733659745,
      "num_tokens": 94456101.0,
      "step": 42305
    },
    {
      "entropy": 5.294980859756469,
      "epoch": 4.7518391643735605,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029669472928251954,
      "loss": 4.8395,
      "mean_token_accuracy": 0.23958063423633574,
      "num_tokens": 94466820.0,
      "step": 42310
    },
    {
      "entropy": 5.319519090652466,
      "epoch": 4.752400741281519,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029665476728571466,
      "loss": 4.8592,
      "mean_token_accuracy": 0.23113391548395157,
      "num_tokens": 94478236.0,
      "step": 42315
    },
    {
      "entropy": 5.3873358249664305,
      "epoch": 4.752962318189476,
      "grad_norm": 1.0,
      "learning_rate": 0.0002966148045994135,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22463198155164718,
      "num_tokens": 94490136.0,
      "step": 42320
    },
    {
      "entropy": 5.368136739730835,
      "epoch": 4.753523895097434,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002965748412248886,
      "loss": 4.9257,
      "mean_token_accuracy": 0.22944828867912292,
      "num_tokens": 94501694.0,
      "step": 42325
    },
    {
      "entropy": 5.244281339645386,
      "epoch": 4.754085472005391,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002965348771634124,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22721097469329835,
      "num_tokens": 94512724.0,
      "step": 42330
    },
    {
      "entropy": 5.414455699920654,
      "epoch": 4.754647048913348,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00029649491241625736,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22722697108983994,
      "num_tokens": 94523055.0,
      "step": 42335
    },
    {
      "entropy": 5.2879150390625,
      "epoch": 4.755208625821306,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029645494698469604,
      "loss": 4.8224,
      "mean_token_accuracy": 0.23019666820764542,
      "num_tokens": 94533656.0,
      "step": 42340
    },
    {
      "entropy": 5.352685976028442,
      "epoch": 4.755770202729264,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00029641498087000083,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22331905215978623,
      "num_tokens": 94544772.0,
      "step": 42345
    },
    {
      "entropy": 5.28442645072937,
      "epoch": 4.756331779637222,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002963750140734443,
      "loss": 4.8168,
      "mean_token_accuracy": 0.2409016013145447,
      "num_tokens": 94554386.0,
      "step": 42350
    },
    {
      "entropy": 5.274297046661377,
      "epoch": 4.756893356545179,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002963350465962991,
      "loss": 4.9294,
      "mean_token_accuracy": 0.2230806455016136,
      "num_tokens": 94566314.0,
      "step": 42355
    },
    {
      "entropy": 5.397469997406006,
      "epoch": 4.757454933453136,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002962950784398376,
      "loss": 4.9305,
      "mean_token_accuracy": 0.23076384216547013,
      "num_tokens": 94577420.0,
      "step": 42360
    },
    {
      "entropy": 5.33483510017395,
      "epoch": 4.758016510361094,
      "grad_norm": 1.0,
      "learning_rate": 0.0002962551096053327,
      "loss": 4.8532,
      "mean_token_accuracy": 0.22766718417406082,
      "num_tokens": 94588956.0,
      "step": 42365
    },
    {
      "entropy": 5.416394567489624,
      "epoch": 4.7585780872690515,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029621514009405675,
      "loss": 5.0064,
      "mean_token_accuracy": 0.22231080383062363,
      "num_tokens": 94601446.0,
      "step": 42370
    },
    {
      "entropy": 5.2310144901275635,
      "epoch": 4.759139664177009,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002961751699072826,
      "loss": 4.7867,
      "mean_token_accuracy": 0.24080368429422377,
      "num_tokens": 94612489.0,
      "step": 42375
    },
    {
      "entropy": 5.352566146850586,
      "epoch": 4.759701241084967,
      "grad_norm": 1.125,
      "learning_rate": 0.0002961351990462827,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2280285358428955,
      "num_tokens": 94622576.0,
      "step": 42380
    },
    {
      "entropy": 5.275673246383667,
      "epoch": 4.760262817992924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029609522751232996,
      "loss": 4.7784,
      "mean_token_accuracy": 0.23069059401750563,
      "num_tokens": 94633649.0,
      "step": 42385
    },
    {
      "entropy": 5.3353746891021725,
      "epoch": 4.760824394900881,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029605525530669705,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2224062204360962,
      "num_tokens": 94645229.0,
      "step": 42390
    },
    {
      "entropy": 5.294969463348389,
      "epoch": 4.761385971808839,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029601528243065656,
      "loss": 4.8588,
      "mean_token_accuracy": 0.23314192444086074,
      "num_tokens": 94656374.0,
      "step": 42395
    },
    {
      "entropy": 5.328191947937012,
      "epoch": 4.761947548716797,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029597530888548135,
      "loss": 4.8819,
      "mean_token_accuracy": 0.23007851094007492,
      "num_tokens": 94667634.0,
      "step": 42400
    },
    {
      "entropy": 5.397167110443116,
      "epoch": 4.762509125624755,
      "grad_norm": 1.0,
      "learning_rate": 0.00029593533467244415,
      "loss": 4.9795,
      "mean_token_accuracy": 0.2219239890575409,
      "num_tokens": 94679418.0,
      "step": 42405
    },
    {
      "entropy": 5.428636789321899,
      "epoch": 4.763070702532712,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002958953597928178,
      "loss": 5.0003,
      "mean_token_accuracy": 0.227021224796772,
      "num_tokens": 94690872.0,
      "step": 42410
    },
    {
      "entropy": 5.374027109146118,
      "epoch": 4.763632279440669,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002958553842478751,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2255226954817772,
      "num_tokens": 94701630.0,
      "step": 42415
    },
    {
      "entropy": 5.315361499786377,
      "epoch": 4.764193856348627,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029581540803888886,
      "loss": 4.8413,
      "mean_token_accuracy": 0.23393039405345917,
      "num_tokens": 94713192.0,
      "step": 42420
    },
    {
      "entropy": 5.190341949462891,
      "epoch": 4.7647554332565845,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029577543116713197,
      "loss": 4.7336,
      "mean_token_accuracy": 0.2386300668120384,
      "num_tokens": 94723681.0,
      "step": 42425
    },
    {
      "entropy": 5.358533477783203,
      "epoch": 4.765317010164542,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029573545363387725,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22395992279052734,
      "num_tokens": 94735575.0,
      "step": 42430
    },
    {
      "entropy": 5.418138933181763,
      "epoch": 4.7658785870725,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002956954754403977,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22810962945222854,
      "num_tokens": 94745913.0,
      "step": 42435
    },
    {
      "entropy": 5.4595075130462645,
      "epoch": 4.766440163980457,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002956554965879662,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2266670897603035,
      "num_tokens": 94757429.0,
      "step": 42440
    },
    {
      "entropy": 5.281836366653442,
      "epoch": 4.767001740888414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002956155170778556,
      "loss": 4.7427,
      "mean_token_accuracy": 0.2357871413230896,
      "num_tokens": 94767552.0,
      "step": 42445
    },
    {
      "entropy": 5.38918719291687,
      "epoch": 4.767563317796372,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002955755369113389,
      "loss": 4.9992,
      "mean_token_accuracy": 0.22958276569843292,
      "num_tokens": 94780016.0,
      "step": 42450
    },
    {
      "entropy": 5.167181062698364,
      "epoch": 4.7681248947043295,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029553555608968925,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2442493438720703,
      "num_tokens": 94791529.0,
      "step": 42455
    },
    {
      "entropy": 5.267548847198486,
      "epoch": 4.768686471612288,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002954955746141795,
      "loss": 4.8843,
      "mean_token_accuracy": 0.22789565473794937,
      "num_tokens": 94801580.0,
      "step": 42460
    },
    {
      "entropy": 5.347129583358765,
      "epoch": 4.769248048520245,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002954555924860827,
      "loss": 4.8571,
      "mean_token_accuracy": 0.23230217397212982,
      "num_tokens": 94812143.0,
      "step": 42465
    },
    {
      "entropy": 5.290542650222778,
      "epoch": 4.769809625428202,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002954156097066718,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22464328557252883,
      "num_tokens": 94822541.0,
      "step": 42470
    },
    {
      "entropy": 5.38091025352478,
      "epoch": 4.77037120233616,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002953756262772201,
      "loss": 4.9411,
      "mean_token_accuracy": 0.23001226931810378,
      "num_tokens": 94834298.0,
      "step": 42475
    },
    {
      "entropy": 5.408822298049927,
      "epoch": 4.770932779244117,
      "grad_norm": 1.0,
      "learning_rate": 0.0002953356421990005,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22896819561719894,
      "num_tokens": 94844753.0,
      "step": 42480
    },
    {
      "entropy": 5.3620133876800535,
      "epoch": 4.7714943561520755,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002952956574732862,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22314756959676743,
      "num_tokens": 94856082.0,
      "step": 42485
    },
    {
      "entropy": 5.295596885681152,
      "epoch": 4.772055933060033,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002952556721013503,
      "loss": 4.8156,
      "mean_token_accuracy": 0.23112361878156662,
      "num_tokens": 94868342.0,
      "step": 42490
    },
    {
      "entropy": 5.392345285415649,
      "epoch": 4.77261750996799,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002952156860844659,
      "loss": 4.9288,
      "mean_token_accuracy": 0.227860327064991,
      "num_tokens": 94879799.0,
      "step": 42495
    },
    {
      "entropy": 5.407217025756836,
      "epoch": 4.773179086875947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002951756994239063,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22482404559850694,
      "num_tokens": 94891309.0,
      "step": 42500
    },
    {
      "entropy": 5.272177886962891,
      "epoch": 4.773740663783905,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00029513571212094443,
      "loss": 4.8656,
      "mean_token_accuracy": 0.23112059086561204,
      "num_tokens": 94902755.0,
      "step": 42505
    },
    {
      "entropy": 5.288668823242188,
      "epoch": 4.7743022406918625,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002950957241768539,
      "loss": 4.7764,
      "mean_token_accuracy": 0.24598341733217238,
      "num_tokens": 94912981.0,
      "step": 42510
    },
    {
      "entropy": 5.234358787536621,
      "epoch": 4.774863817599821,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00029505573559290756,
      "loss": 4.8498,
      "mean_token_accuracy": 0.23153398483991622,
      "num_tokens": 94924843.0,
      "step": 42515
    },
    {
      "entropy": 5.376433086395264,
      "epoch": 4.775425394507778,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029501574637037884,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2277486056089401,
      "num_tokens": 94935575.0,
      "step": 42520
    },
    {
      "entropy": 5.268910264968872,
      "epoch": 4.775986971415735,
      "grad_norm": 1.015625,
      "learning_rate": 0.000294975756510541,
      "loss": 4.781,
      "mean_token_accuracy": 0.24068457931280135,
      "num_tokens": 94946459.0,
      "step": 42525
    },
    {
      "entropy": 5.383422994613648,
      "epoch": 4.776548548323693,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002949357660146673,
      "loss": 4.897,
      "mean_token_accuracy": 0.22915136963129043,
      "num_tokens": 94957454.0,
      "step": 42530
    },
    {
      "entropy": 5.301697492599487,
      "epoch": 4.77711012523165,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002948957748840311,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22467681616544724,
      "num_tokens": 94968394.0,
      "step": 42535
    },
    {
      "entropy": 5.285888719558716,
      "epoch": 4.7776717021396085,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029485578311990574,
      "loss": 4.8565,
      "mean_token_accuracy": 0.23844217956066133,
      "num_tokens": 94980048.0,
      "step": 42540
    },
    {
      "entropy": 5.26178388595581,
      "epoch": 4.778233279047566,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029481579072356456,
      "loss": 4.8214,
      "mean_token_accuracy": 0.23710710257291795,
      "num_tokens": 94991070.0,
      "step": 42545
    },
    {
      "entropy": 5.28102068901062,
      "epoch": 4.778794855955523,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002947757976962809,
      "loss": 4.8057,
      "mean_token_accuracy": 0.23239758610725403,
      "num_tokens": 95001881.0,
      "step": 42550
    },
    {
      "entropy": 5.363073301315308,
      "epoch": 4.779356432863481,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029473580403932814,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22532707303762436,
      "num_tokens": 95011871.0,
      "step": 42555
    },
    {
      "entropy": 5.293043565750122,
      "epoch": 4.779918009771438,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002946958097539798,
      "loss": 4.8164,
      "mean_token_accuracy": 0.22973926067352296,
      "num_tokens": 95022744.0,
      "step": 42560
    },
    {
      "entropy": 5.26733078956604,
      "epoch": 4.780479586679395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002946558148415092,
      "loss": 4.8292,
      "mean_token_accuracy": 0.23056551814079285,
      "num_tokens": 95032905.0,
      "step": 42565
    },
    {
      "entropy": 5.375695514678955,
      "epoch": 4.7810411635873535,
      "grad_norm": 1.0,
      "learning_rate": 0.0002946158193031898,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22271549701690674,
      "num_tokens": 95044051.0,
      "step": 42570
    },
    {
      "entropy": 5.355753803253174,
      "epoch": 4.781602740495311,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029457582314029507,
      "loss": 4.864,
      "mean_token_accuracy": 0.23478852957487106,
      "num_tokens": 95054482.0,
      "step": 42575
    },
    {
      "entropy": 5.382435750961304,
      "epoch": 4.782164317403268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029453582635409866,
      "loss": 4.9026,
      "mean_token_accuracy": 0.22424572408199311,
      "num_tokens": 95064319.0,
      "step": 42580
    },
    {
      "entropy": 5.2300678253173825,
      "epoch": 4.782725894311226,
      "grad_norm": 1.0,
      "learning_rate": 0.0002944958289458739,
      "loss": 4.8004,
      "mean_token_accuracy": 0.22758205682039262,
      "num_tokens": 95075129.0,
      "step": 42585
    },
    {
      "entropy": 5.365240383148193,
      "epoch": 4.783287471219183,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029445583091689437,
      "loss": 4.9031,
      "mean_token_accuracy": 0.2243991121649742,
      "num_tokens": 95086383.0,
      "step": 42590
    },
    {
      "entropy": 5.373481941223145,
      "epoch": 4.783849048127141,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002944158322684336,
      "loss": 4.817,
      "mean_token_accuracy": 0.23565888851881028,
      "num_tokens": 95098804.0,
      "step": 42595
    },
    {
      "entropy": 5.291520786285401,
      "epoch": 4.784410625035099,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029437583300176526,
      "loss": 4.844,
      "mean_token_accuracy": 0.23695716708898545,
      "num_tokens": 95110465.0,
      "step": 42600
    },
    {
      "entropy": 5.237540578842163,
      "epoch": 4.784972201943056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00029433583311816284,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2278318554162979,
      "num_tokens": 95121999.0,
      "step": 42605
    },
    {
      "entropy": 5.304619216918946,
      "epoch": 4.785533778851014,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002942958326189,
      "loss": 4.8195,
      "mean_token_accuracy": 0.24257900416851044,
      "num_tokens": 95132920.0,
      "step": 42610
    },
    {
      "entropy": 5.353788280487061,
      "epoch": 4.786095355758971,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029425583150525044,
      "loss": 4.9196,
      "mean_token_accuracy": 0.23226158618927,
      "num_tokens": 95143361.0,
      "step": 42615
    },
    {
      "entropy": 5.359145450592041,
      "epoch": 4.786656932666928,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029421582977848763,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22206931114196776,
      "num_tokens": 95153509.0,
      "step": 42620
    },
    {
      "entropy": 5.312826204299927,
      "epoch": 4.7872185095748865,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00029417582743988547,
      "loss": 4.8599,
      "mean_token_accuracy": 0.23063335418701172,
      "num_tokens": 95165762.0,
      "step": 42625
    },
    {
      "entropy": 5.284657716751099,
      "epoch": 4.787780086482844,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00029413582449071745,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22656734734773637,
      "num_tokens": 95177038.0,
      "step": 42630
    },
    {
      "entropy": 5.422993803024292,
      "epoch": 4.788341663390801,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002940958209322574,
      "loss": 4.941,
      "mean_token_accuracy": 0.2293344557285309,
      "num_tokens": 95188641.0,
      "step": 42635
    },
    {
      "entropy": 5.41595048904419,
      "epoch": 4.788903240298759,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00029405581676577897,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21891319006681442,
      "num_tokens": 95200792.0,
      "step": 42640
    },
    {
      "entropy": 5.452337121963501,
      "epoch": 4.789464817206716,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000294015811992556,
      "loss": 4.9848,
      "mean_token_accuracy": 0.22601761519908906,
      "num_tokens": 95212578.0,
      "step": 42645
    },
    {
      "entropy": 5.304792785644532,
      "epoch": 4.790026394114674,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002939758066138621,
      "loss": 4.8474,
      "mean_token_accuracy": 0.23232062458992003,
      "num_tokens": 95223744.0,
      "step": 42650
    },
    {
      "entropy": 5.2585015296936035,
      "epoch": 4.790587971022632,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002939358006309712,
      "loss": 4.8154,
      "mean_token_accuracy": 0.2406544417142868,
      "num_tokens": 95234836.0,
      "step": 42655
    },
    {
      "entropy": 5.352051210403443,
      "epoch": 4.791149547930589,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002938957940451571,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23248167484998702,
      "num_tokens": 95246293.0,
      "step": 42660
    },
    {
      "entropy": 5.439797115325928,
      "epoch": 4.791711124838547,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00029385578685769346,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22030054926872253,
      "num_tokens": 95257240.0,
      "step": 42665
    },
    {
      "entropy": 5.367308807373047,
      "epoch": 4.792272701746504,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002938157790698544,
      "loss": 4.8723,
      "mean_token_accuracy": 0.23114923536777496,
      "num_tokens": 95267270.0,
      "step": 42670
    },
    {
      "entropy": 5.29309310913086,
      "epoch": 4.792834278654462,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002937757706829136,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22460008710622786,
      "num_tokens": 95278451.0,
      "step": 42675
    },
    {
      "entropy": 5.284264183044433,
      "epoch": 4.793395855562419,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000293735761698145,
      "loss": 4.9194,
      "mean_token_accuracy": 0.23590585887432097,
      "num_tokens": 95290367.0,
      "step": 42680
    },
    {
      "entropy": 5.3096435546875,
      "epoch": 4.793957432470377,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002936957521168224,
      "loss": 4.7692,
      "mean_token_accuracy": 0.23875954747200012,
      "num_tokens": 95300875.0,
      "step": 42685
    },
    {
      "entropy": 5.2712689399719235,
      "epoch": 4.794519009378335,
      "grad_norm": 1.0,
      "learning_rate": 0.0002936557419402198,
      "loss": 4.7997,
      "mean_token_accuracy": 0.2310612455010414,
      "num_tokens": 95311299.0,
      "step": 42690
    },
    {
      "entropy": 5.329549932479859,
      "epoch": 4.795080586286292,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002936157311696112,
      "loss": 4.8938,
      "mean_token_accuracy": 0.232401442527771,
      "num_tokens": 95322298.0,
      "step": 42695
    },
    {
      "entropy": 5.366584825515747,
      "epoch": 4.795642163194249,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002935757198062705,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21869388073682786,
      "num_tokens": 95333190.0,
      "step": 42700
    },
    {
      "entropy": 5.341529178619385,
      "epoch": 4.796203740102207,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002935357078514717,
      "loss": 4.8582,
      "mean_token_accuracy": 0.23462610691785812,
      "num_tokens": 95344008.0,
      "step": 42705
    },
    {
      "entropy": 5.312467002868653,
      "epoch": 4.7967653170101645,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029349569530648877,
      "loss": 4.8256,
      "mean_token_accuracy": 0.23169538825750352,
      "num_tokens": 95354746.0,
      "step": 42710
    },
    {
      "entropy": 5.180278205871582,
      "epoch": 4.797326893918122,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002934556821725957,
      "loss": 4.744,
      "mean_token_accuracy": 0.2360079139471054,
      "num_tokens": 95365963.0,
      "step": 42715
    },
    {
      "entropy": 5.314266920089722,
      "epoch": 4.79788847082608,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00029341566845106647,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22951397448778152,
      "num_tokens": 95376807.0,
      "step": 42720
    },
    {
      "entropy": 5.350525283813477,
      "epoch": 4.798450047734037,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002933756541431753,
      "loss": 4.9266,
      "mean_token_accuracy": 0.22568430751562119,
      "num_tokens": 95387467.0,
      "step": 42725
    },
    {
      "entropy": 5.343288135528565,
      "epoch": 4.799011624641995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029333563925019616,
      "loss": 4.8647,
      "mean_token_accuracy": 0.23177625238895416,
      "num_tokens": 95398275.0,
      "step": 42730
    },
    {
      "entropy": 5.299396657943726,
      "epoch": 4.799573201549952,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029329562377340327,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22968343496322632,
      "num_tokens": 95410231.0,
      "step": 42735
    },
    {
      "entropy": 5.251077318191529,
      "epoch": 4.80013477845791,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002932556077140704,
      "loss": 4.7813,
      "mean_token_accuracy": 0.2487647458910942,
      "num_tokens": 95421008.0,
      "step": 42740
    },
    {
      "entropy": 5.299357843399048,
      "epoch": 4.800696355365868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029321559107347206,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22979382276535035,
      "num_tokens": 95430972.0,
      "step": 42745
    },
    {
      "entropy": 5.359325122833252,
      "epoch": 4.801257932273825,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00029317557385288216,
      "loss": 4.887,
      "mean_token_accuracy": 0.2319648891687393,
      "num_tokens": 95442486.0,
      "step": 42750
    },
    {
      "entropy": 5.307155895233154,
      "epoch": 4.801819509181782,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000293135556053575,
      "loss": 4.8676,
      "mean_token_accuracy": 0.23343969732522965,
      "num_tokens": 95453994.0,
      "step": 42755
    },
    {
      "entropy": 5.315494632720947,
      "epoch": 4.80238108608974,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029309553767682467,
      "loss": 4.8318,
      "mean_token_accuracy": 0.23557461351156234,
      "num_tokens": 95463541.0,
      "step": 42760
    },
    {
      "entropy": 5.335611009597779,
      "epoch": 4.8029426629976975,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002930555187239054,
      "loss": 4.9404,
      "mean_token_accuracy": 0.22833583503961563,
      "num_tokens": 95474919.0,
      "step": 42765
    },
    {
      "entropy": 5.294150733947754,
      "epoch": 4.803504239905655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029301549919609146,
      "loss": 4.7749,
      "mean_token_accuracy": 0.24282611310482025,
      "num_tokens": 95485110.0,
      "step": 42770
    },
    {
      "entropy": 5.2900879859924315,
      "epoch": 4.804065816813613,
      "grad_norm": 1.0,
      "learning_rate": 0.00029297547909465697,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22216370701789856,
      "num_tokens": 95497210.0,
      "step": 42775
    },
    {
      "entropy": 5.415588665008545,
      "epoch": 4.80462739372157,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002929354584208763,
      "loss": 5.0285,
      "mean_token_accuracy": 0.22394767999649048,
      "num_tokens": 95510110.0,
      "step": 42780
    },
    {
      "entropy": 5.318846368789673,
      "epoch": 4.805188970629528,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002928954371760237,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2279029294848442,
      "num_tokens": 95520841.0,
      "step": 42785
    },
    {
      "entropy": 5.325888538360596,
      "epoch": 4.805750547537485,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002928554153613734,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23398559540510178,
      "num_tokens": 95531860.0,
      "step": 42790
    },
    {
      "entropy": 5.345814275741577,
      "epoch": 4.8063121244454425,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002928153929781997,
      "loss": 4.8738,
      "mean_token_accuracy": 0.22905177175998687,
      "num_tokens": 95543533.0,
      "step": 42795
    },
    {
      "entropy": 5.341566181182861,
      "epoch": 4.806873701353401,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029277537002777703,
      "loss": 4.8085,
      "mean_token_accuracy": 0.235746431350708,
      "num_tokens": 95554890.0,
      "step": 42800
    },
    {
      "entropy": 5.337465763092041,
      "epoch": 4.807435278261358,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029273534651137967,
      "loss": 4.8698,
      "mean_token_accuracy": 0.23406989872455597,
      "num_tokens": 95567849.0,
      "step": 42805
    },
    {
      "entropy": 5.286644458770752,
      "epoch": 4.807996855169315,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029269532243028195,
      "loss": 4.7586,
      "mean_token_accuracy": 0.24070949852466583,
      "num_tokens": 95580047.0,
      "step": 42810
    },
    {
      "entropy": 5.321757650375366,
      "epoch": 4.808558432077273,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002926552977857584,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22867271155118943,
      "num_tokens": 95590690.0,
      "step": 42815
    },
    {
      "entropy": 5.287789392471313,
      "epoch": 4.80912000898523,
      "grad_norm": 1.0,
      "learning_rate": 0.00029261527257908325,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22744768857955933,
      "num_tokens": 95601439.0,
      "step": 42820
    },
    {
      "entropy": 5.33438925743103,
      "epoch": 4.809681585893188,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029257524681153106,
      "loss": 4.8532,
      "mean_token_accuracy": 0.23291106671094894,
      "num_tokens": 95611970.0,
      "step": 42825
    },
    {
      "entropy": 5.356077003479004,
      "epoch": 4.810243162801146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002925352204843761,
      "loss": 4.8413,
      "mean_token_accuracy": 0.23313669264316558,
      "num_tokens": 95622695.0,
      "step": 42830
    },
    {
      "entropy": 5.315057277679443,
      "epoch": 4.810804739709103,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000292495193598893,
      "loss": 4.8406,
      "mean_token_accuracy": 0.2332966849207878,
      "num_tokens": 95633774.0,
      "step": 42835
    },
    {
      "entropy": 5.246612119674682,
      "epoch": 4.811366316617061,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029245516615635615,
      "loss": 4.8091,
      "mean_token_accuracy": 0.23598286062479018,
      "num_tokens": 95645287.0,
      "step": 42840
    },
    {
      "entropy": 5.364112281799317,
      "epoch": 4.811927893525018,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029241513815804,
      "loss": 4.8701,
      "mean_token_accuracy": 0.2278810277581215,
      "num_tokens": 95656073.0,
      "step": 42845
    },
    {
      "entropy": 5.361810922622681,
      "epoch": 4.8124894704329755,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029237510960521904,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22782557904720308,
      "num_tokens": 95668049.0,
      "step": 42850
    },
    {
      "entropy": 5.233780431747436,
      "epoch": 4.813051047340934,
      "grad_norm": 1.0625,
      "learning_rate": 0.000292335080499168,
      "loss": 4.7445,
      "mean_token_accuracy": 0.24447658956050872,
      "num_tokens": 95679539.0,
      "step": 42855
    },
    {
      "entropy": 5.2559043884277346,
      "epoch": 4.813612624248891,
      "grad_norm": 1.0,
      "learning_rate": 0.00029229505084116123,
      "loss": 4.8266,
      "mean_token_accuracy": 0.23099856376647948,
      "num_tokens": 95690759.0,
      "step": 42860
    },
    {
      "entropy": 5.3213907241821286,
      "epoch": 4.814174201156849,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002922550206324732,
      "loss": 4.8619,
      "mean_token_accuracy": 0.2291916251182556,
      "num_tokens": 95701590.0,
      "step": 42865
    },
    {
      "entropy": 5.4052464962005615,
      "epoch": 4.814735778064806,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002922149898743787,
      "loss": 4.9198,
      "mean_token_accuracy": 0.23098693490028382,
      "num_tokens": 95711934.0,
      "step": 42870
    },
    {
      "entropy": 5.357414197921753,
      "epoch": 4.815297354972763,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002921749585681523,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22864381223917007,
      "num_tokens": 95722569.0,
      "step": 42875
    },
    {
      "entropy": 5.3060918807983395,
      "epoch": 4.8158589318807214,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029213492671506855,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2228565275669098,
      "num_tokens": 95733043.0,
      "step": 42880
    },
    {
      "entropy": 5.291569089889526,
      "epoch": 4.816420508788679,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00029209489431640203,
      "loss": 4.8017,
      "mean_token_accuracy": 0.23368646800518036,
      "num_tokens": 95743262.0,
      "step": 42885
    },
    {
      "entropy": 5.375045824050903,
      "epoch": 4.816982085696636,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0002920548613734275,
      "loss": 4.9345,
      "mean_token_accuracy": 0.23007966578006744,
      "num_tokens": 95754708.0,
      "step": 42890
    },
    {
      "entropy": 5.3585028648376465,
      "epoch": 4.817543662604594,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029201482788741953,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22821191251277922,
      "num_tokens": 95765255.0,
      "step": 42895
    },
    {
      "entropy": 5.399783515930176,
      "epoch": 4.818105239512551,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002919747938596528,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2254250705242157,
      "num_tokens": 95775367.0,
      "step": 42900
    },
    {
      "entropy": 5.3890105247497555,
      "epoch": 4.818666816420508,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002919347592914021,
      "loss": 4.927,
      "mean_token_accuracy": 0.23353558778762817,
      "num_tokens": 95785952.0,
      "step": 42905
    },
    {
      "entropy": 5.414256238937378,
      "epoch": 4.8192283933284665,
      "grad_norm": 1.125,
      "learning_rate": 0.00029189472418394215,
      "loss": 5.0412,
      "mean_token_accuracy": 0.22387515008449554,
      "num_tokens": 95796407.0,
      "step": 42910
    },
    {
      "entropy": 5.319532585144043,
      "epoch": 4.819789970236424,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029185468853854755,
      "loss": 4.8497,
      "mean_token_accuracy": 0.23301519304513932,
      "num_tokens": 95806698.0,
      "step": 42915
    },
    {
      "entropy": 5.2897931098937985,
      "epoch": 4.820351547144382,
      "grad_norm": 1.0,
      "learning_rate": 0.0002918146523564932,
      "loss": 4.8162,
      "mean_token_accuracy": 0.23302064538002015,
      "num_tokens": 95817975.0,
      "step": 42920
    },
    {
      "entropy": 5.342338037490845,
      "epoch": 4.820913124052339,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029177461563905375,
      "loss": 4.8661,
      "mean_token_accuracy": 0.2357890248298645,
      "num_tokens": 95828831.0,
      "step": 42925
    },
    {
      "entropy": 5.348430967330932,
      "epoch": 4.821474700960296,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029173457838750405,
      "loss": 4.8822,
      "mean_token_accuracy": 0.230085888504982,
      "num_tokens": 95840375.0,
      "step": 42930
    },
    {
      "entropy": 5.281631851196289,
      "epoch": 4.822036277868254,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029169454060311886,
      "loss": 4.8949,
      "mean_token_accuracy": 0.23167502284049987,
      "num_tokens": 95852629.0,
      "step": 42935
    },
    {
      "entropy": 5.345697021484375,
      "epoch": 4.822597854776212,
      "grad_norm": 1.046875,
      "learning_rate": 0.000291654502287173,
      "loss": 4.9229,
      "mean_token_accuracy": 0.23165742456912994,
      "num_tokens": 95862844.0,
      "step": 42940
    },
    {
      "entropy": 5.342270994186402,
      "epoch": 4.823159431684169,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002916144634409414,
      "loss": 4.8389,
      "mean_token_accuracy": 0.23228463977575303,
      "num_tokens": 95873165.0,
      "step": 42945
    },
    {
      "entropy": 5.254854536056518,
      "epoch": 4.823721008592127,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029157442406569885,
      "loss": 4.8418,
      "mean_token_accuracy": 0.22796654403209687,
      "num_tokens": 95884349.0,
      "step": 42950
    },
    {
      "entropy": 5.2913782596588135,
      "epoch": 4.824282585500084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002915343841627201,
      "loss": 4.8444,
      "mean_token_accuracy": 0.2326574519276619,
      "num_tokens": 95894614.0,
      "step": 42955
    },
    {
      "entropy": 5.324865531921387,
      "epoch": 4.824844162408041,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002914943437332802,
      "loss": 4.886,
      "mean_token_accuracy": 0.23467847257852553,
      "num_tokens": 95905644.0,
      "step": 42960
    },
    {
      "entropy": 5.268607950210571,
      "epoch": 4.8254057393159995,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029145430277865406,
      "loss": 4.8385,
      "mean_token_accuracy": 0.2309998616576195,
      "num_tokens": 95915318.0,
      "step": 42965
    },
    {
      "entropy": 5.272212266921997,
      "epoch": 4.825967316223957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002914142613001165,
      "loss": 4.8087,
      "mean_token_accuracy": 0.2320857286453247,
      "num_tokens": 95926144.0,
      "step": 42970
    },
    {
      "entropy": 5.259996223449707,
      "epoch": 4.826528893131915,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029137421929894235,
      "loss": 4.8256,
      "mean_token_accuracy": 0.2258915677666664,
      "num_tokens": 95936384.0,
      "step": 42975
    },
    {
      "entropy": 5.353667974472046,
      "epoch": 4.827090470039872,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002913341767764069,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2254423752427101,
      "num_tokens": 95946597.0,
      "step": 42980
    },
    {
      "entropy": 5.323237800598145,
      "epoch": 4.827652046947829,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029129413373378484,
      "loss": 4.8917,
      "mean_token_accuracy": 0.23017434328794478,
      "num_tokens": 95958054.0,
      "step": 42985
    },
    {
      "entropy": 5.320249128341675,
      "epoch": 4.828213623855787,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00029125409017235126,
      "loss": 4.907,
      "mean_token_accuracy": 0.23131458908319474,
      "num_tokens": 95969167.0,
      "step": 42990
    },
    {
      "entropy": 5.35586109161377,
      "epoch": 4.828775200763745,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002912140460933812,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2334311917424202,
      "num_tokens": 95980685.0,
      "step": 42995
    },
    {
      "entropy": 5.350427770614624,
      "epoch": 4.829336777671702,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029117400149814953,
      "loss": 4.966,
      "mean_token_accuracy": 0.22223259508609772,
      "num_tokens": 95991588.0,
      "step": 43000
    },
    {
      "entropy": 5.368867063522339,
      "epoch": 4.82989835457966,
      "grad_norm": 1.125,
      "learning_rate": 0.0002911339563879315,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22808744609355927,
      "num_tokens": 96003035.0,
      "step": 43005
    },
    {
      "entropy": 5.379842138290405,
      "epoch": 4.830459931487617,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002910939107640019,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22827474623918534,
      "num_tokens": 96013365.0,
      "step": 43010
    },
    {
      "entropy": 5.341946125030518,
      "epoch": 4.831021508395574,
      "grad_norm": 1.078125,
      "learning_rate": 0.000291053864627636,
      "loss": 4.8171,
      "mean_token_accuracy": 0.23325565606355667,
      "num_tokens": 96024140.0,
      "step": 43015
    },
    {
      "entropy": 5.294568872451782,
      "epoch": 4.831583085303532,
      "grad_norm": 1.0,
      "learning_rate": 0.00029101381798010885,
      "loss": 4.9,
      "mean_token_accuracy": 0.2232532262802124,
      "num_tokens": 96035647.0,
      "step": 43020
    },
    {
      "entropy": 5.344024658203125,
      "epoch": 4.83214466221149,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002909737708226954,
      "loss": 4.8177,
      "mean_token_accuracy": 0.23930782526731492,
      "num_tokens": 96046190.0,
      "step": 43025
    },
    {
      "entropy": 5.345743560791016,
      "epoch": 4.832706239119448,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029093372315667105,
      "loss": 4.8486,
      "mean_token_accuracy": 0.23587086498737336,
      "num_tokens": 96056939.0,
      "step": 43030
    },
    {
      "entropy": 5.257515478134155,
      "epoch": 4.833267816027405,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029089367498331075,
      "loss": 4.7719,
      "mean_token_accuracy": 0.23428597897291184,
      "num_tokens": 96068111.0,
      "step": 43035
    },
    {
      "entropy": 5.25619330406189,
      "epoch": 4.833829392935362,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029085362630388965,
      "loss": 4.8198,
      "mean_token_accuracy": 0.23603614270687104,
      "num_tokens": 96078303.0,
      "step": 43040
    },
    {
      "entropy": 5.310703706741333,
      "epoch": 4.83439096984332,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002908135771196829,
      "loss": 4.8038,
      "mean_token_accuracy": 0.23762328922748566,
      "num_tokens": 96088492.0,
      "step": 43045
    },
    {
      "entropy": 5.278870582580566,
      "epoch": 4.8349525467512775,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002907735274319657,
      "loss": 4.8412,
      "mean_token_accuracy": 0.23347576409578324,
      "num_tokens": 96100401.0,
      "step": 43050
    },
    {
      "entropy": 5.323126983642578,
      "epoch": 4.835514123659236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002907334772420134,
      "loss": 4.8841,
      "mean_token_accuracy": 0.23284367024898528,
      "num_tokens": 96111848.0,
      "step": 43055
    },
    {
      "entropy": 5.339754486083985,
      "epoch": 4.836075700567193,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000290693426551101,
      "loss": 4.8794,
      "mean_token_accuracy": 0.23273054510354996,
      "num_tokens": 96123544.0,
      "step": 43060
    },
    {
      "entropy": 5.251462173461914,
      "epoch": 4.83663727747515,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00029065337536050384,
      "loss": 4.8536,
      "mean_token_accuracy": 0.2332400679588318,
      "num_tokens": 96134649.0,
      "step": 43065
    },
    {
      "entropy": 5.355578804016114,
      "epoch": 4.837198854383108,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029061332367149705,
      "loss": 4.9362,
      "mean_token_accuracy": 0.22919019609689711,
      "num_tokens": 96147597.0,
      "step": 43070
    },
    {
      "entropy": 5.435472440719605,
      "epoch": 4.837760431291065,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029057327148535614,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2293771982192993,
      "num_tokens": 96157285.0,
      "step": 43075
    },
    {
      "entropy": 5.427551364898681,
      "epoch": 4.838322008199023,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000290533218803356,
      "loss": 5.0473,
      "mean_token_accuracy": 0.21637782752513884,
      "num_tokens": 96169632.0,
      "step": 43080
    },
    {
      "entropy": 5.400629043579102,
      "epoch": 4.838883585106981,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029049316562677235,
      "loss": 4.9307,
      "mean_token_accuracy": 0.23053263127803802,
      "num_tokens": 96181311.0,
      "step": 43085
    },
    {
      "entropy": 5.383113813400269,
      "epoch": 4.839445162014938,
      "grad_norm": 0.90625,
      "learning_rate": 0.00029045311195688016,
      "loss": 4.9219,
      "mean_token_accuracy": 0.23532415181398392,
      "num_tokens": 96193836.0,
      "step": 43090
    },
    {
      "entropy": 5.358198404312134,
      "epoch": 4.840006738922895,
      "grad_norm": 1.046875,
      "learning_rate": 0.000290413057794955,
      "loss": 4.8688,
      "mean_token_accuracy": 0.23525242656469345,
      "num_tokens": 96204530.0,
      "step": 43095
    },
    {
      "entropy": 5.348000001907349,
      "epoch": 4.840568315830853,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029037300314227205,
      "loss": 4.8997,
      "mean_token_accuracy": 0.23065423518419265,
      "num_tokens": 96215402.0,
      "step": 43100
    },
    {
      "entropy": 5.298193645477295,
      "epoch": 4.8411298927388104,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029033294800010666,
      "loss": 4.8187,
      "mean_token_accuracy": 0.23463291972875594,
      "num_tokens": 96225799.0,
      "step": 43105
    },
    {
      "entropy": 5.316161680221557,
      "epoch": 4.8416914696467686,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029029289236973436,
      "loss": 4.9351,
      "mean_token_accuracy": 0.2238069951534271,
      "num_tokens": 96236306.0,
      "step": 43110
    },
    {
      "entropy": 5.337806367874146,
      "epoch": 4.842253046554726,
      "grad_norm": 0.921875,
      "learning_rate": 0.00029025283625243043,
      "loss": 4.8873,
      "mean_token_accuracy": 0.23056380301713944,
      "num_tokens": 96247307.0,
      "step": 43115
    },
    {
      "entropy": 5.393820428848267,
      "epoch": 4.842814623462683,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00029021277964947023,
      "loss": 4.8998,
      "mean_token_accuracy": 0.2313714176416397,
      "num_tokens": 96259883.0,
      "step": 43120
    },
    {
      "entropy": 5.286264228820801,
      "epoch": 4.843376200370641,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029017272256212927,
      "loss": 4.8545,
      "mean_token_accuracy": 0.23162703961133957,
      "num_tokens": 96270219.0,
      "step": 43125
    },
    {
      "entropy": 5.318028831481934,
      "epoch": 4.843937777278598,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002901326649916829,
      "loss": 4.9646,
      "mean_token_accuracy": 0.21764818280935289,
      "num_tokens": 96281318.0,
      "step": 43130
    },
    {
      "entropy": 5.324691390991211,
      "epoch": 4.8444993541865555,
      "grad_norm": 0.9375,
      "learning_rate": 0.00029009260693940665,
      "loss": 4.8742,
      "mean_token_accuracy": 0.23142812848091127,
      "num_tokens": 96292597.0,
      "step": 43135
    },
    {
      "entropy": 5.281784629821777,
      "epoch": 4.845060931094514,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002900525484065759,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23197828829288483,
      "num_tokens": 96302578.0,
      "step": 43140
    },
    {
      "entropy": 5.331191158294677,
      "epoch": 4.845622508002471,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00029001248939446617,
      "loss": 4.8444,
      "mean_token_accuracy": 0.23380081355571747,
      "num_tokens": 96313268.0,
      "step": 43145
    },
    {
      "entropy": 5.265436506271362,
      "epoch": 4.846184084910428,
      "grad_norm": 0.96875,
      "learning_rate": 0.000289972429904353,
      "loss": 4.8735,
      "mean_token_accuracy": 0.23431734293699263,
      "num_tokens": 96323595.0,
      "step": 43150
    },
    {
      "entropy": 5.340033864974975,
      "epoch": 4.846745661818386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028993236993751174,
      "loss": 4.8788,
      "mean_token_accuracy": 0.23143659979104997,
      "num_tokens": 96334698.0,
      "step": 43155
    },
    {
      "entropy": 5.358072710037232,
      "epoch": 4.847307238726343,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028989230949521817,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21996584236621858,
      "num_tokens": 96346107.0,
      "step": 43160
    },
    {
      "entropy": 5.2907391548156735,
      "epoch": 4.8478688156343015,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028985224857874765,
      "loss": 4.7969,
      "mean_token_accuracy": 0.23691593408584594,
      "num_tokens": 96356301.0,
      "step": 43165
    },
    {
      "entropy": 5.2664165019989015,
      "epoch": 4.848430392542259,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028981218718937573,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23671514093875884,
      "num_tokens": 96368048.0,
      "step": 43170
    },
    {
      "entropy": 5.329429721832275,
      "epoch": 4.848991969450216,
      "grad_norm": 1.0,
      "learning_rate": 0.00028977212532837804,
      "loss": 4.8765,
      "mean_token_accuracy": 0.23309760987758638,
      "num_tokens": 96379074.0,
      "step": 43175
    },
    {
      "entropy": 5.276646566390991,
      "epoch": 4.849553546358174,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002897320629970302,
      "loss": 4.8285,
      "mean_token_accuracy": 0.2424048513174057,
      "num_tokens": 96390881.0,
      "step": 43180
    },
    {
      "entropy": 5.345463657379151,
      "epoch": 4.850115123266131,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028969200019660774,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22648611068725585,
      "num_tokens": 96402211.0,
      "step": 43185
    },
    {
      "entropy": 5.318372392654419,
      "epoch": 4.8506767001740885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002896519369283862,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2298620581626892,
      "num_tokens": 96412712.0,
      "step": 43190
    },
    {
      "entropy": 5.3349690437316895,
      "epoch": 4.851238277082047,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002896118731936414,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2331255003809929,
      "num_tokens": 96423516.0,
      "step": 43195
    },
    {
      "entropy": 5.406678628921509,
      "epoch": 4.851799853990004,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002895718089936489,
      "loss": 4.9225,
      "mean_token_accuracy": 0.22649643272161485,
      "num_tokens": 96434632.0,
      "step": 43200
    },
    {
      "entropy": 5.412808656692505,
      "epoch": 4.852361430897961,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002895317443296843,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2262161836028099,
      "num_tokens": 96445644.0,
      "step": 43205
    },
    {
      "entropy": 5.339436292648315,
      "epoch": 4.852923007805919,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028949167920302337,
      "loss": 4.8608,
      "mean_token_accuracy": 0.22936105877161025,
      "num_tokens": 96455921.0,
      "step": 43210
    },
    {
      "entropy": 5.357323265075683,
      "epoch": 4.853484584713876,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028945161361494177,
      "loss": 4.8815,
      "mean_token_accuracy": 0.23438649326562883,
      "num_tokens": 96466359.0,
      "step": 43215
    },
    {
      "entropy": 5.295104122161865,
      "epoch": 4.854046161621834,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002894115475667151,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22728336751461028,
      "num_tokens": 96478314.0,
      "step": 43220
    },
    {
      "entropy": 5.316946935653687,
      "epoch": 4.854607738529792,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028937148105961924,
      "loss": 4.9051,
      "mean_token_accuracy": 0.23063861429691315,
      "num_tokens": 96488849.0,
      "step": 43225
    },
    {
      "entropy": 5.305208969116211,
      "epoch": 4.855169315437749,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0002893314140949298,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22772152423858644,
      "num_tokens": 96500102.0,
      "step": 43230
    },
    {
      "entropy": 5.31298975944519,
      "epoch": 4.855730892345707,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028929134667392265,
      "loss": 4.8463,
      "mean_token_accuracy": 0.23254792541265487,
      "num_tokens": 96511416.0,
      "step": 43235
    },
    {
      "entropy": 5.297690582275391,
      "epoch": 4.856292469253664,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028925127879787343,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2265419602394104,
      "num_tokens": 96522486.0,
      "step": 43240
    },
    {
      "entropy": 5.326198768615723,
      "epoch": 4.856854046161622,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002892112104680579,
      "loss": 4.8849,
      "mean_token_accuracy": 0.23170798867940903,
      "num_tokens": 96533259.0,
      "step": 43245
    },
    {
      "entropy": 5.222328138351441,
      "epoch": 4.8574156230695795,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028917114168575196,
      "loss": 4.8329,
      "mean_token_accuracy": 0.23302036970853807,
      "num_tokens": 96545409.0,
      "step": 43250
    },
    {
      "entropy": 5.2777549743652346,
      "epoch": 4.857977199977537,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002891310724522314,
      "loss": 4.8151,
      "mean_token_accuracy": 0.22367736846208572,
      "num_tokens": 96556656.0,
      "step": 43255
    },
    {
      "entropy": 5.299331903457642,
      "epoch": 4.858538776885495,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00028909100276877203,
      "loss": 4.8218,
      "mean_token_accuracy": 0.23524726927280426,
      "num_tokens": 96567542.0,
      "step": 43260
    },
    {
      "entropy": 5.26268367767334,
      "epoch": 4.859100353793452,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028905093263664966,
      "loss": 4.7531,
      "mean_token_accuracy": 0.2323642283678055,
      "num_tokens": 96579737.0,
      "step": 43265
    },
    {
      "entropy": 5.345893859863281,
      "epoch": 4.859661930701409,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028901086205714016,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22595739960670472,
      "num_tokens": 96589669.0,
      "step": 43270
    },
    {
      "entropy": 5.344051218032837,
      "epoch": 4.860223507609367,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00028897079103151936,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22652770429849625,
      "num_tokens": 96601852.0,
      "step": 43275
    },
    {
      "entropy": 5.391782522201538,
      "epoch": 4.860785084517325,
      "grad_norm": 1.0,
      "learning_rate": 0.00028893071956106315,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2315346747636795,
      "num_tokens": 96613883.0,
      "step": 43280
    },
    {
      "entropy": 5.259414291381836,
      "epoch": 4.861346661425282,
      "grad_norm": 1.109375,
      "learning_rate": 0.00028889064764704735,
      "loss": 4.7821,
      "mean_token_accuracy": 0.2381630226969719,
      "num_tokens": 96625157.0,
      "step": 43285
    },
    {
      "entropy": 5.325397920608521,
      "epoch": 4.86190823833324,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028885057529074804,
      "loss": 4.8898,
      "mean_token_accuracy": 0.2302596315741539,
      "num_tokens": 96636249.0,
      "step": 43290
    },
    {
      "entropy": 5.405759525299072,
      "epoch": 4.862469815241197,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028881050249344103,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21969633400440217,
      "num_tokens": 96647315.0,
      "step": 43295
    },
    {
      "entropy": 5.355242347717285,
      "epoch": 4.863031392149155,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002887704292564022,
      "loss": 4.8798,
      "mean_token_accuracy": 0.2295886605978012,
      "num_tokens": 96658590.0,
      "step": 43300
    },
    {
      "entropy": 5.430629873275757,
      "epoch": 4.8635929690571125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028873035558090766,
      "loss": 4.9423,
      "mean_token_accuracy": 0.22395009100437163,
      "num_tokens": 96668705.0,
      "step": 43305
    },
    {
      "entropy": 5.315474510192871,
      "epoch": 4.86415454596507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002886902814682332,
      "loss": 4.9102,
      "mean_token_accuracy": 0.23082761466503143,
      "num_tokens": 96679588.0,
      "step": 43310
    },
    {
      "entropy": 5.353763771057129,
      "epoch": 4.864716122873028,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00028865020691965485,
      "loss": 4.9419,
      "mean_token_accuracy": 0.23244519680738449,
      "num_tokens": 96691645.0,
      "step": 43315
    },
    {
      "entropy": 5.3849201679229735,
      "epoch": 4.865277699780985,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028861013193644865,
      "loss": 4.9406,
      "mean_token_accuracy": 0.23095204681158066,
      "num_tokens": 96702530.0,
      "step": 43320
    },
    {
      "entropy": 5.259098911285401,
      "epoch": 4.865839276688942,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002885700565198906,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22545968145132064,
      "num_tokens": 96714753.0,
      "step": 43325
    },
    {
      "entropy": 5.315648984909058,
      "epoch": 4.8664008535969,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028852998067125675,
      "loss": 4.8933,
      "mean_token_accuracy": 0.23347335010766984,
      "num_tokens": 96726482.0,
      "step": 43330
    },
    {
      "entropy": 5.349673318862915,
      "epoch": 4.8669624305048576,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002884899043918229,
      "loss": 4.8393,
      "mean_token_accuracy": 0.23442807346582412,
      "num_tokens": 96736926.0,
      "step": 43335
    },
    {
      "entropy": 5.34988317489624,
      "epoch": 4.867524007412815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002884498276828653,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22408774197101594,
      "num_tokens": 96749429.0,
      "step": 43340
    },
    {
      "entropy": 5.355040884017944,
      "epoch": 4.868085584320773,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028840975054566007,
      "loss": 4.9554,
      "mean_token_accuracy": 0.22587487548589708,
      "num_tokens": 96761451.0,
      "step": 43345
    },
    {
      "entropy": 5.3646934032440186,
      "epoch": 4.86864716122873,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028836967298148317,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22248540222644805,
      "num_tokens": 96772717.0,
      "step": 43350
    },
    {
      "entropy": 5.295404291152954,
      "epoch": 4.869208738136688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028832959499161065,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22597357779741287,
      "num_tokens": 96784152.0,
      "step": 43355
    },
    {
      "entropy": 5.361203241348266,
      "epoch": 4.869770315044645,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002882895165773187,
      "loss": 4.8195,
      "mean_token_accuracy": 0.23361522257328032,
      "num_tokens": 96794312.0,
      "step": 43360
    },
    {
      "entropy": 5.264232873916626,
      "epoch": 4.870331891952603,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002882494377398833,
      "loss": 4.8558,
      "mean_token_accuracy": 0.23449060916900635,
      "num_tokens": 96806327.0,
      "step": 43365
    },
    {
      "entropy": 5.305569696426391,
      "epoch": 4.870893468860561,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00028820935848058066,
      "loss": 4.897,
      "mean_token_accuracy": 0.22720257937908173,
      "num_tokens": 96818264.0,
      "step": 43370
    },
    {
      "entropy": 5.3297031879425045,
      "epoch": 4.871455045768518,
      "grad_norm": 1.015625,
      "learning_rate": 0.000288169278800687,
      "loss": 4.8295,
      "mean_token_accuracy": 0.23555526435375213,
      "num_tokens": 96828082.0,
      "step": 43375
    },
    {
      "entropy": 5.397797775268555,
      "epoch": 4.872016622676475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028812919870147837,
      "loss": 4.9364,
      "mean_token_accuracy": 0.23351891040802003,
      "num_tokens": 96838269.0,
      "step": 43380
    },
    {
      "entropy": 5.255085754394531,
      "epoch": 4.872578199584433,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028808911818423096,
      "loss": 4.7832,
      "mean_token_accuracy": 0.23676198720932007,
      "num_tokens": 96847931.0,
      "step": 43385
    },
    {
      "entropy": 5.329758787155152,
      "epoch": 4.8731397764923905,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028804903725022094,
      "loss": 4.8874,
      "mean_token_accuracy": 0.22854242622852325,
      "num_tokens": 96858721.0,
      "step": 43390
    },
    {
      "entropy": 5.296429824829102,
      "epoch": 4.873701353400348,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002880089559007245,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23064084798097612,
      "num_tokens": 96869928.0,
      "step": 43395
    },
    {
      "entropy": 5.285370397567749,
      "epoch": 4.874262930308306,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002879688741370179,
      "loss": 4.8081,
      "mean_token_accuracy": 0.23379067778587342,
      "num_tokens": 96880343.0,
      "step": 43400
    },
    {
      "entropy": 5.3252222537994385,
      "epoch": 4.874824507216263,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002879287919603773,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22885009944438933,
      "num_tokens": 96892175.0,
      "step": 43405
    },
    {
      "entropy": 5.312579107284546,
      "epoch": 4.875386084124221,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028788870937207896,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22775400280952454,
      "num_tokens": 96903312.0,
      "step": 43410
    },
    {
      "entropy": 5.2522852420806885,
      "epoch": 4.875947661032178,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002878486263733991,
      "loss": 4.7285,
      "mean_token_accuracy": 0.23988295942544938,
      "num_tokens": 96914383.0,
      "step": 43415
    },
    {
      "entropy": 5.25719690322876,
      "epoch": 4.876509237940136,
      "grad_norm": 0.96875,
      "learning_rate": 0.000287808542965614,
      "loss": 4.7243,
      "mean_token_accuracy": 0.2466825872659683,
      "num_tokens": 96925340.0,
      "step": 43420
    },
    {
      "entropy": 5.266987085342407,
      "epoch": 4.877070814848094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028776845914999996,
      "loss": 4.8718,
      "mean_token_accuracy": 0.22936150431632996,
      "num_tokens": 96937138.0,
      "step": 43425
    },
    {
      "entropy": 5.3050812721252445,
      "epoch": 4.877632391756051,
      "grad_norm": 1.0,
      "learning_rate": 0.0002877283749278332,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22736894935369492,
      "num_tokens": 96948645.0,
      "step": 43430
    },
    {
      "entropy": 5.267594242095948,
      "epoch": 4.878193968664009,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028768829030039,
      "loss": 4.8196,
      "mean_token_accuracy": 0.23601215481758117,
      "num_tokens": 96959090.0,
      "step": 43435
    },
    {
      "entropy": 5.329309797286987,
      "epoch": 4.878755545571966,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028764820526894675,
      "loss": 4.9405,
      "mean_token_accuracy": 0.22125795781612395,
      "num_tokens": 96968679.0,
      "step": 43440
    },
    {
      "entropy": 5.357925081253052,
      "epoch": 4.879317122479923,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028760811983477976,
      "loss": 4.926,
      "mean_token_accuracy": 0.23245888203382492,
      "num_tokens": 96980180.0,
      "step": 43445
    },
    {
      "entropy": 5.33661150932312,
      "epoch": 4.8798786993878815,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028756803399916534,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22769194096326828,
      "num_tokens": 96991418.0,
      "step": 43450
    },
    {
      "entropy": 5.321838760375977,
      "epoch": 4.880440276295839,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002875279477633798,
      "loss": 4.8555,
      "mean_token_accuracy": 0.23519182354211807,
      "num_tokens": 97002052.0,
      "step": 43455
    },
    {
      "entropy": 5.252468824386597,
      "epoch": 4.881001853203796,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002874878611286995,
      "loss": 4.8535,
      "mean_token_accuracy": 0.229627887904644,
      "num_tokens": 97014303.0,
      "step": 43460
    },
    {
      "entropy": 5.239798498153687,
      "epoch": 4.881563430111754,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000287447774096401,
      "loss": 4.734,
      "mean_token_accuracy": 0.24361055642366408,
      "num_tokens": 97025716.0,
      "step": 43465
    },
    {
      "entropy": 5.245723295211792,
      "epoch": 4.882125007019711,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028740768666776046,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22692623287439345,
      "num_tokens": 97037285.0,
      "step": 43470
    },
    {
      "entropy": 5.258013916015625,
      "epoch": 4.8826865839276685,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00028736759884405433,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22613944709300995,
      "num_tokens": 97048601.0,
      "step": 43475
    },
    {
      "entropy": 5.309021377563477,
      "epoch": 4.883248160835627,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002873275106265591,
      "loss": 4.7743,
      "mean_token_accuracy": 0.23627099841833116,
      "num_tokens": 97059763.0,
      "step": 43480
    },
    {
      "entropy": 5.363088893890381,
      "epoch": 4.883809737743584,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028728742201655116,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22974938601255418,
      "num_tokens": 97070210.0,
      "step": 43485
    },
    {
      "entropy": 5.421697282791138,
      "epoch": 4.884371314651542,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000287247333015307,
      "loss": 4.999,
      "mean_token_accuracy": 0.2225547879934311,
      "num_tokens": 97081625.0,
      "step": 43490
    },
    {
      "entropy": 5.3555773258209225,
      "epoch": 4.884932891559499,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002872072436241029,
      "loss": 4.9047,
      "mean_token_accuracy": 0.22872087508440017,
      "num_tokens": 97092414.0,
      "step": 43495
    },
    {
      "entropy": 5.292194557189942,
      "epoch": 4.885494468467456,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002871671538442155,
      "loss": 4.8677,
      "mean_token_accuracy": 0.23707842826843262,
      "num_tokens": 97103360.0,
      "step": 43500
    },
    {
      "entropy": 5.24903244972229,
      "epoch": 4.8860560453754145,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002871270636769212,
      "loss": 4.7843,
      "mean_token_accuracy": 0.23223959803581237,
      "num_tokens": 97115128.0,
      "step": 43505
    },
    {
      "entropy": 5.268277978897094,
      "epoch": 4.886617622283372,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002870869731234965,
      "loss": 4.8032,
      "mean_token_accuracy": 0.24308907985687256,
      "num_tokens": 97126696.0,
      "step": 43510
    },
    {
      "entropy": 5.360691404342651,
      "epoch": 4.887179199191329,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002870468821852178,
      "loss": 4.8967,
      "mean_token_accuracy": 0.2323077902197838,
      "num_tokens": 97139072.0,
      "step": 43515
    },
    {
      "entropy": 5.317084074020386,
      "epoch": 4.887740776099287,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028700679086336175,
      "loss": 4.8581,
      "mean_token_accuracy": 0.23282939791679383,
      "num_tokens": 97150525.0,
      "step": 43520
    },
    {
      "entropy": 5.260127973556519,
      "epoch": 4.888302353007244,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002869666991592048,
      "loss": 4.8186,
      "mean_token_accuracy": 0.2285119354724884,
      "num_tokens": 97161862.0,
      "step": 43525
    },
    {
      "entropy": 5.317036294937134,
      "epoch": 4.8888639299152015,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002869266070740235,
      "loss": 4.8305,
      "mean_token_accuracy": 0.23852730095386504,
      "num_tokens": 97171875.0,
      "step": 43530
    },
    {
      "entropy": 5.34832911491394,
      "epoch": 4.88942550682316,
      "grad_norm": 1.0,
      "learning_rate": 0.0002868865146090944,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23183713555336,
      "num_tokens": 97182683.0,
      "step": 43535
    },
    {
      "entropy": 5.406647396087647,
      "epoch": 4.889987083731117,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00028684642176569414,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22994759529829026,
      "num_tokens": 97192983.0,
      "step": 43540
    },
    {
      "entropy": 5.290046072006225,
      "epoch": 4.890548660639075,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002868063285450992,
      "loss": 4.8771,
      "mean_token_accuracy": 0.2355479910969734,
      "num_tokens": 97203000.0,
      "step": 43545
    },
    {
      "entropy": 5.289004611968994,
      "epoch": 4.891110237547032,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002867662349485861,
      "loss": 4.8066,
      "mean_token_accuracy": 0.23648800551891327,
      "num_tokens": 97213103.0,
      "step": 43550
    },
    {
      "entropy": 5.378039264678955,
      "epoch": 4.891671814454989,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028672614097743156,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2248881533741951,
      "num_tokens": 97224123.0,
      "step": 43555
    },
    {
      "entropy": 5.249356365203857,
      "epoch": 4.892233391362947,
      "grad_norm": 0.96875,
      "learning_rate": 0.00028668604663291213,
      "loss": 4.8289,
      "mean_token_accuracy": 0.22726954221725465,
      "num_tokens": 97235849.0,
      "step": 43560
    },
    {
      "entropy": 5.371703052520752,
      "epoch": 4.892794968270905,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028664595191630455,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22203558087348937,
      "num_tokens": 97247382.0,
      "step": 43565
    },
    {
      "entropy": 5.372456455230713,
      "epoch": 4.893356545178862,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028660585682888524,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22958250045776368,
      "num_tokens": 97258704.0,
      "step": 43570
    },
    {
      "entropy": 5.3771453380584715,
      "epoch": 4.89391812208682,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028656576137193096,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22788091748952866,
      "num_tokens": 97270080.0,
      "step": 43575
    },
    {
      "entropy": 5.342940998077393,
      "epoch": 4.894479698994777,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002865256655467184,
      "loss": 4.827,
      "mean_token_accuracy": 0.23564284145832062,
      "num_tokens": 97281560.0,
      "step": 43580
    },
    {
      "entropy": 5.2959856510162355,
      "epoch": 4.895041275902734,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002864855693545241,
      "loss": 4.841,
      "mean_token_accuracy": 0.22396003305912018,
      "num_tokens": 97292970.0,
      "step": 43585
    },
    {
      "entropy": 5.227031898498535,
      "epoch": 4.8956028528106925,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028644547279662485,
      "loss": 4.7918,
      "mean_token_accuracy": 0.24001364260911942,
      "num_tokens": 97302481.0,
      "step": 43590
    },
    {
      "entropy": 5.447815990447998,
      "epoch": 4.89616442971865,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028640537587429734,
      "loss": 5.0401,
      "mean_token_accuracy": 0.2186543047428131,
      "num_tokens": 97315382.0,
      "step": 43595
    },
    {
      "entropy": 5.335455369949341,
      "epoch": 4.896726006626608,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002863652785888182,
      "loss": 4.7676,
      "mean_token_accuracy": 0.24200854748487471,
      "num_tokens": 97325657.0,
      "step": 43600
    },
    {
      "entropy": 5.29167628288269,
      "epoch": 4.897287583534565,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002863251809414642,
      "loss": 4.8577,
      "mean_token_accuracy": 0.23164136707782745,
      "num_tokens": 97336792.0,
      "step": 43605
    },
    {
      "entropy": 5.270018148422241,
      "epoch": 4.897849160442522,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000286285082933512,
      "loss": 4.8939,
      "mean_token_accuracy": 0.23587997257709503,
      "num_tokens": 97348439.0,
      "step": 43610
    },
    {
      "entropy": 5.348113107681274,
      "epoch": 4.89841073735048,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002862449845662385,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22562716752290726,
      "num_tokens": 97360429.0,
      "step": 43615
    },
    {
      "entropy": 5.386110639572143,
      "epoch": 4.898972314258438,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002862048858409201,
      "loss": 4.9264,
      "mean_token_accuracy": 0.23741941303014755,
      "num_tokens": 97371105.0,
      "step": 43620
    },
    {
      "entropy": 5.269165706634522,
      "epoch": 4.899533891166396,
      "grad_norm": 1.015625,
      "learning_rate": 0.000286164786758834,
      "loss": 4.8166,
      "mean_token_accuracy": 0.2354582816362381,
      "num_tokens": 97381789.0,
      "step": 43625
    },
    {
      "entropy": 5.261893081665039,
      "epoch": 4.900095468074353,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028612468732125667,
      "loss": 4.9326,
      "mean_token_accuracy": 0.2203940525650978,
      "num_tokens": 97393653.0,
      "step": 43630
    },
    {
      "entropy": 5.306213855743408,
      "epoch": 4.90065704498231,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002860845875294651,
      "loss": 4.8415,
      "mean_token_accuracy": 0.231693859398365,
      "num_tokens": 97403475.0,
      "step": 43635
    },
    {
      "entropy": 5.371659851074218,
      "epoch": 4.901218621890268,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002860444873847358,
      "loss": 4.9388,
      "mean_token_accuracy": 0.22735889703035356,
      "num_tokens": 97415124.0,
      "step": 43640
    },
    {
      "entropy": 5.346455335617065,
      "epoch": 4.9017801987982255,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028600438688834587,
      "loss": 4.9051,
      "mean_token_accuracy": 0.23143184036016465,
      "num_tokens": 97426019.0,
      "step": 43645
    },
    {
      "entropy": 5.300188875198364,
      "epoch": 4.902341775706183,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000285964286041572,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22373491674661636,
      "num_tokens": 97437678.0,
      "step": 43650
    },
    {
      "entropy": 5.30559287071228,
      "epoch": 4.902903352614141,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028592418484569095,
      "loss": 4.817,
      "mean_token_accuracy": 0.23448545783758162,
      "num_tokens": 97448884.0,
      "step": 43655
    },
    {
      "entropy": 5.318466138839722,
      "epoch": 4.903464929522098,
      "grad_norm": 1.0,
      "learning_rate": 0.0002858840833019795,
      "loss": 4.8364,
      "mean_token_accuracy": 0.23811340630054473,
      "num_tokens": 97459875.0,
      "step": 43660
    },
    {
      "entropy": 5.299627876281738,
      "epoch": 4.904026506430055,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002858439814117148,
      "loss": 4.7693,
      "mean_token_accuracy": 0.24229296892881394,
      "num_tokens": 97470238.0,
      "step": 43665
    },
    {
      "entropy": 5.3799244403839115,
      "epoch": 4.904588083338013,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028580387917617343,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23038016110658646,
      "num_tokens": 97481561.0,
      "step": 43670
    },
    {
      "entropy": 5.373349237442016,
      "epoch": 4.9051496602459705,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002857637765966324,
      "loss": 4.9768,
      "mean_token_accuracy": 0.22437910735607147,
      "num_tokens": 97493959.0,
      "step": 43675
    },
    {
      "entropy": 5.264508390426636,
      "epoch": 4.905711237153929,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002857236736743685,
      "loss": 4.8345,
      "mean_token_accuracy": 0.236747707426548,
      "num_tokens": 97505613.0,
      "step": 43680
    },
    {
      "entropy": 5.349529647827149,
      "epoch": 4.906272814061886,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002856835704106587,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22886948734521867,
      "num_tokens": 97517730.0,
      "step": 43685
    },
    {
      "entropy": 5.274074554443359,
      "epoch": 4.906834390969843,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028564346680677995,
      "loss": 4.7701,
      "mean_token_accuracy": 0.2371492028236389,
      "num_tokens": 97528194.0,
      "step": 43690
    },
    {
      "entropy": 5.287001609802246,
      "epoch": 4.907395967877801,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002856033628640089,
      "loss": 4.8541,
      "mean_token_accuracy": 0.22040757536888123,
      "num_tokens": 97538980.0,
      "step": 43695
    },
    {
      "entropy": 5.326306962966919,
      "epoch": 4.907957544785758,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002855632585836228,
      "loss": 4.896,
      "mean_token_accuracy": 0.22839174419641495,
      "num_tokens": 97550250.0,
      "step": 43700
    },
    {
      "entropy": 5.363025236129761,
      "epoch": 4.908519121693716,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028552315396689846,
      "loss": 4.929,
      "mean_token_accuracy": 0.2265319123864174,
      "num_tokens": 97561687.0,
      "step": 43705
    },
    {
      "entropy": 5.389482307434082,
      "epoch": 4.909080698601674,
      "grad_norm": 1.0,
      "learning_rate": 0.0002854830490151128,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22430268079042434,
      "num_tokens": 97572820.0,
      "step": 43710
    },
    {
      "entropy": 5.263682889938354,
      "epoch": 4.909642275509631,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028544294372954276,
      "loss": 4.8469,
      "mean_token_accuracy": 0.23596428632736205,
      "num_tokens": 97583556.0,
      "step": 43715
    },
    {
      "entropy": 5.36049747467041,
      "epoch": 4.910203852417588,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002854028381114654,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22820814549922944,
      "num_tokens": 97594503.0,
      "step": 43720
    },
    {
      "entropy": 5.275243806838989,
      "epoch": 4.910765429325546,
      "grad_norm": 1.0,
      "learning_rate": 0.00028536273216215756,
      "loss": 4.7679,
      "mean_token_accuracy": 0.23475221246480943,
      "num_tokens": 97603811.0,
      "step": 43725
    },
    {
      "entropy": 5.405625104904175,
      "epoch": 4.9113270062335035,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028532262588289636,
      "loss": 4.9555,
      "mean_token_accuracy": 0.22812681049108505,
      "num_tokens": 97614573.0,
      "step": 43730
    },
    {
      "entropy": 5.3233311653137205,
      "epoch": 4.911888583141462,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028528251927495874,
      "loss": 4.8184,
      "mean_token_accuracy": 0.23509712964296342,
      "num_tokens": 97625928.0,
      "step": 43735
    },
    {
      "entropy": 5.285038328170776,
      "epoch": 4.912450160049419,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00028524241233962176,
      "loss": 4.8217,
      "mean_token_accuracy": 0.2273826077580452,
      "num_tokens": 97635763.0,
      "step": 43740
    },
    {
      "entropy": 5.3496777534484865,
      "epoch": 4.913011736957376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028520230507816236,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22293906062841415,
      "num_tokens": 97648460.0,
      "step": 43745
    },
    {
      "entropy": 5.312964248657226,
      "epoch": 4.913573313865334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028516219749185755,
      "loss": 4.8902,
      "mean_token_accuracy": 0.23015811890363694,
      "num_tokens": 97659997.0,
      "step": 43750
    },
    {
      "entropy": 5.3918619632720945,
      "epoch": 4.914134890773291,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002851220895819845,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2331821545958519,
      "num_tokens": 97671709.0,
      "step": 43755
    },
    {
      "entropy": 5.3747790336608885,
      "epoch": 4.914696467681249,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028508198134982016,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22697795629501344,
      "num_tokens": 97683237.0,
      "step": 43760
    },
    {
      "entropy": 5.375089073181153,
      "epoch": 4.915258044589207,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028504187279664164,
      "loss": 4.899,
      "mean_token_accuracy": 0.2325640082359314,
      "num_tokens": 97694979.0,
      "step": 43765
    },
    {
      "entropy": 5.313180875778198,
      "epoch": 4.915819621497164,
      "grad_norm": 0.984375,
      "learning_rate": 0.000285001763923726,
      "loss": 4.8439,
      "mean_token_accuracy": 0.23218737095594405,
      "num_tokens": 97705525.0,
      "step": 43770
    },
    {
      "entropy": 5.316745567321777,
      "epoch": 4.916381198405121,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002849616547323503,
      "loss": 4.9281,
      "mean_token_accuracy": 0.2307622402906418,
      "num_tokens": 97715679.0,
      "step": 43775
    },
    {
      "entropy": 5.36646466255188,
      "epoch": 4.916942775313079,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028492154522379167,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22269349843263625,
      "num_tokens": 97727567.0,
      "step": 43780
    },
    {
      "entropy": 5.3884687423706055,
      "epoch": 4.917504352221036,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002848814353993273,
      "loss": 4.8699,
      "mean_token_accuracy": 0.23473931699991227,
      "num_tokens": 97738066.0,
      "step": 43785
    },
    {
      "entropy": 5.293538379669189,
      "epoch": 4.9180659291289945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028484132526023404,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2364138826727867,
      "num_tokens": 97748839.0,
      "step": 43790
    },
    {
      "entropy": 5.361705541610718,
      "epoch": 4.918627506036952,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002848012148077892,
      "loss": 4.908,
      "mean_token_accuracy": 0.2308024361729622,
      "num_tokens": 97759762.0,
      "step": 43795
    },
    {
      "entropy": 5.359376668930054,
      "epoch": 4.919189082944909,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028476110404326984,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22927218526601792,
      "num_tokens": 97770546.0,
      "step": 43800
    },
    {
      "entropy": 5.287531423568725,
      "epoch": 4.919750659852867,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0002847209929679532,
      "loss": 4.8303,
      "mean_token_accuracy": 0.2370566263794899,
      "num_tokens": 97781768.0,
      "step": 43805
    },
    {
      "entropy": 5.304828691482544,
      "epoch": 4.920312236760824,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002846808815831164,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23243265450000763,
      "num_tokens": 97793132.0,
      "step": 43810
    },
    {
      "entropy": 5.310479593276978,
      "epoch": 4.920873813668782,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002846407698900365,
      "loss": 4.9424,
      "mean_token_accuracy": 0.2304024338722229,
      "num_tokens": 97804136.0,
      "step": 43815
    },
    {
      "entropy": 5.399616813659668,
      "epoch": 4.92143539057674,
      "grad_norm": 1.125,
      "learning_rate": 0.00028460065788999073,
      "loss": 4.9547,
      "mean_token_accuracy": 0.23554362952709199,
      "num_tokens": 97814722.0,
      "step": 43820
    },
    {
      "entropy": 5.410502290725708,
      "epoch": 4.921996967484697,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002845605455842564,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2298137679696083,
      "num_tokens": 97826131.0,
      "step": 43825
    },
    {
      "entropy": 5.374788808822632,
      "epoch": 4.922558544392655,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002845204329741106,
      "loss": 4.966,
      "mean_token_accuracy": 0.22769310027360917,
      "num_tokens": 97837401.0,
      "step": 43830
    },
    {
      "entropy": 5.364683341979981,
      "epoch": 4.923120121300612,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002844803200608304,
      "loss": 4.9016,
      "mean_token_accuracy": 0.23265491873025895,
      "num_tokens": 97848559.0,
      "step": 43835
    },
    {
      "entropy": 5.367771244049072,
      "epoch": 4.923681698208569,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002844402068456931,
      "loss": 4.8362,
      "mean_token_accuracy": 0.22819226682186128,
      "num_tokens": 97860384.0,
      "step": 43840
    },
    {
      "entropy": 5.285113143920898,
      "epoch": 4.9242432751165275,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002844000933299761,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2321562111377716,
      "num_tokens": 97871060.0,
      "step": 43845
    },
    {
      "entropy": 5.247565126419067,
      "epoch": 4.924804852024485,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00028435997951495643,
      "loss": 4.8099,
      "mean_token_accuracy": 0.2360501617193222,
      "num_tokens": 97881449.0,
      "step": 43850
    },
    {
      "entropy": 5.257348823547363,
      "epoch": 4.925366428932442,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002843198654019114,
      "loss": 4.7791,
      "mean_token_accuracy": 0.2396544799208641,
      "num_tokens": 97894204.0,
      "step": 43855
    },
    {
      "entropy": 5.364230680465698,
      "epoch": 4.9259280058404,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002842797509921182,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22929206639528274,
      "num_tokens": 97906690.0,
      "step": 43860
    },
    {
      "entropy": 5.317640924453736,
      "epoch": 4.926489582748357,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028423963628685414,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22905247956514357,
      "num_tokens": 97918006.0,
      "step": 43865
    },
    {
      "entropy": 5.314087724685669,
      "epoch": 4.927051159656315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002841995212873965,
      "loss": 4.8337,
      "mean_token_accuracy": 0.23329581916332245,
      "num_tokens": 97928904.0,
      "step": 43870
    },
    {
      "entropy": 5.394802188873291,
      "epoch": 4.927612736564273,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002841594059950225,
      "loss": 5.0332,
      "mean_token_accuracy": 0.21941060125827788,
      "num_tokens": 97942379.0,
      "step": 43875
    },
    {
      "entropy": 5.3418519496917725,
      "epoch": 4.92817431347223,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028411929041100947,
      "loss": 4.8208,
      "mean_token_accuracy": 0.23636738657951356,
      "num_tokens": 97952462.0,
      "step": 43880
    },
    {
      "entropy": 5.343882894515991,
      "epoch": 4.928735890380188,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002840791745366347,
      "loss": 4.8899,
      "mean_token_accuracy": 0.22578718811273574,
      "num_tokens": 97964798.0,
      "step": 43885
    },
    {
      "entropy": 5.294370412826538,
      "epoch": 4.929297467288145,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028403905837317547,
      "loss": 4.9477,
      "mean_token_accuracy": 0.22700662910938263,
      "num_tokens": 97975617.0,
      "step": 43890
    },
    {
      "entropy": 5.2771800518035885,
      "epoch": 4.929859044196102,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028399894192190916,
      "loss": 4.8346,
      "mean_token_accuracy": 0.23581611216068268,
      "num_tokens": 97986404.0,
      "step": 43895
    },
    {
      "entropy": 5.3818747997283936,
      "epoch": 4.93042062110406,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00028395882518411305,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22851255983114244,
      "num_tokens": 97999256.0,
      "step": 43900
    },
    {
      "entropy": 5.391334676742554,
      "epoch": 4.930982198012018,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002839187081610645,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22823070883750915,
      "num_tokens": 98010018.0,
      "step": 43905
    },
    {
      "entropy": 5.358220148086548,
      "epoch": 4.931543774919975,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002838785908540407,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22924142330884933,
      "num_tokens": 98020485.0,
      "step": 43910
    },
    {
      "entropy": 5.271506643295288,
      "epoch": 4.932105351827933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028383847326431924,
      "loss": 4.854,
      "mean_token_accuracy": 0.23586402982473373,
      "num_tokens": 98030495.0,
      "step": 43915
    },
    {
      "entropy": 5.283480834960938,
      "epoch": 4.93266692873589,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00028379835539317727,
      "loss": 4.8869,
      "mean_token_accuracy": 0.2284668579697609,
      "num_tokens": 98041623.0,
      "step": 43920
    },
    {
      "entropy": 5.322345876693726,
      "epoch": 4.933228505643848,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002837582372418923,
      "loss": 4.7975,
      "mean_token_accuracy": 0.24158926606178283,
      "num_tokens": 98052045.0,
      "step": 43925
    },
    {
      "entropy": 5.406316757202148,
      "epoch": 4.9337900825518055,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002837181188117417,
      "loss": 4.9347,
      "mean_token_accuracy": 0.2243686005473137,
      "num_tokens": 98064358.0,
      "step": 43930
    },
    {
      "entropy": 5.282062959671021,
      "epoch": 4.934351659459763,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002836780001040028,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23010243028402327,
      "num_tokens": 98075328.0,
      "step": 43935
    },
    {
      "entropy": 5.309149932861328,
      "epoch": 4.934913236367721,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000283637881119953,
      "loss": 4.9097,
      "mean_token_accuracy": 0.2245876297354698,
      "num_tokens": 98085588.0,
      "step": 43940
    },
    {
      "entropy": 5.288299322128296,
      "epoch": 4.935474813275678,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028359776186086965,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22747464925050737,
      "num_tokens": 98096249.0,
      "step": 43945
    },
    {
      "entropy": 5.39266939163208,
      "epoch": 4.936036390183635,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002835576423280303,
      "loss": 4.8984,
      "mean_token_accuracy": 0.22641366273164748,
      "num_tokens": 98107498.0,
      "step": 43950
    },
    {
      "entropy": 5.388674068450928,
      "epoch": 4.936597967091593,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002835175225227123,
      "loss": 4.9344,
      "mean_token_accuracy": 0.23377058804035186,
      "num_tokens": 98119918.0,
      "step": 43955
    },
    {
      "entropy": 5.293115234375,
      "epoch": 4.937159543999551,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002834774024461931,
      "loss": 4.8412,
      "mean_token_accuracy": 0.22545841485261917,
      "num_tokens": 98130923.0,
      "step": 43960
    },
    {
      "entropy": 5.265460586547851,
      "epoch": 4.937721120907508,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028343728209975,
      "loss": 4.7945,
      "mean_token_accuracy": 0.238722762465477,
      "num_tokens": 98142192.0,
      "step": 43965
    },
    {
      "entropy": 5.290746641159058,
      "epoch": 4.938282697815466,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028339716148466067,
      "loss": 4.849,
      "mean_token_accuracy": 0.2260967805981636,
      "num_tokens": 98152479.0,
      "step": 43970
    },
    {
      "entropy": 5.318809652328492,
      "epoch": 4.938844274723423,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028335704060220246,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23584267795085906,
      "num_tokens": 98162479.0,
      "step": 43975
    },
    {
      "entropy": 5.34360637664795,
      "epoch": 4.939405851631381,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00028331691945365274,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22377211600542068,
      "num_tokens": 98173310.0,
      "step": 43980
    },
    {
      "entropy": 5.44387264251709,
      "epoch": 4.9399674285393385,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002832767980402892,
      "loss": 5.0435,
      "mean_token_accuracy": 0.22565365731716155,
      "num_tokens": 98185444.0,
      "step": 43985
    },
    {
      "entropy": 5.322139263153076,
      "epoch": 4.940529005447296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002832366763633892,
      "loss": 4.8656,
      "mean_token_accuracy": 0.23601166009902955,
      "num_tokens": 98195898.0,
      "step": 43990
    },
    {
      "entropy": 5.329060792922974,
      "epoch": 4.941090582355254,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028319655442423015,
      "loss": 4.8746,
      "mean_token_accuracy": 0.22898177057504654,
      "num_tokens": 98205957.0,
      "step": 43995
    },
    {
      "entropy": 5.301036691665649,
      "epoch": 4.941652159263211,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028315643222408965,
      "loss": 4.875,
      "mean_token_accuracy": 0.22958087176084518,
      "num_tokens": 98216502.0,
      "step": 44000
    },
    {
      "entropy": 5.3527936935424805,
      "epoch": 4.942213736171169,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028311630976424526,
      "loss": 4.8756,
      "mean_token_accuracy": 0.23058597445487977,
      "num_tokens": 98226798.0,
      "step": 44005
    },
    {
      "entropy": 5.416969442367554,
      "epoch": 4.942775313079126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002830761870459743,
      "loss": 4.9766,
      "mean_token_accuracy": 0.22520416975021362,
      "num_tokens": 98238100.0,
      "step": 44010
    },
    {
      "entropy": 5.406939172744751,
      "epoch": 4.9433368899870835,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002830360640705545,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2317142218351364,
      "num_tokens": 98248800.0,
      "step": 44015
    },
    {
      "entropy": 5.36112060546875,
      "epoch": 4.943898466895042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028299594083926324,
      "loss": 4.8513,
      "mean_token_accuracy": 0.2330041527748108,
      "num_tokens": 98258326.0,
      "step": 44020
    },
    {
      "entropy": 5.298646402359009,
      "epoch": 4.944460043802999,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028295581735337817,
      "loss": 4.8698,
      "mean_token_accuracy": 0.23104314804077147,
      "num_tokens": 98269055.0,
      "step": 44025
    },
    {
      "entropy": 5.307069110870361,
      "epoch": 4.945021620710956,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00028291569361417674,
      "loss": 4.8413,
      "mean_token_accuracy": 0.2314005896449089,
      "num_tokens": 98280683.0,
      "step": 44030
    },
    {
      "entropy": 5.451658058166504,
      "epoch": 4.945583197618914,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002828755696229366,
      "loss": 5.0606,
      "mean_token_accuracy": 0.2241984024643898,
      "num_tokens": 98292427.0,
      "step": 44035
    },
    {
      "entropy": 5.338675022125244,
      "epoch": 4.946144774526871,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002828354453809352,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22993558794260024,
      "num_tokens": 98302760.0,
      "step": 44040
    },
    {
      "entropy": 5.343034172058106,
      "epoch": 4.946706351434829,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002827953208894503,
      "loss": 4.853,
      "mean_token_accuracy": 0.22973467260599137,
      "num_tokens": 98313643.0,
      "step": 44045
    },
    {
      "entropy": 5.409980964660645,
      "epoch": 4.947267928342787,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002827551961497593,
      "loss": 4.9646,
      "mean_token_accuracy": 0.23023361414670945,
      "num_tokens": 98323866.0,
      "step": 44050
    },
    {
      "entropy": 5.3541711330413815,
      "epoch": 4.947829505250744,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002827150711631398,
      "loss": 4.9181,
      "mean_token_accuracy": 0.22667600810527802,
      "num_tokens": 98335683.0,
      "step": 44055
    },
    {
      "entropy": 5.38443660736084,
      "epoch": 4.948391082158702,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00028267494593086945,
      "loss": 4.9362,
      "mean_token_accuracy": 0.23792639076709748,
      "num_tokens": 98347553.0,
      "step": 44060
    },
    {
      "entropy": 5.342494201660156,
      "epoch": 4.948952659066659,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00028263482045422597,
      "loss": 4.8744,
      "mean_token_accuracy": 0.2372132882475853,
      "num_tokens": 98359729.0,
      "step": 44065
    },
    {
      "entropy": 5.406559991836548,
      "epoch": 4.9495142359746165,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002825946947344867,
      "loss": 4.9567,
      "mean_token_accuracy": 0.23028545528650285,
      "num_tokens": 98371029.0,
      "step": 44070
    },
    {
      "entropy": 5.334176921844483,
      "epoch": 4.950075812882575,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002825545687729295,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22301337718963624,
      "num_tokens": 98380911.0,
      "step": 44075
    },
    {
      "entropy": 5.327294731140137,
      "epoch": 4.950637389790532,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028251444257083186,
      "loss": 4.7744,
      "mean_token_accuracy": 0.241789810359478,
      "num_tokens": 98390341.0,
      "step": 44080
    },
    {
      "entropy": 5.356728029251099,
      "epoch": 4.951198966698489,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028247431612947146,
      "loss": 4.9695,
      "mean_token_accuracy": 0.22658490091562272,
      "num_tokens": 98401707.0,
      "step": 44085
    },
    {
      "entropy": 5.379615306854248,
      "epoch": 4.951760543606447,
      "grad_norm": 1.0,
      "learning_rate": 0.0002824341894501259,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22883762419223785,
      "num_tokens": 98413226.0,
      "step": 44090
    },
    {
      "entropy": 5.35163836479187,
      "epoch": 4.952322120514404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028239406253407286,
      "loss": 4.895,
      "mean_token_accuracy": 0.23150481134653092,
      "num_tokens": 98424805.0,
      "step": 44095
    },
    {
      "entropy": 5.34906234741211,
      "epoch": 4.952883697422362,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028235393538259007,
      "loss": 4.907,
      "mean_token_accuracy": 0.22492167502641677,
      "num_tokens": 98435725.0,
      "step": 44100
    },
    {
      "entropy": 5.356379795074463,
      "epoch": 4.95344527433032,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002823138079969551,
      "loss": 4.8522,
      "mean_token_accuracy": 0.23155504912137986,
      "num_tokens": 98447014.0,
      "step": 44105
    },
    {
      "entropy": 5.3878199577331545,
      "epoch": 4.954006851238277,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028227368037844557,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22939899563789368,
      "num_tokens": 98458902.0,
      "step": 44110
    },
    {
      "entropy": 5.298256063461304,
      "epoch": 4.954568428146235,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00028223355252833933,
      "loss": 4.815,
      "mean_token_accuracy": 0.24126108437776567,
      "num_tokens": 98471173.0,
      "step": 44115
    },
    {
      "entropy": 5.2723030090332035,
      "epoch": 4.955130005054192,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028219342444791396,
      "loss": 4.8057,
      "mean_token_accuracy": 0.23419490456581116,
      "num_tokens": 98481882.0,
      "step": 44120
    },
    {
      "entropy": 5.323552131652832,
      "epoch": 4.955691581962149,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028215329613844715,
      "loss": 4.8283,
      "mean_token_accuracy": 0.23322323113679885,
      "num_tokens": 98492938.0,
      "step": 44125
    },
    {
      "entropy": 5.398375034332275,
      "epoch": 4.9562531588701075,
      "grad_norm": 0.90625,
      "learning_rate": 0.00028211316760121656,
      "loss": 5.0263,
      "mean_token_accuracy": 0.2238294944167137,
      "num_tokens": 98505980.0,
      "step": 44130
    },
    {
      "entropy": 5.391941785812378,
      "epoch": 4.956814735778065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002820730388375001,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2223310276865959,
      "num_tokens": 98517868.0,
      "step": 44135
    },
    {
      "entropy": 5.344036293029785,
      "epoch": 4.957376312686022,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002820329098485752,
      "loss": 4.7313,
      "mean_token_accuracy": 0.24245921522378922,
      "num_tokens": 98527557.0,
      "step": 44140
    },
    {
      "entropy": 5.290171718597412,
      "epoch": 4.95793788959398,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028199278063571976,
      "loss": 4.8946,
      "mean_token_accuracy": 0.22890387177467347,
      "num_tokens": 98540502.0,
      "step": 44145
    },
    {
      "entropy": 5.310486125946045,
      "epoch": 4.958499466501937,
      "grad_norm": 1.0,
      "learning_rate": 0.00028195265120021157,
      "loss": 4.9051,
      "mean_token_accuracy": 0.23402635902166366,
      "num_tokens": 98551875.0,
      "step": 44150
    },
    {
      "entropy": 5.254725933074951,
      "epoch": 4.9590610434098945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028191252154332816,
      "loss": 4.8557,
      "mean_token_accuracy": 0.22986187189817428,
      "num_tokens": 98561702.0,
      "step": 44155
    },
    {
      "entropy": 5.359773063659668,
      "epoch": 4.959622620317853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028187239166634737,
      "loss": 4.9115,
      "mean_token_accuracy": 0.23624062836170195,
      "num_tokens": 98571112.0,
      "step": 44160
    },
    {
      "entropy": 5.33473014831543,
      "epoch": 4.96018419722581,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028183226157054705,
      "loss": 4.8991,
      "mean_token_accuracy": 0.22513955384492873,
      "num_tokens": 98581943.0,
      "step": 44165
    },
    {
      "entropy": 5.371310663223267,
      "epoch": 4.960745774133768,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002817921312572048,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22784174978733063,
      "num_tokens": 98592452.0,
      "step": 44170
    },
    {
      "entropy": 5.347080135345459,
      "epoch": 4.961307351041725,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002817520007275985,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2303934246301651,
      "num_tokens": 98603527.0,
      "step": 44175
    },
    {
      "entropy": 5.37602744102478,
      "epoch": 4.961868927949682,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002817118699830058,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22198678106069564,
      "num_tokens": 98614359.0,
      "step": 44180
    },
    {
      "entropy": 5.2531122207641605,
      "epoch": 4.9624305048576405,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002816717390247046,
      "loss": 4.7933,
      "mean_token_accuracy": 0.23269341588020326,
      "num_tokens": 98625452.0,
      "step": 44185
    },
    {
      "entropy": 5.290101766586304,
      "epoch": 4.962992081765598,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002816316078539727,
      "loss": 4.8491,
      "mean_token_accuracy": 0.23625773042440415,
      "num_tokens": 98636918.0,
      "step": 44190
    },
    {
      "entropy": 5.338339710235596,
      "epoch": 4.963553658673556,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028159147647208784,
      "loss": 4.9336,
      "mean_token_accuracy": 0.23187313079833985,
      "num_tokens": 98647744.0,
      "step": 44195
    },
    {
      "entropy": 5.273560047149658,
      "epoch": 4.964115235581513,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002815513448803277,
      "loss": 4.7975,
      "mean_token_accuracy": 0.2349267154932022,
      "num_tokens": 98658655.0,
      "step": 44200
    },
    {
      "entropy": 5.331953525543213,
      "epoch": 4.96467681248947,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002815112130799703,
      "loss": 4.9406,
      "mean_token_accuracy": 0.22487879246473313,
      "num_tokens": 98670858.0,
      "step": 44205
    },
    {
      "entropy": 5.2744081020355225,
      "epoch": 4.965238389397428,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028147108107229324,
      "loss": 4.7845,
      "mean_token_accuracy": 0.23304370939731597,
      "num_tokens": 98683044.0,
      "step": 44210
    },
    {
      "entropy": 5.3803290843963625,
      "epoch": 4.965799966305386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028143094885857455,
      "loss": 4.884,
      "mean_token_accuracy": 0.22484272420406343,
      "num_tokens": 98694124.0,
      "step": 44215
    },
    {
      "entropy": 5.320252799987793,
      "epoch": 4.966361543213343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000281390816440092,
      "loss": 4.8668,
      "mean_token_accuracy": 0.2363376334309578,
      "num_tokens": 98704102.0,
      "step": 44220
    },
    {
      "entropy": 5.403891754150391,
      "epoch": 4.966923120121301,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002813506838181233,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22905370146036147,
      "num_tokens": 98715594.0,
      "step": 44225
    },
    {
      "entropy": 5.322282838821411,
      "epoch": 4.967484697029258,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002813105509939464,
      "loss": 4.8572,
      "mean_token_accuracy": 0.23039425164461136,
      "num_tokens": 98727388.0,
      "step": 44230
    },
    {
      "entropy": 5.366693782806396,
      "epoch": 4.968046273937215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000281270417968839,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22971024364233017,
      "num_tokens": 98739471.0,
      "step": 44235
    },
    {
      "entropy": 5.340255546569824,
      "epoch": 4.968607850845173,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002812302847440791,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22576503753662108,
      "num_tokens": 98750490.0,
      "step": 44240
    },
    {
      "entropy": 5.272222995758057,
      "epoch": 4.969169427753131,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028119015132094455,
      "loss": 4.7976,
      "mean_token_accuracy": 0.2317809358239174,
      "num_tokens": 98761170.0,
      "step": 44245
    },
    {
      "entropy": 5.264775657653809,
      "epoch": 4.969731004661089,
      "grad_norm": 1.0,
      "learning_rate": 0.0002811500177007132,
      "loss": 4.8038,
      "mean_token_accuracy": 0.23310671001672745,
      "num_tokens": 98772473.0,
      "step": 44250
    },
    {
      "entropy": 5.355880451202393,
      "epoch": 4.970292581569046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028110988388466287,
      "loss": 4.9135,
      "mean_token_accuracy": 0.2276088833808899,
      "num_tokens": 98783304.0,
      "step": 44255
    },
    {
      "entropy": 5.314635944366455,
      "epoch": 4.970854158477003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002810697498740715,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22704823315143585,
      "num_tokens": 98794201.0,
      "step": 44260
    },
    {
      "entropy": 5.300666761398316,
      "epoch": 4.971415735384961,
      "grad_norm": 1.0,
      "learning_rate": 0.0002810296156702169,
      "loss": 4.7749,
      "mean_token_accuracy": 0.2405395984649658,
      "num_tokens": 98804781.0,
      "step": 44265
    },
    {
      "entropy": 5.364297485351562,
      "epoch": 4.9719773122929185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028098948127437695,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22813138067722322,
      "num_tokens": 98815213.0,
      "step": 44270
    },
    {
      "entropy": 5.294503784179687,
      "epoch": 4.972538889200876,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002809493466878297,
      "loss": 4.8586,
      "mean_token_accuracy": 0.23013588190078735,
      "num_tokens": 98826352.0,
      "step": 44275
    },
    {
      "entropy": 5.284305143356323,
      "epoch": 4.973100466108834,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028090921191185295,
      "loss": 4.7878,
      "mean_token_accuracy": 0.2349633514881134,
      "num_tokens": 98837509.0,
      "step": 44280
    },
    {
      "entropy": 5.349946928024292,
      "epoch": 4.973662043016791,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002808690769477245,
      "loss": 4.916,
      "mean_token_accuracy": 0.2275216683745384,
      "num_tokens": 98848256.0,
      "step": 44285
    },
    {
      "entropy": 5.370415163040161,
      "epoch": 4.974223619924748,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002808289417967225,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2315392017364502,
      "num_tokens": 98859894.0,
      "step": 44290
    },
    {
      "entropy": 5.339399433135986,
      "epoch": 4.974785196832706,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028078880646012467,
      "loss": 4.8563,
      "mean_token_accuracy": 0.2334185242652893,
      "num_tokens": 98871848.0,
      "step": 44295
    },
    {
      "entropy": 5.258039808273315,
      "epoch": 4.975346773740664,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028074867093920897,
      "loss": 4.7564,
      "mean_token_accuracy": 0.24237384051084518,
      "num_tokens": 98882640.0,
      "step": 44300
    },
    {
      "entropy": 5.268483972549438,
      "epoch": 4.975908350648622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028070853523525337,
      "loss": 4.8387,
      "mean_token_accuracy": 0.23144348263740538,
      "num_tokens": 98893478.0,
      "step": 44305
    },
    {
      "entropy": 5.289311027526855,
      "epoch": 4.976469927556579,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028066839934953573,
      "loss": 4.8099,
      "mean_token_accuracy": 0.23863814175128936,
      "num_tokens": 98903713.0,
      "step": 44310
    },
    {
      "entropy": 5.388206672668457,
      "epoch": 4.977031504464536,
      "grad_norm": 0.890625,
      "learning_rate": 0.00028062826328333415,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2344930127263069,
      "num_tokens": 98915622.0,
      "step": 44315
    },
    {
      "entropy": 5.282104730606079,
      "epoch": 4.977593081372494,
      "grad_norm": 1.125,
      "learning_rate": 0.0002805881270379265,
      "loss": 4.7976,
      "mean_token_accuracy": 0.23761697560548783,
      "num_tokens": 98926676.0,
      "step": 44320
    },
    {
      "entropy": 5.3667596817016605,
      "epoch": 4.9781546582804515,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002805479906145906,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2209734782576561,
      "num_tokens": 98937562.0,
      "step": 44325
    },
    {
      "entropy": 5.289053440093994,
      "epoch": 4.978716235188409,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002805078540146046,
      "loss": 4.8313,
      "mean_token_accuracy": 0.23370797485113143,
      "num_tokens": 98948968.0,
      "step": 44330
    },
    {
      "entropy": 5.384174489974976,
      "epoch": 4.979277812096367,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002804677172392464,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22409718185663224,
      "num_tokens": 98960222.0,
      "step": 44335
    },
    {
      "entropy": 5.489224863052368,
      "epoch": 4.979839389004324,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000280427580289794,
      "loss": 4.9482,
      "mean_token_accuracy": 0.22779795229434968,
      "num_tokens": 98973613.0,
      "step": 44340
    },
    {
      "entropy": 5.32267804145813,
      "epoch": 4.980400965912281,
      "grad_norm": 1.140625,
      "learning_rate": 0.00028038744316752524,
      "loss": 4.9047,
      "mean_token_accuracy": 0.22170582711696624,
      "num_tokens": 98984431.0,
      "step": 44345
    },
    {
      "entropy": 5.243948364257813,
      "epoch": 4.980962542820239,
      "grad_norm": 1.0,
      "learning_rate": 0.00028034730587371824,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23071282356977463,
      "num_tokens": 98995860.0,
      "step": 44350
    },
    {
      "entropy": 5.315756511688233,
      "epoch": 4.9815241197281965,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00028030716840965096,
      "loss": 4.7933,
      "mean_token_accuracy": 0.23636251240968703,
      "num_tokens": 99006892.0,
      "step": 44355
    },
    {
      "entropy": 5.38933629989624,
      "epoch": 4.982085696636155,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028026703077660143,
      "loss": 4.8342,
      "mean_token_accuracy": 0.2365269511938095,
      "num_tokens": 99018072.0,
      "step": 44360
    },
    {
      "entropy": 5.320294761657715,
      "epoch": 4.982647273544112,
      "grad_norm": 1.0,
      "learning_rate": 0.0002802268929758475,
      "loss": 4.9299,
      "mean_token_accuracy": 0.2276756688952446,
      "num_tokens": 99030295.0,
      "step": 44365
    },
    {
      "entropy": 5.360086059570312,
      "epoch": 4.983208850452069,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002801867550086674,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2243215411901474,
      "num_tokens": 99042456.0,
      "step": 44370
    },
    {
      "entropy": 5.33536581993103,
      "epoch": 4.983770427360027,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002801466168763389,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2301416203379631,
      "num_tokens": 99053639.0,
      "step": 44375
    },
    {
      "entropy": 5.400546598434448,
      "epoch": 4.984332004267984,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002801064785801401,
      "loss": 4.8409,
      "mean_token_accuracy": 0.2357956662774086,
      "num_tokens": 99065350.0,
      "step": 44380
    },
    {
      "entropy": 5.402433681488037,
      "epoch": 4.9848935811759425,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028006634012134915,
      "loss": 4.94,
      "mean_token_accuracy": 0.2324439376592636,
      "num_tokens": 99075675.0,
      "step": 44385
    },
    {
      "entropy": 5.313912677764892,
      "epoch": 4.9854551580839,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028002620150124386,
      "loss": 4.876,
      "mean_token_accuracy": 0.2301669090986252,
      "num_tokens": 99086431.0,
      "step": 44390
    },
    {
      "entropy": 5.293459939956665,
      "epoch": 4.986016734991857,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00027998606272110237,
      "loss": 4.8,
      "mean_token_accuracy": 0.2427807107567787,
      "num_tokens": 99097417.0,
      "step": 44395
    },
    {
      "entropy": 5.2105019092559814,
      "epoch": 4.986578311899815,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002799459237822028,
      "loss": 4.7429,
      "mean_token_accuracy": 0.23698496222496032,
      "num_tokens": 99107762.0,
      "step": 44400
    },
    {
      "entropy": 5.342716836929322,
      "epoch": 4.987139888807772,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027990578468582304,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22431514859199525,
      "num_tokens": 99118566.0,
      "step": 44405
    },
    {
      "entropy": 5.350826692581177,
      "epoch": 4.9877014657157295,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002798656454332412,
      "loss": 4.859,
      "mean_token_accuracy": 0.23182294368743897,
      "num_tokens": 99129327.0,
      "step": 44410
    },
    {
      "entropy": 5.4494575500488285,
      "epoch": 4.988263042623688,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027982550602573524,
      "loss": 4.9703,
      "mean_token_accuracy": 0.22818288803100586,
      "num_tokens": 99140811.0,
      "step": 44415
    },
    {
      "entropy": 5.337866306304932,
      "epoch": 4.988824619531645,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002797853664645834,
      "loss": 4.8725,
      "mean_token_accuracy": 0.23053045868873595,
      "num_tokens": 99151860.0,
      "step": 44420
    },
    {
      "entropy": 5.326477575302124,
      "epoch": 4.989386196439602,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002797452267510637,
      "loss": 4.8413,
      "mean_token_accuracy": 0.2286055341362953,
      "num_tokens": 99163235.0,
      "step": 44425
    },
    {
      "entropy": 5.316936683654785,
      "epoch": 4.98994777334756,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000279705086886454,
      "loss": 4.8397,
      "mean_token_accuracy": 0.23527706414461136,
      "num_tokens": 99176239.0,
      "step": 44430
    },
    {
      "entropy": 5.2372802734375,
      "epoch": 4.990509350255517,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002796649468720326,
      "loss": 4.7824,
      "mean_token_accuracy": 0.23237699866294861,
      "num_tokens": 99187221.0,
      "step": 44435
    },
    {
      "entropy": 5.299254608154297,
      "epoch": 4.9910709271634754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027962480670907743,
      "loss": 4.8679,
      "mean_token_accuracy": 0.23690046817064286,
      "num_tokens": 99198803.0,
      "step": 44440
    },
    {
      "entropy": 5.348174715042115,
      "epoch": 4.991632504071433,
      "grad_norm": 0.953125,
      "learning_rate": 0.00027958466639886663,
      "loss": 4.891,
      "mean_token_accuracy": 0.227015221118927,
      "num_tokens": 99210179.0,
      "step": 44445
    },
    {
      "entropy": 5.455660724639893,
      "epoch": 4.99219408097939,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027954452594267825,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21731373220682143,
      "num_tokens": 99221633.0,
      "step": 44450
    },
    {
      "entropy": 5.316402816772461,
      "epoch": 4.992755657887348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027950438534179046,
      "loss": 4.8297,
      "mean_token_accuracy": 0.2369336426258087,
      "num_tokens": 99232292.0,
      "step": 44455
    },
    {
      "entropy": 5.253886413574219,
      "epoch": 4.993317234795305,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00027946424459748126,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23046937584877014,
      "num_tokens": 99244577.0,
      "step": 44460
    },
    {
      "entropy": 5.407855701446533,
      "epoch": 4.993878811703262,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027942410371102877,
      "loss": 4.9669,
      "mean_token_accuracy": 0.22711817175149918,
      "num_tokens": 99254698.0,
      "step": 44465
    },
    {
      "entropy": 5.362670612335205,
      "epoch": 4.9944403886112205,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002793839626837111,
      "loss": 4.944,
      "mean_token_accuracy": 0.23160551637411117,
      "num_tokens": 99266289.0,
      "step": 44470
    },
    {
      "entropy": 5.358116388320923,
      "epoch": 4.995001965519178,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027934382151680644,
      "loss": 4.885,
      "mean_token_accuracy": 0.23612411171197892,
      "num_tokens": 99278493.0,
      "step": 44475
    },
    {
      "entropy": 5.265798139572143,
      "epoch": 4.995563542427135,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027930368021159273,
      "loss": 4.8235,
      "mean_token_accuracy": 0.23651092797517775,
      "num_tokens": 99289095.0,
      "step": 44480
    },
    {
      "entropy": 5.337602138519287,
      "epoch": 4.996125119335093,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002792635387693481,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23498842418193816,
      "num_tokens": 99300314.0,
      "step": 44485
    },
    {
      "entropy": 5.3936927318573,
      "epoch": 4.99668669624305,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027922339719135093,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22946917563676833,
      "num_tokens": 99311460.0,
      "step": 44490
    },
    {
      "entropy": 5.363899755477905,
      "epoch": 4.997248273151008,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000279183255478879,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22423994541168213,
      "num_tokens": 99323755.0,
      "step": 44495
    },
    {
      "entropy": 5.392146587371826,
      "epoch": 4.997809850058966,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00027914311363321073,
      "loss": 4.9697,
      "mean_token_accuracy": 0.221455679833889,
      "num_tokens": 99335079.0,
      "step": 44500
    },
    {
      "entropy": 5.376827335357666,
      "epoch": 4.998371426966923,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027910297165562406,
      "loss": 4.9296,
      "mean_token_accuracy": 0.2300969272851944,
      "num_tokens": 99347051.0,
      "step": 44505
    },
    {
      "entropy": 5.305168724060058,
      "epoch": 4.998933003874881,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027906282954739714,
      "loss": 4.8362,
      "mean_token_accuracy": 0.23675844669342042,
      "num_tokens": 99357038.0,
      "step": 44510
    },
    {
      "entropy": 5.330075120925903,
      "epoch": 4.999494580782838,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00027902268730980816,
      "loss": 4.8395,
      "mean_token_accuracy": 0.23686676174402238,
      "num_tokens": 99368909.0,
      "step": 44515
    },
    {
      "entropy": 5.3331469429863825,
      "epoch": 5.0,
      "grad_norm": 3.390625,
      "learning_rate": 0.0002789825449441353,
      "loss": 4.8789,
      "mean_token_accuracy": 0.2276321682665083,
      "num_tokens": 99377525.0,
      "step": 44520
    },
    {
      "entropy": 5.303949975967408,
      "epoch": 5.000561576907957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002789424024516566,
      "loss": 4.7182,
      "mean_token_accuracy": 0.24421388506889344,
      "num_tokens": 99388238.0,
      "step": 44525
    },
    {
      "entropy": 5.287517738342285,
      "epoch": 5.001123153815915,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027890225983365034,
      "loss": 4.8203,
      "mean_token_accuracy": 0.23427336513996125,
      "num_tokens": 99399823.0,
      "step": 44530
    },
    {
      "entropy": 5.252340030670166,
      "epoch": 5.0016847307238725,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002788621170913946,
      "loss": 4.7802,
      "mean_token_accuracy": 0.23298331648111342,
      "num_tokens": 99411593.0,
      "step": 44535
    },
    {
      "entropy": 5.406036996841431,
      "epoch": 5.00224630763183,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027882197422616746,
      "loss": 4.878,
      "mean_token_accuracy": 0.23435810953378677,
      "num_tokens": 99423185.0,
      "step": 44540
    },
    {
      "entropy": 5.35374174118042,
      "epoch": 5.002807884539788,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002787818312392473,
      "loss": 4.825,
      "mean_token_accuracy": 0.23475240170955658,
      "num_tokens": 99434306.0,
      "step": 44545
    },
    {
      "entropy": 5.303316450119018,
      "epoch": 5.003369461447745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027874168813191207,
      "loss": 4.8354,
      "mean_token_accuracy": 0.23187606483697892,
      "num_tokens": 99445328.0,
      "step": 44550
    },
    {
      "entropy": 5.356036424636841,
      "epoch": 5.003931038355703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002787015449054401,
      "loss": 4.8957,
      "mean_token_accuracy": 0.23820963948965074,
      "num_tokens": 99457101.0,
      "step": 44555
    },
    {
      "entropy": 5.3451367855072025,
      "epoch": 5.00449261526366,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002786614015611095,
      "loss": 4.7731,
      "mean_token_accuracy": 0.2389104649424553,
      "num_tokens": 99466862.0,
      "step": 44560
    },
    {
      "entropy": 5.287978935241699,
      "epoch": 5.005054192171618,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027862125810019844,
      "loss": 4.7714,
      "mean_token_accuracy": 0.23365640491247178,
      "num_tokens": 99477113.0,
      "step": 44565
    },
    {
      "entropy": 5.17185001373291,
      "epoch": 5.005615769079576,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027858111452398515,
      "loss": 4.6495,
      "mean_token_accuracy": 0.24367730617523192,
      "num_tokens": 99488308.0,
      "step": 44570
    },
    {
      "entropy": 5.345585441589355,
      "epoch": 5.006177345987533,
      "grad_norm": 1.0,
      "learning_rate": 0.00027854097083374773,
      "loss": 4.8666,
      "mean_token_accuracy": 0.23174157440662385,
      "num_tokens": 99499082.0,
      "step": 44575
    },
    {
      "entropy": 5.333270645141601,
      "epoch": 5.00673892289549,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002785008270307645,
      "loss": 4.7902,
      "mean_token_accuracy": 0.2316268801689148,
      "num_tokens": 99509703.0,
      "step": 44580
    },
    {
      "entropy": 5.256840753555298,
      "epoch": 5.007300499803448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002784606831163136,
      "loss": 4.7068,
      "mean_token_accuracy": 0.24941868036985398,
      "num_tokens": 99519893.0,
      "step": 44585
    },
    {
      "entropy": 5.306329822540283,
      "epoch": 5.0078620767114055,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027842053909167314,
      "loss": 4.8003,
      "mean_token_accuracy": 0.23638997972011566,
      "num_tokens": 99530795.0,
      "step": 44590
    },
    {
      "entropy": 5.303034067153931,
      "epoch": 5.008423653619364,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027838039495812133,
      "loss": 4.7357,
      "mean_token_accuracy": 0.24044784009456635,
      "num_tokens": 99540733.0,
      "step": 44595
    },
    {
      "entropy": 5.315994691848755,
      "epoch": 5.008985230527321,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027834025071693655,
      "loss": 4.7891,
      "mean_token_accuracy": 0.23771920204162597,
      "num_tokens": 99551671.0,
      "step": 44600
    },
    {
      "entropy": 5.291541910171508,
      "epoch": 5.009546807435278,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027830010636939684,
      "loss": 4.7813,
      "mean_token_accuracy": 0.2412249580025673,
      "num_tokens": 99561894.0,
      "step": 44605
    },
    {
      "entropy": 5.231776475906372,
      "epoch": 5.010108384343236,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027825996191678043,
      "loss": 4.6943,
      "mean_token_accuracy": 0.2490280330181122,
      "num_tokens": 99573309.0,
      "step": 44610
    },
    {
      "entropy": 5.292166233062744,
      "epoch": 5.010669961251193,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027821981736036555,
      "loss": 4.7933,
      "mean_token_accuracy": 0.23290181457996367,
      "num_tokens": 99584517.0,
      "step": 44615
    },
    {
      "entropy": 5.401762437820435,
      "epoch": 5.011231538159151,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002781796727014305,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2303318977355957,
      "num_tokens": 99596034.0,
      "step": 44620
    },
    {
      "entropy": 5.398155689239502,
      "epoch": 5.011793115067109,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002781395279412535,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22965058237314223,
      "num_tokens": 99608399.0,
      "step": 44625
    },
    {
      "entropy": 5.327514028549194,
      "epoch": 5.012354691975066,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00027809938308111254,
      "loss": 4.7469,
      "mean_token_accuracy": 0.23899840265512468,
      "num_tokens": 99620328.0,
      "step": 44630
    },
    {
      "entropy": 5.299961280822754,
      "epoch": 5.012916268883023,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027805923812228615,
      "loss": 4.8127,
      "mean_token_accuracy": 0.23348730355501174,
      "num_tokens": 99630527.0,
      "step": 44635
    },
    {
      "entropy": 5.349582624435425,
      "epoch": 5.013477845790981,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027801909306605243,
      "loss": 4.8645,
      "mean_token_accuracy": 0.23481377512216567,
      "num_tokens": 99641192.0,
      "step": 44640
    },
    {
      "entropy": 5.28320837020874,
      "epoch": 5.014039422698938,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002779789479136896,
      "loss": 4.7569,
      "mean_token_accuracy": 0.23828925788402558,
      "num_tokens": 99653335.0,
      "step": 44645
    },
    {
      "entropy": 5.331968498229981,
      "epoch": 5.0146009996068965,
      "grad_norm": 1.125,
      "learning_rate": 0.0002779388026664759,
      "loss": 4.7688,
      "mean_token_accuracy": 0.2405068963766098,
      "num_tokens": 99663627.0,
      "step": 44650
    },
    {
      "entropy": 5.31217622756958,
      "epoch": 5.015162576514854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002778986573256896,
      "loss": 4.8086,
      "mean_token_accuracy": 0.23821263015270233,
      "num_tokens": 99673988.0,
      "step": 44655
    },
    {
      "entropy": 5.309736108779907,
      "epoch": 5.015724153422811,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002778585118926089,
      "loss": 4.8139,
      "mean_token_accuracy": 0.22898896187543868,
      "num_tokens": 99684150.0,
      "step": 44660
    },
    {
      "entropy": 5.3033661365509035,
      "epoch": 5.016285730330769,
      "grad_norm": 0.953125,
      "learning_rate": 0.00027781836636851207,
      "loss": 4.7693,
      "mean_token_accuracy": 0.24040007144212722,
      "num_tokens": 99696144.0,
      "step": 44665
    },
    {
      "entropy": 5.349575757980347,
      "epoch": 5.016847307238726,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027777822075467735,
      "loss": 4.7768,
      "mean_token_accuracy": 0.23750172406435013,
      "num_tokens": 99707405.0,
      "step": 44670
    },
    {
      "entropy": 5.283712482452392,
      "epoch": 5.0174088841466835,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000277738075052383,
      "loss": 4.6969,
      "mean_token_accuracy": 0.25905880331993103,
      "num_tokens": 99720316.0,
      "step": 44675
    },
    {
      "entropy": 5.235405826568604,
      "epoch": 5.017970461054642,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002776979292629074,
      "loss": 4.6917,
      "mean_token_accuracy": 0.2434634044766426,
      "num_tokens": 99732184.0,
      "step": 44680
    },
    {
      "entropy": 5.248032140731811,
      "epoch": 5.018532037962599,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002776577833875285,
      "loss": 4.7723,
      "mean_token_accuracy": 0.23761207461357117,
      "num_tokens": 99743349.0,
      "step": 44685
    },
    {
      "entropy": 5.330416345596314,
      "epoch": 5.019093614870557,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027761763742752485,
      "loss": 4.8311,
      "mean_token_accuracy": 0.23845785707235337,
      "num_tokens": 99754569.0,
      "step": 44690
    },
    {
      "entropy": 5.406059694290161,
      "epoch": 5.019655191778514,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002775774913841746,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21944546699523926,
      "num_tokens": 99766643.0,
      "step": 44695
    },
    {
      "entropy": 5.247896718978882,
      "epoch": 5.020216768686471,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000277537345258756,
      "loss": 4.6613,
      "mean_token_accuracy": 0.24383610635995864,
      "num_tokens": 99777829.0,
      "step": 44700
    },
    {
      "entropy": 5.284526586532593,
      "epoch": 5.0207783455944295,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027749719905254737,
      "loss": 4.7613,
      "mean_token_accuracy": 0.2390891820192337,
      "num_tokens": 99788815.0,
      "step": 44705
    },
    {
      "entropy": 5.2233256816864015,
      "epoch": 5.021339922502387,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000277457052766827,
      "loss": 4.7915,
      "mean_token_accuracy": 0.24463313072919846,
      "num_tokens": 99799343.0,
      "step": 44710
    },
    {
      "entropy": 5.4329140186309814,
      "epoch": 5.021901499410344,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027741690640287306,
      "loss": 4.9657,
      "mean_token_accuracy": 0.21930896639823913,
      "num_tokens": 99810926.0,
      "step": 44715
    },
    {
      "entropy": 5.336809492111206,
      "epoch": 5.022463076318302,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002773767599619639,
      "loss": 4.8134,
      "mean_token_accuracy": 0.24083648920059203,
      "num_tokens": 99824506.0,
      "step": 44720
    },
    {
      "entropy": 5.40251054763794,
      "epoch": 5.023024653226259,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027733661344537776,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2357860177755356,
      "num_tokens": 99835751.0,
      "step": 44725
    },
    {
      "entropy": 5.276918649673462,
      "epoch": 5.0235862301342165,
      "grad_norm": 1.09375,
      "learning_rate": 0.000277296466854393,
      "loss": 4.6888,
      "mean_token_accuracy": 0.2441596731543541,
      "num_tokens": 99846793.0,
      "step": 44730
    },
    {
      "entropy": 5.2190086364746096,
      "epoch": 5.024147807042175,
      "grad_norm": 1.0,
      "learning_rate": 0.0002772563201902877,
      "loss": 4.6778,
      "mean_token_accuracy": 0.24783345311880112,
      "num_tokens": 99858881.0,
      "step": 44735
    },
    {
      "entropy": 5.316756534576416,
      "epoch": 5.024709383950132,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027721617345434035,
      "loss": 4.8346,
      "mean_token_accuracy": 0.23651431500911713,
      "num_tokens": 99869902.0,
      "step": 44740
    },
    {
      "entropy": 5.274245309829712,
      "epoch": 5.02527096085809,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00027717602664782925,
      "loss": 4.7535,
      "mean_token_accuracy": 0.24220282286405564,
      "num_tokens": 99879899.0,
      "step": 44745
    },
    {
      "entropy": 5.217467880249023,
      "epoch": 5.025832537766047,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002771358797720326,
      "loss": 4.6689,
      "mean_token_accuracy": 0.24848663359880446,
      "num_tokens": 99891751.0,
      "step": 44750
    },
    {
      "entropy": 5.306231546401977,
      "epoch": 5.026394114674004,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027709573282822864,
      "loss": 4.7705,
      "mean_token_accuracy": 0.24466101080179214,
      "num_tokens": 99901660.0,
      "step": 44755
    },
    {
      "entropy": 5.328562498092651,
      "epoch": 5.026955691581962,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002770555858176958,
      "loss": 4.883,
      "mean_token_accuracy": 0.22767579257488252,
      "num_tokens": 99912319.0,
      "step": 44760
    },
    {
      "entropy": 5.298930549621582,
      "epoch": 5.02751726848992,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002770154387417123,
      "loss": 4.7799,
      "mean_token_accuracy": 0.24284012317657472,
      "num_tokens": 99922425.0,
      "step": 44765
    },
    {
      "entropy": 5.305669355392456,
      "epoch": 5.028078845397877,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002769752916015565,
      "loss": 4.7462,
      "mean_token_accuracy": 0.23931794613599777,
      "num_tokens": 99934210.0,
      "step": 44770
    },
    {
      "entropy": 5.305175018310547,
      "epoch": 5.028640422305835,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027693514439850646,
      "loss": 4.778,
      "mean_token_accuracy": 0.23506115823984147,
      "num_tokens": 99947094.0,
      "step": 44775
    },
    {
      "entropy": 5.265874767303467,
      "epoch": 5.029201999213792,
      "grad_norm": 1.125,
      "learning_rate": 0.00027689499713384085,
      "loss": 4.8028,
      "mean_token_accuracy": 0.23250661343336104,
      "num_tokens": 99957876.0,
      "step": 44780
    },
    {
      "entropy": 5.245990514755249,
      "epoch": 5.02976357612175,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002768548498088378,
      "loss": 4.8071,
      "mean_token_accuracy": 0.23538009822368622,
      "num_tokens": 99970033.0,
      "step": 44785
    },
    {
      "entropy": 5.311634159088134,
      "epoch": 5.0303251530297075,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002768147024247756,
      "loss": 4.8364,
      "mean_token_accuracy": 0.23247957676649095,
      "num_tokens": 99980708.0,
      "step": 44790
    },
    {
      "entropy": 5.328367805480957,
      "epoch": 5.030886729937665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002767745549829325,
      "loss": 4.7763,
      "mean_token_accuracy": 0.23279207795858384,
      "num_tokens": 99991504.0,
      "step": 44795
    },
    {
      "entropy": 5.359219264984131,
      "epoch": 5.031448306845623,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027673440748458687,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2294816792011261,
      "num_tokens": 100002722.0,
      "step": 44800
    },
    {
      "entropy": 5.313094615936279,
      "epoch": 5.03200988375358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027669425993101716,
      "loss": 4.8193,
      "mean_token_accuracy": 0.23241813331842423,
      "num_tokens": 100014337.0,
      "step": 44805
    },
    {
      "entropy": 5.269993638992309,
      "epoch": 5.032571460661537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002766541123235014,
      "loss": 4.7312,
      "mean_token_accuracy": 0.23760635107755662,
      "num_tokens": 100025717.0,
      "step": 44810
    },
    {
      "entropy": 5.3128660202026365,
      "epoch": 5.033133037569495,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002766139646633182,
      "loss": 4.7666,
      "mean_token_accuracy": 0.23344726413488387,
      "num_tokens": 100036882.0,
      "step": 44815
    },
    {
      "entropy": 5.255267572402954,
      "epoch": 5.033694614477453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027657381695174563,
      "loss": 4.7346,
      "mean_token_accuracy": 0.2389051392674446,
      "num_tokens": 100047744.0,
      "step": 44820
    },
    {
      "entropy": 5.397951078414917,
      "epoch": 5.03425619138541,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027653366919006217,
      "loss": 4.9018,
      "mean_token_accuracy": 0.23275411128997803,
      "num_tokens": 100059877.0,
      "step": 44825
    },
    {
      "entropy": 5.279456710815429,
      "epoch": 5.034817768293368,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002764935213795461,
      "loss": 4.7415,
      "mean_token_accuracy": 0.2421242907643318,
      "num_tokens": 100072294.0,
      "step": 44830
    },
    {
      "entropy": 5.293368148803711,
      "epoch": 5.035379345201325,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002764533735214757,
      "loss": 4.7823,
      "mean_token_accuracy": 0.2416826829314232,
      "num_tokens": 100084302.0,
      "step": 44835
    },
    {
      "entropy": 5.3939775943756105,
      "epoch": 5.035940922109283,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027641322561712935,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2360362932085991,
      "num_tokens": 100094694.0,
      "step": 44840
    },
    {
      "entropy": 5.278969097137451,
      "epoch": 5.0365024990172405,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027637307766778524,
      "loss": 4.7604,
      "mean_token_accuracy": 0.2422454759478569,
      "num_tokens": 100105824.0,
      "step": 44845
    },
    {
      "entropy": 5.324724388122559,
      "epoch": 5.037064075925198,
      "grad_norm": 1.109375,
      "learning_rate": 0.000276332929674722,
      "loss": 4.8436,
      "mean_token_accuracy": 0.23639057129621505,
      "num_tokens": 100117040.0,
      "step": 44850
    },
    {
      "entropy": 5.316409301757813,
      "epoch": 5.037625652833156,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027629278163921763,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23150028735399247,
      "num_tokens": 100128747.0,
      "step": 44855
    },
    {
      "entropy": 5.293092107772827,
      "epoch": 5.038187229741113,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027625263356255067,
      "loss": 4.8037,
      "mean_token_accuracy": 0.23068335354328157,
      "num_tokens": 100138839.0,
      "step": 44860
    },
    {
      "entropy": 5.181653308868408,
      "epoch": 5.03874880664907,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002762124854459992,
      "loss": 4.6746,
      "mean_token_accuracy": 0.24150864481925965,
      "num_tokens": 100148898.0,
      "step": 44865
    },
    {
      "entropy": 5.2901581764221195,
      "epoch": 5.039310383557028,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027617233729084194,
      "loss": 4.8236,
      "mean_token_accuracy": 0.23563172072172164,
      "num_tokens": 100160160.0,
      "step": 44870
    },
    {
      "entropy": 5.263094234466553,
      "epoch": 5.0398719604649855,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027613218909835693,
      "loss": 4.7399,
      "mean_token_accuracy": 0.2462816432118416,
      "num_tokens": 100170556.0,
      "step": 44875
    },
    {
      "entropy": 5.337083625793457,
      "epoch": 5.040433537372944,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002760920408698225,
      "loss": 4.7939,
      "mean_token_accuracy": 0.2324121907353401,
      "num_tokens": 100181534.0,
      "step": 44880
    },
    {
      "entropy": 5.27929105758667,
      "epoch": 5.040995114280901,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00027605189260651716,
      "loss": 4.8598,
      "mean_token_accuracy": 0.23064640313386917,
      "num_tokens": 100193086.0,
      "step": 44885
    },
    {
      "entropy": 5.389669275283813,
      "epoch": 5.041556691188858,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002760117443097191,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23510549515485762,
      "num_tokens": 100204222.0,
      "step": 44890
    },
    {
      "entropy": 5.269767236709595,
      "epoch": 5.042118268096816,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027597159598070675,
      "loss": 4.7544,
      "mean_token_accuracy": 0.24486927837133407,
      "num_tokens": 100215260.0,
      "step": 44895
    },
    {
      "entropy": 5.2542260646820065,
      "epoch": 5.042679845004773,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027593144762075837,
      "loss": 4.7413,
      "mean_token_accuracy": 0.2402457669377327,
      "num_tokens": 100225803.0,
      "step": 44900
    },
    {
      "entropy": 5.31165714263916,
      "epoch": 5.043241421912731,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027589129923115234,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2288212835788727,
      "num_tokens": 100237351.0,
      "step": 44905
    },
    {
      "entropy": 5.292793464660645,
      "epoch": 5.043802998820689,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000275851150813167,
      "loss": 4.7729,
      "mean_token_accuracy": 0.2498647913336754,
      "num_tokens": 100247484.0,
      "step": 44910
    },
    {
      "entropy": 5.327408313751221,
      "epoch": 5.044364575728646,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002758110023680807,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23849890530109405,
      "num_tokens": 100259159.0,
      "step": 44915
    },
    {
      "entropy": 5.371684217453003,
      "epoch": 5.044926152636603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002757708538971717,
      "loss": 4.9348,
      "mean_token_accuracy": 0.22241273075342177,
      "num_tokens": 100270673.0,
      "step": 44920
    },
    {
      "entropy": 5.304699563980103,
      "epoch": 5.045487729544561,
      "grad_norm": 1.0,
      "learning_rate": 0.0002757307054017185,
      "loss": 4.8079,
      "mean_token_accuracy": 0.22975546568632127,
      "num_tokens": 100281825.0,
      "step": 44925
    },
    {
      "entropy": 5.3554059028625485,
      "epoch": 5.0460493064525185,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002756905568829994,
      "loss": 4.7304,
      "mean_token_accuracy": 0.23920682817697525,
      "num_tokens": 100292610.0,
      "step": 44930
    },
    {
      "entropy": 5.337239980697632,
      "epoch": 5.046610883360477,
      "grad_norm": 1.0,
      "learning_rate": 0.00027565040834229266,
      "loss": 4.8027,
      "mean_token_accuracy": 0.23677806705236434,
      "num_tokens": 100303589.0,
      "step": 44935
    },
    {
      "entropy": 5.252290534973144,
      "epoch": 5.047172460268434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027561025978087663,
      "loss": 4.7515,
      "mean_token_accuracy": 0.24129058569669723,
      "num_tokens": 100314843.0,
      "step": 44940
    },
    {
      "entropy": 5.289244747161865,
      "epoch": 5.047734037176391,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027557011120002975,
      "loss": 4.7892,
      "mean_token_accuracy": 0.23798831850290297,
      "num_tokens": 100325250.0,
      "step": 44945
    },
    {
      "entropy": 5.34537706375122,
      "epoch": 5.048295614084349,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002755299626010302,
      "loss": 4.8331,
      "mean_token_accuracy": 0.23407106399536132,
      "num_tokens": 100336280.0,
      "step": 44950
    },
    {
      "entropy": 5.2791344165802006,
      "epoch": 5.048857190992306,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002754898139851565,
      "loss": 4.8329,
      "mean_token_accuracy": 0.2336358904838562,
      "num_tokens": 100349040.0,
      "step": 44955
    },
    {
      "entropy": 5.37589168548584,
      "epoch": 5.049418767900264,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027544966535368695,
      "loss": 4.8912,
      "mean_token_accuracy": 0.23421404510736465,
      "num_tokens": 100360602.0,
      "step": 44960
    },
    {
      "entropy": 5.382865285873413,
      "epoch": 5.049980344808222,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002754095167078999,
      "loss": 4.9194,
      "mean_token_accuracy": 0.22897761613130568,
      "num_tokens": 100371627.0,
      "step": 44965
    },
    {
      "entropy": 5.357540035247803,
      "epoch": 5.050541921716179,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002753693680490736,
      "loss": 4.8615,
      "mean_token_accuracy": 0.2335910454392433,
      "num_tokens": 100382897.0,
      "step": 44970
    },
    {
      "entropy": 5.2880110263824465,
      "epoch": 5.051103498624137,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002753292193784865,
      "loss": 4.7454,
      "mean_token_accuracy": 0.24296204298734664,
      "num_tokens": 100394764.0,
      "step": 44975
    },
    {
      "entropy": 5.34711012840271,
      "epoch": 5.051665075532094,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027528907069741696,
      "loss": 4.8901,
      "mean_token_accuracy": 0.22749956250190734,
      "num_tokens": 100406603.0,
      "step": 44980
    },
    {
      "entropy": 5.166896104812622,
      "epoch": 5.052226652440051,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002752489220071433,
      "loss": 4.5887,
      "mean_token_accuracy": 0.25253511369228365,
      "num_tokens": 100417050.0,
      "step": 44985
    },
    {
      "entropy": 5.335669803619385,
      "epoch": 5.0527882293480095,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002752087733089438,
      "loss": 4.8025,
      "mean_token_accuracy": 0.23743102997541427,
      "num_tokens": 100428344.0,
      "step": 44990
    },
    {
      "entropy": 5.29627799987793,
      "epoch": 5.053349806255967,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002751686246040969,
      "loss": 4.7213,
      "mean_token_accuracy": 0.24333713203668594,
      "num_tokens": 100439887.0,
      "step": 44995
    },
    {
      "entropy": 5.316780090332031,
      "epoch": 5.053911383163924,
      "grad_norm": 1.09375,
      "learning_rate": 0.000275128475893881,
      "loss": 4.8223,
      "mean_token_accuracy": 0.2441769376397133,
      "num_tokens": 100450297.0,
      "step": 45000
    },
    {
      "epoch": 5.053911383163924,
      "eval_entropy": 5.091480728889329,
      "eval_loss": 5.0214948654174805,
      "eval_mean_token_accuracy": 0.23153242991546827,
      "eval_num_tokens": 100450297.0,
      "eval_runtime": 23.9715,
      "eval_samples_per_second": 1293.744,
      "eval_steps_per_second": 161.734,
      "step": 45000
    },
    {
      "entropy": 5.425415658950806,
      "epoch": 5.054472960071882,
      "grad_norm": 1.0,
      "learning_rate": 0.00027508832717957433,
      "loss": 5.0265,
      "mean_token_accuracy": 0.22369078546762466,
      "num_tokens": 100463165.0,
      "step": 45005
    },
    {
      "entropy": 5.381993341445923,
      "epoch": 5.055034536979839,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002750481784624553,
      "loss": 4.8213,
      "mean_token_accuracy": 0.245214082300663,
      "num_tokens": 100474737.0,
      "step": 45010
    },
    {
      "entropy": 5.351296234130859,
      "epoch": 5.0555961138877965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027500802974380224,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22939803898334504,
      "num_tokens": 100487235.0,
      "step": 45015
    },
    {
      "entropy": 5.360181474685669,
      "epoch": 5.056157690795755,
      "grad_norm": 1.0,
      "learning_rate": 0.0002749678810248935,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2375738024711609,
      "num_tokens": 100498886.0,
      "step": 45020
    },
    {
      "entropy": 5.314907264709473,
      "epoch": 5.056719267703712,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002749277323070075,
      "loss": 4.7459,
      "mean_token_accuracy": 0.2407187655568123,
      "num_tokens": 100509748.0,
      "step": 45025
    },
    {
      "entropy": 5.40738320350647,
      "epoch": 5.05728084461167,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027488758359142245,
      "loss": 4.8677,
      "mean_token_accuracy": 0.23281890153884888,
      "num_tokens": 100522189.0,
      "step": 45030
    },
    {
      "entropy": 5.227555274963379,
      "epoch": 5.057842421519627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002748474348794167,
      "loss": 4.7297,
      "mean_token_accuracy": 0.2392197370529175,
      "num_tokens": 100533437.0,
      "step": 45035
    },
    {
      "entropy": 5.451576375961304,
      "epoch": 5.058403998427584,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027480728617226884,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2280246540904045,
      "num_tokens": 100545268.0,
      "step": 45040
    },
    {
      "entropy": 5.31919264793396,
      "epoch": 5.0589655753355425,
      "grad_norm": 1.15625,
      "learning_rate": 0.000274767137471257,
      "loss": 4.8211,
      "mean_token_accuracy": 0.23375478088855745,
      "num_tokens": 100555650.0,
      "step": 45045
    },
    {
      "entropy": 5.294924592971801,
      "epoch": 5.0595271522435,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027472698877765967,
      "loss": 4.8222,
      "mean_token_accuracy": 0.23947843611240388,
      "num_tokens": 100567173.0,
      "step": 45050
    },
    {
      "entropy": 5.408014917373658,
      "epoch": 5.060088729151457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002746868400927551,
      "loss": 4.8527,
      "mean_token_accuracy": 0.2306646853685379,
      "num_tokens": 100577091.0,
      "step": 45055
    },
    {
      "entropy": 5.40376467704773,
      "epoch": 5.060650306059415,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002746466914178216,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2394551992416382,
      "num_tokens": 100587917.0,
      "step": 45060
    },
    {
      "entropy": 5.210582208633423,
      "epoch": 5.061211882967372,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002746065427541376,
      "loss": 4.747,
      "mean_token_accuracy": 0.2391575112938881,
      "num_tokens": 100598861.0,
      "step": 45065
    },
    {
      "entropy": 5.275238084793091,
      "epoch": 5.06177345987533,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00027456639410298144,
      "loss": 4.7445,
      "mean_token_accuracy": 0.24094553142786027,
      "num_tokens": 100609293.0,
      "step": 45070
    },
    {
      "entropy": 5.369404888153076,
      "epoch": 5.062335036783288,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002745262454656314,
      "loss": 4.8905,
      "mean_token_accuracy": 0.225148943066597,
      "num_tokens": 100620062.0,
      "step": 45075
    },
    {
      "entropy": 5.452035331726075,
      "epoch": 5.062896613691245,
      "grad_norm": 1.125,
      "learning_rate": 0.000274486096843366,
      "loss": 4.9365,
      "mean_token_accuracy": 0.2333208590745926,
      "num_tokens": 100630099.0,
      "step": 45080
    },
    {
      "entropy": 5.352411556243896,
      "epoch": 5.063458190599203,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027444594823746344,
      "loss": 4.7913,
      "mean_token_accuracy": 0.23640179336071016,
      "num_tokens": 100641710.0,
      "step": 45085
    },
    {
      "entropy": 5.249108123779297,
      "epoch": 5.06401976750716,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002744057996492021,
      "loss": 4.7095,
      "mean_token_accuracy": 0.24001881927251817,
      "num_tokens": 100652189.0,
      "step": 45090
    },
    {
      "entropy": 5.311280059814453,
      "epoch": 5.064581344415117,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002743656510798603,
      "loss": 4.8105,
      "mean_token_accuracy": 0.23841976076364518,
      "num_tokens": 100663982.0,
      "step": 45095
    },
    {
      "entropy": 5.342454147338867,
      "epoch": 5.065142921323075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002743255025307165,
      "loss": 4.8284,
      "mean_token_accuracy": 0.23870106190443038,
      "num_tokens": 100675070.0,
      "step": 45100
    },
    {
      "entropy": 5.439078283309937,
      "epoch": 5.065704498231033,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027428535400304886,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2298881307244301,
      "num_tokens": 100686852.0,
      "step": 45105
    },
    {
      "entropy": 5.305760908126831,
      "epoch": 5.06626607513899,
      "grad_norm": 1.0,
      "learning_rate": 0.0002742452054981358,
      "loss": 4.7961,
      "mean_token_accuracy": 0.23273055553436278,
      "num_tokens": 100697397.0,
      "step": 45110
    },
    {
      "entropy": 5.276540327072143,
      "epoch": 5.066827652046948,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002742050570172558,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23807377219200135,
      "num_tokens": 100709543.0,
      "step": 45115
    },
    {
      "entropy": 5.322295188903809,
      "epoch": 5.067389228954905,
      "grad_norm": 1.0625,
      "learning_rate": 0.000274164908561687,
      "loss": 4.8779,
      "mean_token_accuracy": 0.2331213504076004,
      "num_tokens": 100721742.0,
      "step": 45120
    },
    {
      "entropy": 5.401753330230713,
      "epoch": 5.067950805862863,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027412476013270786,
      "loss": 4.9792,
      "mean_token_accuracy": 0.22605310380458832,
      "num_tokens": 100733954.0,
      "step": 45125
    },
    {
      "entropy": 5.42852463722229,
      "epoch": 5.0685123827708205,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027408461173159676,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22979721873998643,
      "num_tokens": 100745855.0,
      "step": 45130
    },
    {
      "entropy": 5.240601062774658,
      "epoch": 5.069073959678778,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00027404446335963195,
      "loss": 4.6975,
      "mean_token_accuracy": 0.23887763917446136,
      "num_tokens": 100756673.0,
      "step": 45135
    },
    {
      "entropy": 5.227750205993653,
      "epoch": 5.069635536586736,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002740043150180918,
      "loss": 4.7575,
      "mean_token_accuracy": 0.2381819099187851,
      "num_tokens": 100768088.0,
      "step": 45140
    },
    {
      "entropy": 5.233298015594483,
      "epoch": 5.070197113494693,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002739641667082546,
      "loss": 4.7746,
      "mean_token_accuracy": 0.240044566988945,
      "num_tokens": 100779275.0,
      "step": 45145
    },
    {
      "entropy": 5.379465484619141,
      "epoch": 5.07075869040265,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027392401843139876,
      "loss": 4.8124,
      "mean_token_accuracy": 0.2395699918270111,
      "num_tokens": 100789614.0,
      "step": 45150
    },
    {
      "entropy": 5.2215508937835695,
      "epoch": 5.071320267310608,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002738838701888026,
      "loss": 4.6521,
      "mean_token_accuracy": 0.24645081907510757,
      "num_tokens": 100800784.0,
      "step": 45155
    },
    {
      "entropy": 5.2364600658416744,
      "epoch": 5.071881844218566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002738437219817444,
      "loss": 4.7617,
      "mean_token_accuracy": 0.2370292827486992,
      "num_tokens": 100811280.0,
      "step": 45160
    },
    {
      "entropy": 5.322388648986816,
      "epoch": 5.072443421126524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027380357381150263,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22986701279878616,
      "num_tokens": 100821214.0,
      "step": 45165
    },
    {
      "entropy": 5.3450428485870365,
      "epoch": 5.073004998034481,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002737634256793555,
      "loss": 4.8121,
      "mean_token_accuracy": 0.23324092626571655,
      "num_tokens": 100834363.0,
      "step": 45170
    },
    {
      "entropy": 5.3375711917877195,
      "epoch": 5.073566574942438,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027372327758658135,
      "loss": 4.7909,
      "mean_token_accuracy": 0.2373301699757576,
      "num_tokens": 100845515.0,
      "step": 45175
    },
    {
      "entropy": 5.2282956600189205,
      "epoch": 5.074128151850396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027368312953445845,
      "loss": 4.7763,
      "mean_token_accuracy": 0.23462757617235183,
      "num_tokens": 100856714.0,
      "step": 45180
    },
    {
      "entropy": 5.248512697219849,
      "epoch": 5.0746897287583534,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002736429815242653,
      "loss": 4.7445,
      "mean_token_accuracy": 0.24191382825374602,
      "num_tokens": 100867828.0,
      "step": 45185
    },
    {
      "entropy": 5.363536930084228,
      "epoch": 5.075251305666311,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002736028335572802,
      "loss": 4.863,
      "mean_token_accuracy": 0.2341613292694092,
      "num_tokens": 100879589.0,
      "step": 45190
    },
    {
      "entropy": 5.3106166362762455,
      "epoch": 5.075812882574269,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00027356268563478135,
      "loss": 4.7612,
      "mean_token_accuracy": 0.23888947069644928,
      "num_tokens": 100889218.0,
      "step": 45195
    },
    {
      "entropy": 5.297762489318847,
      "epoch": 5.076374459482226,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002735225377580472,
      "loss": 4.758,
      "mean_token_accuracy": 0.2479991853237152,
      "num_tokens": 100899482.0,
      "step": 45200
    },
    {
      "entropy": 5.266585016250611,
      "epoch": 5.076936036390183,
      "grad_norm": 1.015625,
      "learning_rate": 0.000273482389928356,
      "loss": 4.7287,
      "mean_token_accuracy": 0.23895819932222367,
      "num_tokens": 100910752.0,
      "step": 45205
    },
    {
      "entropy": 5.315157270431518,
      "epoch": 5.077497613298141,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002734422421469861,
      "loss": 4.8179,
      "mean_token_accuracy": 0.2356388345360756,
      "num_tokens": 100922491.0,
      "step": 45210
    },
    {
      "entropy": 5.284455251693726,
      "epoch": 5.0780591902060985,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002734020944152157,
      "loss": 4.7623,
      "mean_token_accuracy": 0.24758197069168092,
      "num_tokens": 100933558.0,
      "step": 45215
    },
    {
      "entropy": 5.211187505722046,
      "epoch": 5.078620767114057,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002733619467343234,
      "loss": 4.6727,
      "mean_token_accuracy": 0.24591450542211532,
      "num_tokens": 100944973.0,
      "step": 45220
    },
    {
      "entropy": 5.362029361724853,
      "epoch": 5.079182344022014,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002733217991055873,
      "loss": 4.8812,
      "mean_token_accuracy": 0.23375518023967742,
      "num_tokens": 100954673.0,
      "step": 45225
    },
    {
      "entropy": 5.317120742797852,
      "epoch": 5.079743920929971,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002732816515302858,
      "loss": 4.7394,
      "mean_token_accuracy": 0.24129631072282792,
      "num_tokens": 100965797.0,
      "step": 45230
    },
    {
      "entropy": 5.332632207870484,
      "epoch": 5.080305497837929,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002732415040096971,
      "loss": 4.8157,
      "mean_token_accuracy": 0.23491228222846985,
      "num_tokens": 100977277.0,
      "step": 45235
    },
    {
      "entropy": 5.3242823600769045,
      "epoch": 5.080867074745886,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027320135654509964,
      "loss": 4.7455,
      "mean_token_accuracy": 0.24510975778102875,
      "num_tokens": 100989147.0,
      "step": 45240
    },
    {
      "entropy": 5.359854364395142,
      "epoch": 5.081428651653844,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002731612091377717,
      "loss": 4.8723,
      "mean_token_accuracy": 0.23540589958429337,
      "num_tokens": 100998317.0,
      "step": 45245
    },
    {
      "entropy": 5.270200300216675,
      "epoch": 5.081990228561802,
      "grad_norm": 1.0,
      "learning_rate": 0.00027312106178899154,
      "loss": 4.7865,
      "mean_token_accuracy": 0.2325945407152176,
      "num_tokens": 101010268.0,
      "step": 45250
    },
    {
      "entropy": 5.3294202327728275,
      "epoch": 5.082551805469759,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002730809145000374,
      "loss": 4.8263,
      "mean_token_accuracy": 0.23664241135120392,
      "num_tokens": 101020344.0,
      "step": 45255
    },
    {
      "entropy": 5.244296646118164,
      "epoch": 5.083113382377717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002730407672721878,
      "loss": 4.6912,
      "mean_token_accuracy": 0.24902569204568864,
      "num_tokens": 101031559.0,
      "step": 45260
    },
    {
      "entropy": 5.283180141448975,
      "epoch": 5.083674959285674,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000273000620106721,
      "loss": 4.7446,
      "mean_token_accuracy": 0.23678601533174515,
      "num_tokens": 101043585.0,
      "step": 45265
    },
    {
      "entropy": 5.329788541793823,
      "epoch": 5.0842365361936315,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002729604730049151,
      "loss": 4.7251,
      "mean_token_accuracy": 0.24079617857933044,
      "num_tokens": 101053506.0,
      "step": 45270
    },
    {
      "entropy": 5.359857273101807,
      "epoch": 5.08479811310159,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002729203259680486,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2261781007051468,
      "num_tokens": 101065179.0,
      "step": 45275
    },
    {
      "entropy": 5.326670074462891,
      "epoch": 5.085359690009547,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002728801789973998,
      "loss": 4.7783,
      "mean_token_accuracy": 0.23147362768650054,
      "num_tokens": 101075843.0,
      "step": 45280
    },
    {
      "entropy": 5.403021144866943,
      "epoch": 5.085921266917504,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002728400320942468,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22077289074659348,
      "num_tokens": 101087305.0,
      "step": 45285
    },
    {
      "entropy": 5.325952243804932,
      "epoch": 5.086482843825462,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000272799885259868,
      "loss": 4.8106,
      "mean_token_accuracy": 0.2429259479045868,
      "num_tokens": 101098738.0,
      "step": 45290
    },
    {
      "entropy": 5.3207625389099125,
      "epoch": 5.087044420733419,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002727597384955418,
      "loss": 4.7399,
      "mean_token_accuracy": 0.24181464910507203,
      "num_tokens": 101108476.0,
      "step": 45295
    },
    {
      "entropy": 5.228541851043701,
      "epoch": 5.0876059976413766,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027271959180254636,
      "loss": 4.7479,
      "mean_token_accuracy": 0.23916487395763397,
      "num_tokens": 101119547.0,
      "step": 45300
    },
    {
      "entropy": 5.29707407951355,
      "epoch": 5.088167574549335,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027267944518215997,
      "loss": 4.8339,
      "mean_token_accuracy": 0.24210966378450394,
      "num_tokens": 101130820.0,
      "step": 45305
    },
    {
      "entropy": 5.342971897125244,
      "epoch": 5.088729151457292,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027263929863566104,
      "loss": 4.8496,
      "mean_token_accuracy": 0.23438386470079423,
      "num_tokens": 101142565.0,
      "step": 45310
    },
    {
      "entropy": 5.379041337966919,
      "epoch": 5.08929072836525,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002725991521643277,
      "loss": 4.8549,
      "mean_token_accuracy": 0.2321729615330696,
      "num_tokens": 101152963.0,
      "step": 45315
    },
    {
      "entropy": 5.355437326431274,
      "epoch": 5.089852305273207,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002725590057694383,
      "loss": 4.8138,
      "mean_token_accuracy": 0.2283369317650795,
      "num_tokens": 101163932.0,
      "step": 45320
    },
    {
      "entropy": 5.22455677986145,
      "epoch": 5.090413882181164,
      "grad_norm": 0.984375,
      "learning_rate": 0.00027251885945227093,
      "loss": 4.7837,
      "mean_token_accuracy": 0.2390478640794754,
      "num_tokens": 101176333.0,
      "step": 45325
    },
    {
      "entropy": 5.294744396209717,
      "epoch": 5.0909754590891225,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00027247871321410425,
      "loss": 4.7887,
      "mean_token_accuracy": 0.235759137570858,
      "num_tokens": 101186443.0,
      "step": 45330
    },
    {
      "entropy": 5.294404792785644,
      "epoch": 5.09153703599708,
      "grad_norm": 1.0,
      "learning_rate": 0.00027243856705621623,
      "loss": 4.7801,
      "mean_token_accuracy": 0.24022141844034195,
      "num_tokens": 101197550.0,
      "step": 45335
    },
    {
      "entropy": 5.335787439346314,
      "epoch": 5.092098612905037,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027239842097988526,
      "loss": 4.8093,
      "mean_token_accuracy": 0.23716435879468917,
      "num_tokens": 101208609.0,
      "step": 45340
    },
    {
      "entropy": 5.297311878204345,
      "epoch": 5.092660189812995,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002723582749863896,
      "loss": 4.8344,
      "mean_token_accuracy": 0.23254527747631074,
      "num_tokens": 101219990.0,
      "step": 45345
    },
    {
      "entropy": 5.267515850067139,
      "epoch": 5.093221766720952,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027231812907700746,
      "loss": 4.7151,
      "mean_token_accuracy": 0.2401001051068306,
      "num_tokens": 101230700.0,
      "step": 45350
    },
    {
      "entropy": 5.252775001525879,
      "epoch": 5.09378334362891,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002722779832530171,
      "loss": 4.6843,
      "mean_token_accuracy": 0.24343296587467195,
      "num_tokens": 101242029.0,
      "step": 45355
    },
    {
      "entropy": 5.179193496704102,
      "epoch": 5.094344920536868,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027223783751569687,
      "loss": 4.6507,
      "mean_token_accuracy": 0.2411127954721451,
      "num_tokens": 101252649.0,
      "step": 45360
    },
    {
      "entropy": 5.389099788665772,
      "epoch": 5.094906497444825,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00027219769186632493,
      "loss": 5.0311,
      "mean_token_accuracy": 0.21986891329288483,
      "num_tokens": 101263744.0,
      "step": 45365
    },
    {
      "entropy": 5.392523527145386,
      "epoch": 5.095468074352783,
      "grad_norm": 1.0,
      "learning_rate": 0.0002721575463061796,
      "loss": 4.8545,
      "mean_token_accuracy": 0.23605295419692993,
      "num_tokens": 101276497.0,
      "step": 45370
    },
    {
      "entropy": 5.396787309646607,
      "epoch": 5.09602965126074,
      "grad_norm": 1.0,
      "learning_rate": 0.00027211740083653903,
      "loss": 4.8899,
      "mean_token_accuracy": 0.2293536588549614,
      "num_tokens": 101287759.0,
      "step": 45375
    },
    {
      "entropy": 5.404470157623291,
      "epoch": 5.096591228168697,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002720772554586816,
      "loss": 4.9114,
      "mean_token_accuracy": 0.23735624104738234,
      "num_tokens": 101299037.0,
      "step": 45380
    },
    {
      "entropy": 5.33274245262146,
      "epoch": 5.0971528050766555,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002720371101738855,
      "loss": 4.8678,
      "mean_token_accuracy": 0.23403965681791306,
      "num_tokens": 101309921.0,
      "step": 45385
    },
    {
      "entropy": 5.249806070327759,
      "epoch": 5.097714381984613,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027199696498342893,
      "loss": 4.7363,
      "mean_token_accuracy": 0.2505019798874855,
      "num_tokens": 101322265.0,
      "step": 45390
    },
    {
      "entropy": 5.273978662490845,
      "epoch": 5.09827595889257,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002719568198885901,
      "loss": 4.7363,
      "mean_token_accuracy": 0.24427465349435806,
      "num_tokens": 101334795.0,
      "step": 45395
    },
    {
      "entropy": 5.311184597015381,
      "epoch": 5.098837535800528,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002719166748906473,
      "loss": 4.8524,
      "mean_token_accuracy": 0.23572950661182404,
      "num_tokens": 101346216.0,
      "step": 45400
    },
    {
      "entropy": 5.349716901779175,
      "epoch": 5.099399112708485,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00027187652999087887,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2299988627433777,
      "num_tokens": 101357705.0,
      "step": 45405
    },
    {
      "entropy": 5.3707081317901615,
      "epoch": 5.099960689616443,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027183638519056284,
      "loss": 4.8985,
      "mean_token_accuracy": 0.23500577956438065,
      "num_tokens": 101370598.0,
      "step": 45410
    },
    {
      "entropy": 5.335847663879394,
      "epoch": 5.1005222665244005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002717962404909776,
      "loss": 4.8221,
      "mean_token_accuracy": 0.23786651343107224,
      "num_tokens": 101381942.0,
      "step": 45415
    },
    {
      "entropy": 5.277440309524536,
      "epoch": 5.101083843432358,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027175609589340135,
      "loss": 4.7292,
      "mean_token_accuracy": 0.23367058485746384,
      "num_tokens": 101392861.0,
      "step": 45420
    },
    {
      "entropy": 5.295115900039673,
      "epoch": 5.101645420340316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002717159513991122,
      "loss": 4.8391,
      "mean_token_accuracy": 0.23846681863069535,
      "num_tokens": 101403626.0,
      "step": 45425
    },
    {
      "entropy": 5.37615270614624,
      "epoch": 5.102206997248273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027167580700938844,
      "loss": 4.8674,
      "mean_token_accuracy": 0.2312498301267624,
      "num_tokens": 101415235.0,
      "step": 45430
    },
    {
      "entropy": 5.322395992279053,
      "epoch": 5.10276857415623,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00027163566272550826,
      "loss": 4.8872,
      "mean_token_accuracy": 0.2284943863749504,
      "num_tokens": 101426873.0,
      "step": 45435
    },
    {
      "entropy": 5.380699396133423,
      "epoch": 5.103330151064188,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027159551854874986,
      "loss": 4.838,
      "mean_token_accuracy": 0.23669617176055907,
      "num_tokens": 101437943.0,
      "step": 45440
    },
    {
      "entropy": 5.3982501983642575,
      "epoch": 5.103891727972146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002715553744803915,
      "loss": 4.9049,
      "mean_token_accuracy": 0.23057395815849305,
      "num_tokens": 101451141.0,
      "step": 45445
    },
    {
      "entropy": 5.495499563217163,
      "epoch": 5.104453304880104,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002715152305217114,
      "loss": 5.0169,
      "mean_token_accuracy": 0.22851019203662873,
      "num_tokens": 101463652.0,
      "step": 45450
    },
    {
      "entropy": 5.366817426681519,
      "epoch": 5.105014881788061,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002714750866739877,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22865240275859833,
      "num_tokens": 101475172.0,
      "step": 45455
    },
    {
      "entropy": 5.2803913116455075,
      "epoch": 5.105576458696018,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002714349429384986,
      "loss": 4.7283,
      "mean_token_accuracy": 0.2495747059583664,
      "num_tokens": 101486386.0,
      "step": 45460
    },
    {
      "entropy": 5.233789920806885,
      "epoch": 5.106138035603976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027139479931652226,
      "loss": 4.6834,
      "mean_token_accuracy": 0.24208881258964537,
      "num_tokens": 101497267.0,
      "step": 45465
    },
    {
      "entropy": 5.376898336410522,
      "epoch": 5.1066996125119335,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002713546558093368,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2249480217695236,
      "num_tokens": 101508089.0,
      "step": 45470
    },
    {
      "entropy": 5.256248807907104,
      "epoch": 5.107261189419891,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002713145124182207,
      "loss": 4.6955,
      "mean_token_accuracy": 0.24082479774951934,
      "num_tokens": 101519841.0,
      "step": 45475
    },
    {
      "entropy": 5.291567993164063,
      "epoch": 5.107822766327849,
      "grad_norm": 1.0,
      "learning_rate": 0.00027127436914445194,
      "loss": 4.7863,
      "mean_token_accuracy": 0.23747073262929916,
      "num_tokens": 101531396.0,
      "step": 45480
    },
    {
      "entropy": 5.2341633319854735,
      "epoch": 5.108384343235806,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002712342259893086,
      "loss": 4.6901,
      "mean_token_accuracy": 0.2474234566092491,
      "num_tokens": 101541648.0,
      "step": 45485
    },
    {
      "entropy": 5.385581541061401,
      "epoch": 5.108945920143763,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00027119408295406906,
      "loss": 4.9275,
      "mean_token_accuracy": 0.23242573738098143,
      "num_tokens": 101552975.0,
      "step": 45490
    },
    {
      "entropy": 5.276618003845215,
      "epoch": 5.109507497051721,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002711539400400114,
      "loss": 4.8046,
      "mean_token_accuracy": 0.23313484191894532,
      "num_tokens": 101564563.0,
      "step": 45495
    },
    {
      "entropy": 5.246607160568237,
      "epoch": 5.110069073959679,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027111379724841373,
      "loss": 4.7205,
      "mean_token_accuracy": 0.23748010098934175,
      "num_tokens": 101575333.0,
      "step": 45500
    },
    {
      "entropy": 5.332402896881104,
      "epoch": 5.110630650867637,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027107365458055424,
      "loss": 4.8884,
      "mean_token_accuracy": 0.23066612780094148,
      "num_tokens": 101587764.0,
      "step": 45505
    },
    {
      "entropy": 5.268567514419556,
      "epoch": 5.111192227775594,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002710335120377112,
      "loss": 4.6926,
      "mean_token_accuracy": 0.2454391285777092,
      "num_tokens": 101598611.0,
      "step": 45510
    },
    {
      "entropy": 5.338338565826416,
      "epoch": 5.111753804683551,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00027099336962116253,
      "loss": 4.8228,
      "mean_token_accuracy": 0.23702238649129867,
      "num_tokens": 101609407.0,
      "step": 45515
    },
    {
      "entropy": 5.303874206542969,
      "epoch": 5.112315381591509,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027095322733218665,
      "loss": 4.765,
      "mean_token_accuracy": 0.24403916150331498,
      "num_tokens": 101620248.0,
      "step": 45520
    },
    {
      "entropy": 5.275510025024414,
      "epoch": 5.112876958499466,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002709130851720616,
      "loss": 4.7379,
      "mean_token_accuracy": 0.23651669621467591,
      "num_tokens": 101630474.0,
      "step": 45525
    },
    {
      "entropy": 5.247062063217163,
      "epoch": 5.113438535407424,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027087294314206544,
      "loss": 4.6871,
      "mean_token_accuracy": 0.24628996700048447,
      "num_tokens": 101640986.0,
      "step": 45530
    },
    {
      "entropy": 5.3085578918457035,
      "epoch": 5.114000112315382,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002708328012434764,
      "loss": 4.7901,
      "mean_token_accuracy": 0.24007055163383484,
      "num_tokens": 101652808.0,
      "step": 45535
    },
    {
      "entropy": 5.334859418869018,
      "epoch": 5.114561689223339,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002707926594775725,
      "loss": 4.8346,
      "mean_token_accuracy": 0.2303402081131935,
      "num_tokens": 101663260.0,
      "step": 45540
    },
    {
      "entropy": 5.321711874008178,
      "epoch": 5.115123266131297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027075251784563196,
      "loss": 4.7586,
      "mean_token_accuracy": 0.23698304146528243,
      "num_tokens": 101674350.0,
      "step": 45545
    },
    {
      "entropy": 5.296667242050171,
      "epoch": 5.115684843039254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000270712376348933,
      "loss": 4.8094,
      "mean_token_accuracy": 0.23043066412210464,
      "num_tokens": 101685138.0,
      "step": 45550
    },
    {
      "entropy": 5.277134752273559,
      "epoch": 5.1162464199472115,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002706722349887536,
      "loss": 4.7558,
      "mean_token_accuracy": 0.2398645430803299,
      "num_tokens": 101696049.0,
      "step": 45555
    },
    {
      "entropy": 5.341930961608886,
      "epoch": 5.11680799685517,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002706320937663719,
      "loss": 4.8475,
      "mean_token_accuracy": 0.23165763318538665,
      "num_tokens": 101706416.0,
      "step": 45560
    },
    {
      "entropy": 5.287344169616699,
      "epoch": 5.117369573763127,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002705919526830661,
      "loss": 4.7504,
      "mean_token_accuracy": 0.2532858371734619,
      "num_tokens": 101716682.0,
      "step": 45565
    },
    {
      "entropy": 5.280338859558105,
      "epoch": 5.117931150671084,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027055181174011415,
      "loss": 4.7668,
      "mean_token_accuracy": 0.24526961892843246,
      "num_tokens": 101727957.0,
      "step": 45570
    },
    {
      "entropy": 5.278573751449585,
      "epoch": 5.118492727579042,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027051167093879424,
      "loss": 4.8016,
      "mean_token_accuracy": 0.24180636554956436,
      "num_tokens": 101738543.0,
      "step": 45575
    },
    {
      "entropy": 5.285947847366333,
      "epoch": 5.119054304486999,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002704715302803846,
      "loss": 4.7897,
      "mean_token_accuracy": 0.23612888306379318,
      "num_tokens": 101748295.0,
      "step": 45580
    },
    {
      "entropy": 5.30083532333374,
      "epoch": 5.119615881394957,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027043138976616304,
      "loss": 4.7755,
      "mean_token_accuracy": 0.23314967304468154,
      "num_tokens": 101759442.0,
      "step": 45585
    },
    {
      "entropy": 5.29158501625061,
      "epoch": 5.120177458302915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027039124939740787,
      "loss": 4.773,
      "mean_token_accuracy": 0.2389997959136963,
      "num_tokens": 101770638.0,
      "step": 45590
    },
    {
      "entropy": 5.372732496261596,
      "epoch": 5.120739035210872,
      "grad_norm": 0.984375,
      "learning_rate": 0.00027035110917539713,
      "loss": 4.7939,
      "mean_token_accuracy": 0.23876810520887376,
      "num_tokens": 101782260.0,
      "step": 45595
    },
    {
      "entropy": 5.336882162094116,
      "epoch": 5.12130061211883,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027031096910140877,
      "loss": 4.8466,
      "mean_token_accuracy": 0.23518606126308442,
      "num_tokens": 101792682.0,
      "step": 45600
    },
    {
      "entropy": 5.293027257919311,
      "epoch": 5.121862189026787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027027082917672107,
      "loss": 4.8305,
      "mean_token_accuracy": 0.2372559502720833,
      "num_tokens": 101804285.0,
      "step": 45605
    },
    {
      "entropy": 5.283727502822876,
      "epoch": 5.1224237659347445,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027023068940261203,
      "loss": 4.7381,
      "mean_token_accuracy": 0.24679071009159087,
      "num_tokens": 101815672.0,
      "step": 45610
    },
    {
      "entropy": 5.281565999984741,
      "epoch": 5.122985342842703,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002701905497803595,
      "loss": 4.7606,
      "mean_token_accuracy": 0.24343560636043549,
      "num_tokens": 101825528.0,
      "step": 45615
    },
    {
      "entropy": 5.225766563415528,
      "epoch": 5.12354691975066,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002701504103112419,
      "loss": 4.7798,
      "mean_token_accuracy": 0.23707717806100845,
      "num_tokens": 101837003.0,
      "step": 45620
    },
    {
      "entropy": 5.342323207855225,
      "epoch": 5.124108496658617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027011027099653706,
      "loss": 4.7843,
      "mean_token_accuracy": 0.23723017573356628,
      "num_tokens": 101848348.0,
      "step": 45625
    },
    {
      "entropy": 5.372262001037598,
      "epoch": 5.124670073566575,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002700701318375232,
      "loss": 4.8319,
      "mean_token_accuracy": 0.24270790219306945,
      "num_tokens": 101858444.0,
      "step": 45630
    },
    {
      "entropy": 5.324036169052124,
      "epoch": 5.125231650474532,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002700299928354781,
      "loss": 4.7848,
      "mean_token_accuracy": 0.23736221790313722,
      "num_tokens": 101868906.0,
      "step": 45635
    },
    {
      "entropy": 5.334273481369019,
      "epoch": 5.12579322738249,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00026998985399168005,
      "loss": 4.8432,
      "mean_token_accuracy": 0.23627472072839736,
      "num_tokens": 101879463.0,
      "step": 45640
    },
    {
      "entropy": 5.309261608123779,
      "epoch": 5.126354804290448,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026994971530740696,
      "loss": 4.8295,
      "mean_token_accuracy": 0.23788015991449357,
      "num_tokens": 101889433.0,
      "step": 45645
    },
    {
      "entropy": 5.413313961029052,
      "epoch": 5.126916381198405,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026990957678393684,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2246471181511879,
      "num_tokens": 101900554.0,
      "step": 45650
    },
    {
      "entropy": 5.327274560928345,
      "epoch": 5.127477958106363,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00026986943842254783,
      "loss": 4.7611,
      "mean_token_accuracy": 0.24421665966510772,
      "num_tokens": 101911205.0,
      "step": 45655
    },
    {
      "entropy": 5.189579820632934,
      "epoch": 5.12803953501432,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002698293002245179,
      "loss": 4.6868,
      "mean_token_accuracy": 0.24794643223285676,
      "num_tokens": 101924145.0,
      "step": 45660
    },
    {
      "entropy": 5.372443294525146,
      "epoch": 5.128601111922277,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026978916219112505,
      "loss": 4.8223,
      "mean_token_accuracy": 0.23775478154420854,
      "num_tokens": 101935433.0,
      "step": 45665
    },
    {
      "entropy": 5.272507572174073,
      "epoch": 5.1291626888302355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026974902432364727,
      "loss": 4.7255,
      "mean_token_accuracy": 0.23705395609140395,
      "num_tokens": 101945570.0,
      "step": 45670
    },
    {
      "entropy": 5.343429183959961,
      "epoch": 5.129724265738193,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026970888662336263,
      "loss": 4.9034,
      "mean_token_accuracy": 0.23015503138303756,
      "num_tokens": 101956410.0,
      "step": 45675
    },
    {
      "entropy": 5.284977722167969,
      "epoch": 5.130285842646151,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002696687490915491,
      "loss": 4.7231,
      "mean_token_accuracy": 0.24318659901618958,
      "num_tokens": 101966841.0,
      "step": 45680
    },
    {
      "entropy": 5.298451280593872,
      "epoch": 5.130847419554108,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026962861172948456,
      "loss": 4.7934,
      "mean_token_accuracy": 0.23785817474126816,
      "num_tokens": 101978169.0,
      "step": 45685
    },
    {
      "entropy": 5.285321950912476,
      "epoch": 5.131408996462065,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026958847453844713,
      "loss": 4.7584,
      "mean_token_accuracy": 0.24467013627290726,
      "num_tokens": 101987703.0,
      "step": 45690
    },
    {
      "entropy": 5.2324730396270756,
      "epoch": 5.131970573370023,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026954833751971486,
      "loss": 4.7491,
      "mean_token_accuracy": 0.23927617222070693,
      "num_tokens": 101999192.0,
      "step": 45695
    },
    {
      "entropy": 5.263344192504883,
      "epoch": 5.132532150277981,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002695082006745656,
      "loss": 4.7585,
      "mean_token_accuracy": 0.23799145072698594,
      "num_tokens": 102010610.0,
      "step": 45700
    },
    {
      "entropy": 5.345568561553955,
      "epoch": 5.133093727185938,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002694680640042774,
      "loss": 4.7188,
      "mean_token_accuracy": 0.2420754015445709,
      "num_tokens": 102020441.0,
      "step": 45705
    },
    {
      "entropy": 5.3496051788330075,
      "epoch": 5.133655304093896,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026942792751012814,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23721442520618438,
      "num_tokens": 102030965.0,
      "step": 45710
    },
    {
      "entropy": 5.231669950485229,
      "epoch": 5.134216881001853,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026938779119339585,
      "loss": 4.7561,
      "mean_token_accuracy": 0.2363509029150009,
      "num_tokens": 102041845.0,
      "step": 45715
    },
    {
      "entropy": 5.35624737739563,
      "epoch": 5.13477845790981,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002693476550553584,
      "loss": 4.9536,
      "mean_token_accuracy": 0.22683112174272538,
      "num_tokens": 102053578.0,
      "step": 45720
    },
    {
      "entropy": 5.386167764663696,
      "epoch": 5.1353400348177685,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026930751909729384,
      "loss": 4.8559,
      "mean_token_accuracy": 0.24191514402627945,
      "num_tokens": 102064235.0,
      "step": 45725
    },
    {
      "entropy": 5.25008487701416,
      "epoch": 5.135901611725726,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00026926738332048013,
      "loss": 4.682,
      "mean_token_accuracy": 0.24812377542257308,
      "num_tokens": 102075213.0,
      "step": 45730
    },
    {
      "entropy": 5.350066804885865,
      "epoch": 5.136463188633684,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026922724772619516,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22801416963338852,
      "num_tokens": 102086651.0,
      "step": 45735
    },
    {
      "entropy": 5.332602405548096,
      "epoch": 5.137024765541641,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002691871123157168,
      "loss": 4.8458,
      "mean_token_accuracy": 0.23010269105434417,
      "num_tokens": 102097396.0,
      "step": 45740
    },
    {
      "entropy": 5.287755727767944,
      "epoch": 5.137586342449598,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002691469770903231,
      "loss": 4.7763,
      "mean_token_accuracy": 0.23839548528194426,
      "num_tokens": 102108009.0,
      "step": 45745
    },
    {
      "entropy": 5.255586576461792,
      "epoch": 5.138147919357556,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002691068420512919,
      "loss": 4.6838,
      "mean_token_accuracy": 0.24887241274118424,
      "num_tokens": 102118158.0,
      "step": 45750
    },
    {
      "entropy": 5.324717855453491,
      "epoch": 5.1387094962655135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000269066707199901,
      "loss": 4.8695,
      "mean_token_accuracy": 0.23962584286928176,
      "num_tokens": 102130232.0,
      "step": 45755
    },
    {
      "entropy": 5.344628047943115,
      "epoch": 5.139271073173471,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026902657253742855,
      "loss": 4.8649,
      "mean_token_accuracy": 0.23794818669557571,
      "num_tokens": 102142698.0,
      "step": 45760
    },
    {
      "entropy": 5.310995483398438,
      "epoch": 5.139832650081429,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002689864380651524,
      "loss": 4.769,
      "mean_token_accuracy": 0.23484190702438354,
      "num_tokens": 102156270.0,
      "step": 45765
    },
    {
      "entropy": 5.299819660186768,
      "epoch": 5.140394226989386,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002689463037843504,
      "loss": 4.7619,
      "mean_token_accuracy": 0.2388528913259506,
      "num_tokens": 102168369.0,
      "step": 45770
    },
    {
      "entropy": 5.298230504989624,
      "epoch": 5.140955803897343,
      "grad_norm": 1.125,
      "learning_rate": 0.0002689061696963004,
      "loss": 4.8193,
      "mean_token_accuracy": 0.23165415823459626,
      "num_tokens": 102180191.0,
      "step": 45775
    },
    {
      "entropy": 5.272256374359131,
      "epoch": 5.141517380805301,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002688660358022803,
      "loss": 4.8061,
      "mean_token_accuracy": 0.23398303091526032,
      "num_tokens": 102191589.0,
      "step": 45780
    },
    {
      "entropy": 5.265850830078125,
      "epoch": 5.142078957713259,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000268825902103568,
      "loss": 4.7508,
      "mean_token_accuracy": 0.2422882705926895,
      "num_tokens": 102202847.0,
      "step": 45785
    },
    {
      "entropy": 5.32477388381958,
      "epoch": 5.142640534621217,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002687857686014414,
      "loss": 4.8217,
      "mean_token_accuracy": 0.24039259403944016,
      "num_tokens": 102214353.0,
      "step": 45790
    },
    {
      "entropy": 5.390439891815186,
      "epoch": 5.143202111529174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002687456352971783,
      "loss": 4.856,
      "mean_token_accuracy": 0.2356734022498131,
      "num_tokens": 102225149.0,
      "step": 45795
    },
    {
      "entropy": 5.345282745361328,
      "epoch": 5.143763688437131,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002687055021920566,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23692155182361602,
      "num_tokens": 102236364.0,
      "step": 45800
    },
    {
      "entropy": 5.3132226943969725,
      "epoch": 5.144325265345089,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002686653692873542,
      "loss": 4.8207,
      "mean_token_accuracy": 0.23870041221380234,
      "num_tokens": 102247250.0,
      "step": 45805
    },
    {
      "entropy": 5.283676099777222,
      "epoch": 5.1448868422530465,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00026862523658434883,
      "loss": 4.7715,
      "mean_token_accuracy": 0.24163332283496858,
      "num_tokens": 102257502.0,
      "step": 45810
    },
    {
      "entropy": 5.355433368682862,
      "epoch": 5.145448419161004,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002685851040843185,
      "loss": 4.8529,
      "mean_token_accuracy": 0.23724713176488876,
      "num_tokens": 102268818.0,
      "step": 45815
    },
    {
      "entropy": 5.213472127914429,
      "epoch": 5.146009996068962,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026854497178854087,
      "loss": 4.7452,
      "mean_token_accuracy": 0.244953915476799,
      "num_tokens": 102278406.0,
      "step": 45820
    },
    {
      "entropy": 5.331564950942993,
      "epoch": 5.146571572976919,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002685048396982939,
      "loss": 4.8669,
      "mean_token_accuracy": 0.22809667736291886,
      "num_tokens": 102290053.0,
      "step": 45825
    },
    {
      "entropy": 5.358632946014405,
      "epoch": 5.147133149884877,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026846470781485517,
      "loss": 4.8636,
      "mean_token_accuracy": 0.23477908223867416,
      "num_tokens": 102300803.0,
      "step": 45830
    },
    {
      "entropy": 5.376092910766602,
      "epoch": 5.147694726792834,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002684245761395028,
      "loss": 4.8306,
      "mean_token_accuracy": 0.23488317131996156,
      "num_tokens": 102311605.0,
      "step": 45835
    },
    {
      "entropy": 5.456468296051026,
      "epoch": 5.148256303700792,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026838444467351447,
      "loss": 4.9717,
      "mean_token_accuracy": 0.2214631989598274,
      "num_tokens": 102322888.0,
      "step": 45840
    },
    {
      "entropy": 5.410055255889892,
      "epoch": 5.14881788060875,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026834431341816805,
      "loss": 4.9001,
      "mean_token_accuracy": 0.23285947144031524,
      "num_tokens": 102333724.0,
      "step": 45845
    },
    {
      "entropy": 5.3645124435424805,
      "epoch": 5.149379457516707,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026830418237474126,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22932861596345902,
      "num_tokens": 102345971.0,
      "step": 45850
    },
    {
      "entropy": 5.328705978393555,
      "epoch": 5.149941034424664,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002682640515445118,
      "loss": 4.8102,
      "mean_token_accuracy": 0.23854778259992598,
      "num_tokens": 102357274.0,
      "step": 45855
    },
    {
      "entropy": 5.304092359542847,
      "epoch": 5.150502611332622,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002682239209287577,
      "loss": 4.8295,
      "mean_token_accuracy": 0.23925930708646775,
      "num_tokens": 102368497.0,
      "step": 45860
    },
    {
      "entropy": 5.24492883682251,
      "epoch": 5.151064188240579,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026818379052875646,
      "loss": 4.7508,
      "mean_token_accuracy": 0.2431942567229271,
      "num_tokens": 102378812.0,
      "step": 45865
    },
    {
      "entropy": 5.219981670379639,
      "epoch": 5.1516257651485375,
      "grad_norm": 1.078125,
      "learning_rate": 0.000268143660345786,
      "loss": 4.7561,
      "mean_token_accuracy": 0.2441585659980774,
      "num_tokens": 102389282.0,
      "step": 45870
    },
    {
      "entropy": 5.380130243301392,
      "epoch": 5.152187342056495,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002681035303811241,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2310943990945816,
      "num_tokens": 102400694.0,
      "step": 45875
    },
    {
      "entropy": 5.414079475402832,
      "epoch": 5.152748918964452,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026806340063604845,
      "loss": 4.8098,
      "mean_token_accuracy": 0.2337352767586708,
      "num_tokens": 102411947.0,
      "step": 45880
    },
    {
      "entropy": 5.306844711303711,
      "epoch": 5.15331049587241,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026802327111183686,
      "loss": 4.7339,
      "mean_token_accuracy": 0.24261675477027894,
      "num_tokens": 102422146.0,
      "step": 45885
    },
    {
      "entropy": 5.234015226364136,
      "epoch": 5.153872072780367,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026798314180976696,
      "loss": 4.7592,
      "mean_token_accuracy": 0.2390228033065796,
      "num_tokens": 102434077.0,
      "step": 45890
    },
    {
      "entropy": 5.3289988994598385,
      "epoch": 5.1544336496883245,
      "grad_norm": 0.953125,
      "learning_rate": 0.00026794301273111654,
      "loss": 4.8138,
      "mean_token_accuracy": 0.233797287940979,
      "num_tokens": 102446812.0,
      "step": 45895
    },
    {
      "entropy": 5.3180272579193115,
      "epoch": 5.154995226596283,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002679028838771633,
      "loss": 4.83,
      "mean_token_accuracy": 0.23212722390890123,
      "num_tokens": 102459198.0,
      "step": 45900
    },
    {
      "entropy": 5.369432353973389,
      "epoch": 5.15555680350424,
      "grad_norm": 0.96875,
      "learning_rate": 0.00026786275524918503,
      "loss": 4.8539,
      "mean_token_accuracy": 0.238254676759243,
      "num_tokens": 102470932.0,
      "step": 45905
    },
    {
      "entropy": 5.3772646427154545,
      "epoch": 5.156118380412197,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002678226268484594,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23620584309101106,
      "num_tokens": 102482349.0,
      "step": 45910
    },
    {
      "entropy": 5.353902292251587,
      "epoch": 5.156679957320155,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002677824986762641,
      "loss": 4.8171,
      "mean_token_accuracy": 0.23579344302415847,
      "num_tokens": 102493782.0,
      "step": 45915
    },
    {
      "entropy": 5.259968900680542,
      "epoch": 5.157241534228112,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002677423707338769,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23308838456869124,
      "num_tokens": 102506106.0,
      "step": 45920
    },
    {
      "entropy": 5.282429456710815,
      "epoch": 5.1578031111360705,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002677022430225754,
      "loss": 4.7258,
      "mean_token_accuracy": 0.2401619479060173,
      "num_tokens": 102516482.0,
      "step": 45925
    },
    {
      "entropy": 5.3451117992401125,
      "epoch": 5.158364688044028,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026766211554363727,
      "loss": 4.8514,
      "mean_token_accuracy": 0.23331603556871414,
      "num_tokens": 102526477.0,
      "step": 45930
    },
    {
      "entropy": 5.367437791824341,
      "epoch": 5.158926264951985,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002676219882983403,
      "loss": 4.8349,
      "mean_token_accuracy": 0.23348188847303392,
      "num_tokens": 102538381.0,
      "step": 45935
    },
    {
      "entropy": 5.238291168212891,
      "epoch": 5.159487841859943,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026758186128796204,
      "loss": 4.7097,
      "mean_token_accuracy": 0.24515151530504226,
      "num_tokens": 102548595.0,
      "step": 45940
    },
    {
      "entropy": 5.308560848236084,
      "epoch": 5.1600494187679,
      "grad_norm": 1.0,
      "learning_rate": 0.00026754173451378014,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23282752484083175,
      "num_tokens": 102560560.0,
      "step": 45945
    },
    {
      "entropy": 5.321582508087158,
      "epoch": 5.1606109956758575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026750160797707245,
      "loss": 4.7632,
      "mean_token_accuracy": 0.24069393575191497,
      "num_tokens": 102570623.0,
      "step": 45950
    },
    {
      "entropy": 5.301931571960449,
      "epoch": 5.161172572583816,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002674614816791164,
      "loss": 4.8556,
      "mean_token_accuracy": 0.23833516538143157,
      "num_tokens": 102581662.0,
      "step": 45955
    },
    {
      "entropy": 5.253459644317627,
      "epoch": 5.161734149491773,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00026742135562118967,
      "loss": 4.7917,
      "mean_token_accuracy": 0.2396291509270668,
      "num_tokens": 102593875.0,
      "step": 45960
    },
    {
      "entropy": 5.342164802551269,
      "epoch": 5.16229572639973,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002673812298045699,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23699143081903457,
      "num_tokens": 102604639.0,
      "step": 45965
    },
    {
      "entropy": 5.347805213928223,
      "epoch": 5.162857303307688,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026734110423053483,
      "loss": 4.8084,
      "mean_token_accuracy": 0.23384298682212828,
      "num_tokens": 102614297.0,
      "step": 45970
    },
    {
      "entropy": 5.384401178359985,
      "epoch": 5.163418880215645,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026730097890036175,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23492816239595413,
      "num_tokens": 102625480.0,
      "step": 45975
    },
    {
      "entropy": 5.332390546798706,
      "epoch": 5.163980457123603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026726085381532867,
      "loss": 4.8217,
      "mean_token_accuracy": 0.24194729328155518,
      "num_tokens": 102637342.0,
      "step": 45980
    },
    {
      "entropy": 5.4014825344085695,
      "epoch": 5.164542034031561,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000267220728976713,
      "loss": 4.8661,
      "mean_token_accuracy": 0.23513562381267547,
      "num_tokens": 102646886.0,
      "step": 45985
    },
    {
      "entropy": 5.369273138046265,
      "epoch": 5.165103610939518,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002671806043857923,
      "loss": 4.8445,
      "mean_token_accuracy": 0.23992279618978501,
      "num_tokens": 102658929.0,
      "step": 45990
    },
    {
      "entropy": 5.436691474914551,
      "epoch": 5.165665187847476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002671404800438442,
      "loss": 4.9829,
      "mean_token_accuracy": 0.22102392315864564,
      "num_tokens": 102670308.0,
      "step": 45995
    },
    {
      "entropy": 5.326947641372681,
      "epoch": 5.166226764755433,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002671003559521463,
      "loss": 4.7571,
      "mean_token_accuracy": 0.238910211622715,
      "num_tokens": 102680714.0,
      "step": 46000
    },
    {
      "entropy": 5.350050640106201,
      "epoch": 5.16678834166339,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00026706023211197613,
      "loss": 4.8495,
      "mean_token_accuracy": 0.2287491038441658,
      "num_tokens": 102692016.0,
      "step": 46005
    },
    {
      "entropy": 5.289581298828125,
      "epoch": 5.1673499185713485,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002670201085246111,
      "loss": 4.7417,
      "mean_token_accuracy": 0.24180605709552766,
      "num_tokens": 102702901.0,
      "step": 46010
    },
    {
      "entropy": 5.318958377838134,
      "epoch": 5.167911495479306,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026697998519132905,
      "loss": 4.7595,
      "mean_token_accuracy": 0.241864213347435,
      "num_tokens": 102714830.0,
      "step": 46015
    },
    {
      "entropy": 5.376074552536011,
      "epoch": 5.168473072387264,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026693986211340736,
      "loss": 4.8871,
      "mean_token_accuracy": 0.23174970149993895,
      "num_tokens": 102726417.0,
      "step": 46020
    },
    {
      "entropy": 5.283934783935547,
      "epoch": 5.169034649295221,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026689973929212363,
      "loss": 4.7028,
      "mean_token_accuracy": 0.24186822175979614,
      "num_tokens": 102737399.0,
      "step": 46025
    },
    {
      "entropy": 5.396965551376343,
      "epoch": 5.169596226203178,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026685961672875524,
      "loss": 4.9483,
      "mean_token_accuracy": 0.23507405519485475,
      "num_tokens": 102748412.0,
      "step": 46030
    },
    {
      "entropy": 5.325987863540649,
      "epoch": 5.170157803111136,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002668194944245799,
      "loss": 4.7969,
      "mean_token_accuracy": 0.2398018717765808,
      "num_tokens": 102759538.0,
      "step": 46035
    },
    {
      "entropy": 5.326444673538208,
      "epoch": 5.170719380019094,
      "grad_norm": 1.046875,
      "learning_rate": 0.000266779372380875,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22305290251970292,
      "num_tokens": 102771660.0,
      "step": 46040
    },
    {
      "entropy": 5.352777290344238,
      "epoch": 5.171280956927051,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000266739250598918,
      "loss": 4.8322,
      "mean_token_accuracy": 0.23046321123838426,
      "num_tokens": 102781968.0,
      "step": 46045
    },
    {
      "entropy": 5.321163463592529,
      "epoch": 5.171842533835009,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026669912907998657,
      "loss": 4.8591,
      "mean_token_accuracy": 0.22982878535985946,
      "num_tokens": 102792734.0,
      "step": 46050
    },
    {
      "entropy": 5.271653890609741,
      "epoch": 5.172404110742966,
      "grad_norm": 1.03125,
      "learning_rate": 0.000266659007825358,
      "loss": 4.7592,
      "mean_token_accuracy": 0.24279282838106156,
      "num_tokens": 102804423.0,
      "step": 46055
    },
    {
      "entropy": 5.315888404846191,
      "epoch": 5.172965687650924,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026661888683631,
      "loss": 4.7744,
      "mean_token_accuracy": 0.2453315868973732,
      "num_tokens": 102815615.0,
      "step": 46060
    },
    {
      "entropy": 5.364956665039062,
      "epoch": 5.1735272645588815,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026657876611411974,
      "loss": 4.851,
      "mean_token_accuracy": 0.23277200609445572,
      "num_tokens": 102827389.0,
      "step": 46065
    },
    {
      "entropy": 5.257541656494141,
      "epoch": 5.174088841466839,
      "grad_norm": 1.0,
      "learning_rate": 0.00026653864566006494,
      "loss": 4.759,
      "mean_token_accuracy": 0.23918040841817856,
      "num_tokens": 102838570.0,
      "step": 46070
    },
    {
      "entropy": 5.271478176116943,
      "epoch": 5.174650418374797,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002664985254754229,
      "loss": 4.7991,
      "mean_token_accuracy": 0.23386719673871995,
      "num_tokens": 102850465.0,
      "step": 46075
    },
    {
      "entropy": 5.257231998443603,
      "epoch": 5.175211995282754,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026645840556147103,
      "loss": 4.7565,
      "mean_token_accuracy": 0.25054467767477034,
      "num_tokens": 102861281.0,
      "step": 46080
    },
    {
      "entropy": 5.345813798904419,
      "epoch": 5.175773572190711,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002664182859194869,
      "loss": 4.8366,
      "mean_token_accuracy": 0.23155570328235625,
      "num_tokens": 102873336.0,
      "step": 46085
    },
    {
      "entropy": 5.302256536483765,
      "epoch": 5.176335149098669,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002663781665507479,
      "loss": 4.8176,
      "mean_token_accuracy": 0.23749894946813582,
      "num_tokens": 102884468.0,
      "step": 46090
    },
    {
      "entropy": 5.336627769470215,
      "epoch": 5.1768967260066265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026633804745653146,
      "loss": 4.8521,
      "mean_token_accuracy": 0.23346900343894958,
      "num_tokens": 102896967.0,
      "step": 46095
    },
    {
      "entropy": 5.339280891418457,
      "epoch": 5.177458302914584,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026629792863811484,
      "loss": 4.8466,
      "mean_token_accuracy": 0.23347465693950653,
      "num_tokens": 102908365.0,
      "step": 46100
    },
    {
      "entropy": 5.343844985961914,
      "epoch": 5.178019879822542,
      "grad_norm": 1.125,
      "learning_rate": 0.0002662578100967756,
      "loss": 4.8101,
      "mean_token_accuracy": 0.23865189999341965,
      "num_tokens": 102918256.0,
      "step": 46105
    },
    {
      "entropy": 5.275404691696167,
      "epoch": 5.178581456730499,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002662176918337911,
      "loss": 4.7595,
      "mean_token_accuracy": 0.24222324937582015,
      "num_tokens": 102928547.0,
      "step": 46110
    },
    {
      "entropy": 5.3406671524047855,
      "epoch": 5.179143033638457,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002661775738504387,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2355419546365738,
      "num_tokens": 102939345.0,
      "step": 46115
    },
    {
      "entropy": 5.409253025054932,
      "epoch": 5.179704610546414,
      "grad_norm": 1.0,
      "learning_rate": 0.00026613745614799574,
      "loss": 4.8956,
      "mean_token_accuracy": 0.23028626292943954,
      "num_tokens": 102949962.0,
      "step": 46120
    },
    {
      "entropy": 5.371373844146729,
      "epoch": 5.180266187454372,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002660973387277396,
      "loss": 4.8223,
      "mean_token_accuracy": 0.22753997892141342,
      "num_tokens": 102960951.0,
      "step": 46125
    },
    {
      "entropy": 5.312120008468628,
      "epoch": 5.18082776436233,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002660572215909477,
      "loss": 4.7693,
      "mean_token_accuracy": 0.23935734778642653,
      "num_tokens": 102972350.0,
      "step": 46130
    },
    {
      "entropy": 5.3044655323028564,
      "epoch": 5.181389341270287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002660171047388973,
      "loss": 4.8434,
      "mean_token_accuracy": 0.2335546538233757,
      "num_tokens": 102984177.0,
      "step": 46135
    },
    {
      "entropy": 5.295929670333862,
      "epoch": 5.181950918178244,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00026597698817286576,
      "loss": 4.8277,
      "mean_token_accuracy": 0.23754352033138276,
      "num_tokens": 102995609.0,
      "step": 46140
    },
    {
      "entropy": 5.384040403366089,
      "epoch": 5.182512495086202,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026593687189413044,
      "loss": 4.8546,
      "mean_token_accuracy": 0.24154627621173858,
      "num_tokens": 103006091.0,
      "step": 46145
    },
    {
      "entropy": 5.323021745681762,
      "epoch": 5.1830740719941595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026589675590396865,
      "loss": 4.7877,
      "mean_token_accuracy": 0.23794285506010054,
      "num_tokens": 103016861.0,
      "step": 46150
    },
    {
      "entropy": 5.304147672653198,
      "epoch": 5.183635648902117,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002658566402036576,
      "loss": 4.7806,
      "mean_token_accuracy": 0.23482141494750977,
      "num_tokens": 103027849.0,
      "step": 46155
    },
    {
      "entropy": 5.315965080261231,
      "epoch": 5.184197225810075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002658165247944747,
      "loss": 4.7907,
      "mean_token_accuracy": 0.2403961032629013,
      "num_tokens": 103040235.0,
      "step": 46160
    },
    {
      "entropy": 5.330029392242432,
      "epoch": 5.184758802718032,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026577640967769727,
      "loss": 4.8503,
      "mean_token_accuracy": 0.23993521183729172,
      "num_tokens": 103049823.0,
      "step": 46165
    },
    {
      "entropy": 5.37466139793396,
      "epoch": 5.18532037962599,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026573629485460257,
      "loss": 4.8217,
      "mean_token_accuracy": 0.2323736995458603,
      "num_tokens": 103060957.0,
      "step": 46170
    },
    {
      "entropy": 5.329695749282837,
      "epoch": 5.185881956533947,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026569618032646773,
      "loss": 4.8558,
      "mean_token_accuracy": 0.23460995852947236,
      "num_tokens": 103071092.0,
      "step": 46175
    },
    {
      "entropy": 5.332767820358276,
      "epoch": 5.186443533441905,
      "grad_norm": 0.9375,
      "learning_rate": 0.00026565606609457016,
      "loss": 4.8512,
      "mean_token_accuracy": 0.23423193991184235,
      "num_tokens": 103083021.0,
      "step": 46180
    },
    {
      "entropy": 5.251117944717407,
      "epoch": 5.187005110349863,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002656159521601871,
      "loss": 4.8097,
      "mean_token_accuracy": 0.23787901252508165,
      "num_tokens": 103093569.0,
      "step": 46185
    },
    {
      "entropy": 5.283539295196533,
      "epoch": 5.18756668725782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002655758385245956,
      "loss": 4.7848,
      "mean_token_accuracy": 0.23613962829113005,
      "num_tokens": 103104745.0,
      "step": 46190
    },
    {
      "entropy": 5.3724761486053465,
      "epoch": 5.188128264165777,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002655357251890732,
      "loss": 4.8218,
      "mean_token_accuracy": 0.23936327844858168,
      "num_tokens": 103117090.0,
      "step": 46195
    },
    {
      "entropy": 5.435968542098999,
      "epoch": 5.188689841073735,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026549561215489695,
      "loss": 4.8794,
      "mean_token_accuracy": 0.230899615585804,
      "num_tokens": 103128491.0,
      "step": 46200
    },
    {
      "entropy": 5.3306722164154055,
      "epoch": 5.189251417981692,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002654554994233442,
      "loss": 4.804,
      "mean_token_accuracy": 0.23342931568622588,
      "num_tokens": 103140028.0,
      "step": 46205
    },
    {
      "entropy": 5.297089910507202,
      "epoch": 5.1898129948896505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002654153869956919,
      "loss": 4.8116,
      "mean_token_accuracy": 0.23440655767917634,
      "num_tokens": 103150514.0,
      "step": 46210
    },
    {
      "entropy": 5.288339042663575,
      "epoch": 5.190374571797608,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002653752748732175,
      "loss": 4.7664,
      "mean_token_accuracy": 0.24235253036022186,
      "num_tokens": 103161409.0,
      "step": 46215
    },
    {
      "entropy": 5.382736015319824,
      "epoch": 5.190936148705565,
      "grad_norm": 1.0625,
      "learning_rate": 0.000265335163057198,
      "loss": 4.8271,
      "mean_token_accuracy": 0.24588370323181152,
      "num_tokens": 103172394.0,
      "step": 46220
    },
    {
      "entropy": 5.291476774215698,
      "epoch": 5.191497725613523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002652950515489107,
      "loss": 4.7872,
      "mean_token_accuracy": 0.23129139989614486,
      "num_tokens": 103183461.0,
      "step": 46225
    },
    {
      "entropy": 5.323620891571045,
      "epoch": 5.19205930252148,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002652549403496327,
      "loss": 4.8644,
      "mean_token_accuracy": 0.23399290889501573,
      "num_tokens": 103194495.0,
      "step": 46230
    },
    {
      "entropy": 5.430592727661133,
      "epoch": 5.1926208794294375,
      "grad_norm": 1.125,
      "learning_rate": 0.00026521482946064123,
      "loss": 4.9013,
      "mean_token_accuracy": 0.23201505839824677,
      "num_tokens": 103205287.0,
      "step": 46235
    },
    {
      "entropy": 5.285521602630615,
      "epoch": 5.193182456337396,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002651747188832134,
      "loss": 4.8834,
      "mean_token_accuracy": 0.23339492082595825,
      "num_tokens": 103216447.0,
      "step": 46240
    },
    {
      "entropy": 5.303513288497925,
      "epoch": 5.193744033245353,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002651346086186262,
      "loss": 4.8215,
      "mean_token_accuracy": 0.23778688609600068,
      "num_tokens": 103228663.0,
      "step": 46245
    },
    {
      "entropy": 5.329627799987793,
      "epoch": 5.194305610153311,
      "grad_norm": 1.125,
      "learning_rate": 0.000265094498668157,
      "loss": 4.8242,
      "mean_token_accuracy": 0.23827047944068908,
      "num_tokens": 103239698.0,
      "step": 46250
    },
    {
      "entropy": 5.372756385803223,
      "epoch": 5.194867187061268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026505438903308285,
      "loss": 4.8148,
      "mean_token_accuracy": 0.23956566154956818,
      "num_tokens": 103250498.0,
      "step": 46255
    },
    {
      "entropy": 5.326104021072387,
      "epoch": 5.195428763969225,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00026501427971468056,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22856909930706024,
      "num_tokens": 103262776.0,
      "step": 46260
    },
    {
      "entropy": 5.258111667633057,
      "epoch": 5.1959903408771835,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002649741707142277,
      "loss": 4.7831,
      "mean_token_accuracy": 0.23821176290512086,
      "num_tokens": 103274057.0,
      "step": 46265
    },
    {
      "entropy": 5.362839078903198,
      "epoch": 5.196551917785141,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026493406203300106,
      "loss": 4.7716,
      "mean_token_accuracy": 0.23728837072849274,
      "num_tokens": 103285038.0,
      "step": 46270
    },
    {
      "entropy": 5.3757997989654545,
      "epoch": 5.197113494693098,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026489395367227785,
      "loss": 4.8462,
      "mean_token_accuracy": 0.24165143370628356,
      "num_tokens": 103297135.0,
      "step": 46275
    },
    {
      "entropy": 5.2999162673950195,
      "epoch": 5.197675071601056,
      "grad_norm": 1.03125,
      "learning_rate": 0.000264853845633335,
      "loss": 4.7935,
      "mean_token_accuracy": 0.23148528933525087,
      "num_tokens": 103307597.0,
      "step": 46280
    },
    {
      "entropy": 5.342015886306763,
      "epoch": 5.198236648509013,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002648137379174497,
      "loss": 4.8398,
      "mean_token_accuracy": 0.23172418177127838,
      "num_tokens": 103318070.0,
      "step": 46285
    },
    {
      "entropy": 5.243734836578369,
      "epoch": 5.1987982254169705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002647736305258989,
      "loss": 4.767,
      "mean_token_accuracy": 0.24164282083511351,
      "num_tokens": 103329277.0,
      "step": 46290
    },
    {
      "entropy": 5.302700042724609,
      "epoch": 5.199359802324929,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002647335234599597,
      "loss": 4.7647,
      "mean_token_accuracy": 0.23718852251768113,
      "num_tokens": 103340349.0,
      "step": 46295
    },
    {
      "entropy": 5.291387605667114,
      "epoch": 5.199921379232886,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026469341672090906,
      "loss": 4.8575,
      "mean_token_accuracy": 0.2299127146601677,
      "num_tokens": 103351818.0,
      "step": 46300
    },
    {
      "entropy": 5.281939697265625,
      "epoch": 5.200482956140844,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026465331031002404,
      "loss": 4.8037,
      "mean_token_accuracy": 0.23634101897478105,
      "num_tokens": 103362092.0,
      "step": 46305
    },
    {
      "entropy": 5.368774032592773,
      "epoch": 5.201044533048801,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026461320422858166,
      "loss": 4.866,
      "mean_token_accuracy": 0.23111395686864852,
      "num_tokens": 103372237.0,
      "step": 46310
    },
    {
      "entropy": 5.334552335739136,
      "epoch": 5.201606109956758,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002645730984778589,
      "loss": 4.7697,
      "mean_token_accuracy": 0.24120177775621415,
      "num_tokens": 103382191.0,
      "step": 46315
    },
    {
      "entropy": 5.359958505630493,
      "epoch": 5.202167686864716,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002645329930591327,
      "loss": 4.884,
      "mean_token_accuracy": 0.23547806739807128,
      "num_tokens": 103392299.0,
      "step": 46320
    },
    {
      "entropy": 5.2710168838500975,
      "epoch": 5.202729263772674,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00026449288797368006,
      "loss": 4.7839,
      "mean_token_accuracy": 0.2354392796754837,
      "num_tokens": 103403846.0,
      "step": 46325
    },
    {
      "entropy": 5.299375247955322,
      "epoch": 5.203290840680631,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00026445278322277796,
      "loss": 4.726,
      "mean_token_accuracy": 0.23402297347784043,
      "num_tokens": 103414773.0,
      "step": 46330
    },
    {
      "entropy": 5.3416444778442385,
      "epoch": 5.203852417588589,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002644126788077032,
      "loss": 4.8592,
      "mean_token_accuracy": 0.23241541385650635,
      "num_tokens": 103427991.0,
      "step": 46335
    },
    {
      "entropy": 5.274004220962524,
      "epoch": 5.204413994496546,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00026437257472973296,
      "loss": 4.8468,
      "mean_token_accuracy": 0.238724485039711,
      "num_tokens": 103439605.0,
      "step": 46340
    },
    {
      "entropy": 5.2903495788574215,
      "epoch": 5.204975571404503,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026433247099014406,
      "loss": 4.776,
      "mean_token_accuracy": 0.23571933209896087,
      "num_tokens": 103450327.0,
      "step": 46345
    },
    {
      "entropy": 5.305609321594238,
      "epoch": 5.2055371483124615,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002642923675902135,
      "loss": 4.8351,
      "mean_token_accuracy": 0.2338385209441185,
      "num_tokens": 103461238.0,
      "step": 46350
    },
    {
      "entropy": 5.292350482940674,
      "epoch": 5.206098725220419,
      "grad_norm": 1.0625,
      "learning_rate": 0.000264252264531218,
      "loss": 4.7694,
      "mean_token_accuracy": 0.24021084010601043,
      "num_tokens": 103471625.0,
      "step": 46355
    },
    {
      "entropy": 5.290216302871704,
      "epoch": 5.206660302128377,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026421216181443457,
      "loss": 4.7811,
      "mean_token_accuracy": 0.23807192146778106,
      "num_tokens": 103483758.0,
      "step": 46360
    },
    {
      "entropy": 5.356973886489868,
      "epoch": 5.207221879036334,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002641720594411401,
      "loss": 4.8833,
      "mean_token_accuracy": 0.2327433481812477,
      "num_tokens": 103495733.0,
      "step": 46365
    },
    {
      "entropy": 5.284691333770752,
      "epoch": 5.207783455944291,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026413195741261145,
      "loss": 4.7996,
      "mean_token_accuracy": 0.23395009934902192,
      "num_tokens": 103506689.0,
      "step": 46370
    },
    {
      "entropy": 5.3217857837677,
      "epoch": 5.208345032852249,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002640918557301256,
      "loss": 4.8949,
      "mean_token_accuracy": 0.23327019810676575,
      "num_tokens": 103519707.0,
      "step": 46375
    },
    {
      "entropy": 5.416431283950805,
      "epoch": 5.208906609760207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002640517543949592,
      "loss": 4.867,
      "mean_token_accuracy": 0.22863890677690507,
      "num_tokens": 103530888.0,
      "step": 46380
    },
    {
      "entropy": 5.35195484161377,
      "epoch": 5.209468186668164,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002640116534083892,
      "loss": 4.8266,
      "mean_token_accuracy": 0.23562142997980118,
      "num_tokens": 103541537.0,
      "step": 46385
    },
    {
      "entropy": 5.294554710388184,
      "epoch": 5.210029763576122,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026397155277169246,
      "loss": 4.7894,
      "mean_token_accuracy": 0.2398407682776451,
      "num_tokens": 103552388.0,
      "step": 46390
    },
    {
      "entropy": 5.338511800765991,
      "epoch": 5.210591340484079,
      "grad_norm": 1.0,
      "learning_rate": 0.0002639314524861457,
      "loss": 4.8899,
      "mean_token_accuracy": 0.23471634089946747,
      "num_tokens": 103564029.0,
      "step": 46395
    },
    {
      "entropy": 5.3889354228973385,
      "epoch": 5.211152917392037,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002638913525530258,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23692228496074677,
      "num_tokens": 103574756.0,
      "step": 46400
    },
    {
      "entropy": 5.396828031539917,
      "epoch": 5.2117144942999944,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026385125297360946,
      "loss": 4.848,
      "mean_token_accuracy": 0.23954593986272812,
      "num_tokens": 103585810.0,
      "step": 46405
    },
    {
      "entropy": 5.312682151794434,
      "epoch": 5.212276071207952,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026381115374917363,
      "loss": 4.8426,
      "mean_token_accuracy": 0.24144833385944367,
      "num_tokens": 103597799.0,
      "step": 46410
    },
    {
      "entropy": 5.329310655593872,
      "epoch": 5.21283764811591,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00026377105488099503,
      "loss": 4.8214,
      "mean_token_accuracy": 0.23575979918241502,
      "num_tokens": 103608570.0,
      "step": 46415
    },
    {
      "entropy": 5.323026275634765,
      "epoch": 5.213399225023867,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026373095637035037,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23358115255832673,
      "num_tokens": 103620339.0,
      "step": 46420
    },
    {
      "entropy": 5.23717885017395,
      "epoch": 5.213960801931824,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00026369085821851645,
      "loss": 4.7058,
      "mean_token_accuracy": 0.23941144496202468,
      "num_tokens": 103631471.0,
      "step": 46425
    },
    {
      "entropy": 5.292329883575439,
      "epoch": 5.214522378839782,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026365076042677,
      "loss": 4.7473,
      "mean_token_accuracy": 0.23796988427639007,
      "num_tokens": 103642223.0,
      "step": 46430
    },
    {
      "entropy": 5.304733848571777,
      "epoch": 5.2150839557477395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002636106629963877,
      "loss": 4.7541,
      "mean_token_accuracy": 0.24536751508712767,
      "num_tokens": 103653848.0,
      "step": 46435
    },
    {
      "entropy": 5.323637437820435,
      "epoch": 5.215645532655698,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002635705659286463,
      "loss": 4.8399,
      "mean_token_accuracy": 0.23941079080104827,
      "num_tokens": 103665765.0,
      "step": 46440
    },
    {
      "entropy": 5.397332382202149,
      "epoch": 5.216207109563655,
      "grad_norm": 1.0,
      "learning_rate": 0.00026353046922482256,
      "loss": 4.8786,
      "mean_token_accuracy": 0.2284968227148056,
      "num_tokens": 103678164.0,
      "step": 46445
    },
    {
      "entropy": 5.4142258644104,
      "epoch": 5.216768686471612,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002634903728861931,
      "loss": 4.8839,
      "mean_token_accuracy": 0.2282840356230736,
      "num_tokens": 103689274.0,
      "step": 46450
    },
    {
      "entropy": 5.455453681945801,
      "epoch": 5.21733026337957,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002634502769140346,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22781803458929062,
      "num_tokens": 103701272.0,
      "step": 46455
    },
    {
      "entropy": 5.356705188751221,
      "epoch": 5.217891840287527,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026341018130962384,
      "loss": 4.8762,
      "mean_token_accuracy": 0.22982213348150254,
      "num_tokens": 103711349.0,
      "step": 46460
    },
    {
      "entropy": 5.286958885192871,
      "epoch": 5.218453417195485,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026337008607423733,
      "loss": 4.7892,
      "mean_token_accuracy": 0.23468744158744811,
      "num_tokens": 103722518.0,
      "step": 46465
    },
    {
      "entropy": 5.309659576416015,
      "epoch": 5.219014994103443,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026332999120915184,
      "loss": 4.8219,
      "mean_token_accuracy": 0.23683672547340393,
      "num_tokens": 103732485.0,
      "step": 46470
    },
    {
      "entropy": 5.36706953048706,
      "epoch": 5.2195765710114,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026328989671564395,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22470946609973907,
      "num_tokens": 103742856.0,
      "step": 46475
    },
    {
      "entropy": 5.384916496276856,
      "epoch": 5.220138147919357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002632498025949902,
      "loss": 4.8734,
      "mean_token_accuracy": 0.23125288337469102,
      "num_tokens": 103755021.0,
      "step": 46480
    },
    {
      "entropy": 5.36115083694458,
      "epoch": 5.220699724827315,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002632097088484674,
      "loss": 4.8064,
      "mean_token_accuracy": 0.2420017257332802,
      "num_tokens": 103765856.0,
      "step": 46485
    },
    {
      "entropy": 5.359007692337036,
      "epoch": 5.2212613017352725,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026316961547735203,
      "loss": 4.7885,
      "mean_token_accuracy": 0.23553959429264068,
      "num_tokens": 103777771.0,
      "step": 46490
    },
    {
      "entropy": 5.2952796459198,
      "epoch": 5.221822878643231,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026312952248292067,
      "loss": 4.8112,
      "mean_token_accuracy": 0.23279588520526887,
      "num_tokens": 103789668.0,
      "step": 46495
    },
    {
      "entropy": 5.331391763687134,
      "epoch": 5.222384455551188,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002630894298664499,
      "loss": 4.9025,
      "mean_token_accuracy": 0.2360980376601219,
      "num_tokens": 103799946.0,
      "step": 46500
    },
    {
      "entropy": 5.323740386962891,
      "epoch": 5.222946032459145,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00026304933762921637,
      "loss": 4.7082,
      "mean_token_accuracy": 0.24122352600097657,
      "num_tokens": 103810311.0,
      "step": 46505
    },
    {
      "entropy": 5.321113109588623,
      "epoch": 5.223507609367103,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002630092457724965,
      "loss": 4.8101,
      "mean_token_accuracy": 0.23278290033340454,
      "num_tokens": 103821210.0,
      "step": 46510
    },
    {
      "entropy": 5.296548366546631,
      "epoch": 5.22406918627506,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002629691542975669,
      "loss": 4.7533,
      "mean_token_accuracy": 0.2430556982755661,
      "num_tokens": 103831684.0,
      "step": 46515
    },
    {
      "entropy": 5.259181356430053,
      "epoch": 5.224630763183018,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002629290632057041,
      "loss": 4.7021,
      "mean_token_accuracy": 0.2528118208050728,
      "num_tokens": 103842293.0,
      "step": 46520
    },
    {
      "entropy": 5.238279342651367,
      "epoch": 5.225192340090976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002628889724981846,
      "loss": 4.8109,
      "mean_token_accuracy": 0.23790747076272964,
      "num_tokens": 103854218.0,
      "step": 46525
    },
    {
      "entropy": 5.3169762134552006,
      "epoch": 5.225753916998933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002628488821762849,
      "loss": 4.9117,
      "mean_token_accuracy": 0.23030306547880172,
      "num_tokens": 103865162.0,
      "step": 46530
    },
    {
      "entropy": 5.34351487159729,
      "epoch": 5.22631549390689,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026280879224128165,
      "loss": 4.8713,
      "mean_token_accuracy": 0.24010601192712783,
      "num_tokens": 103876694.0,
      "step": 46535
    },
    {
      "entropy": 5.315881156921387,
      "epoch": 5.226877070814848,
      "grad_norm": 1.078125,
      "learning_rate": 0.000262768702694451,
      "loss": 4.7563,
      "mean_token_accuracy": 0.24081839621067047,
      "num_tokens": 103887665.0,
      "step": 46540
    },
    {
      "entropy": 5.288255214691162,
      "epoch": 5.227438647722805,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002627286135370697,
      "loss": 4.7921,
      "mean_token_accuracy": 0.23431711345911027,
      "num_tokens": 103899396.0,
      "step": 46545
    },
    {
      "entropy": 5.341542530059814,
      "epoch": 5.2280002246307635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000262688524770414,
      "loss": 4.8045,
      "mean_token_accuracy": 0.2354320302605629,
      "num_tokens": 103910592.0,
      "step": 46550
    },
    {
      "entropy": 5.371560525894165,
      "epoch": 5.228561801538721,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026264843639576055,
      "loss": 4.8883,
      "mean_token_accuracy": 0.23524882942438125,
      "num_tokens": 103922657.0,
      "step": 46555
    },
    {
      "entropy": 5.275856637954712,
      "epoch": 5.229123378446678,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002626083484143856,
      "loss": 4.7816,
      "mean_token_accuracy": 0.23905359953641891,
      "num_tokens": 103933129.0,
      "step": 46560
    },
    {
      "entropy": 5.384342288970947,
      "epoch": 5.229684955354636,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002625682608275657,
      "loss": 4.8901,
      "mean_token_accuracy": 0.23655917644500732,
      "num_tokens": 103945485.0,
      "step": 46565
    },
    {
      "entropy": 5.247740650177002,
      "epoch": 5.230246532262593,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002625281736365772,
      "loss": 4.7703,
      "mean_token_accuracy": 0.2390676721930504,
      "num_tokens": 103957465.0,
      "step": 46570
    },
    {
      "entropy": 5.293262672424317,
      "epoch": 5.2308081091705505,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026248808684269646,
      "loss": 4.8032,
      "mean_token_accuracy": 0.2360462635755539,
      "num_tokens": 103968410.0,
      "step": 46575
    },
    {
      "entropy": 5.363087368011475,
      "epoch": 5.231369686078509,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002624480004472,
      "loss": 4.8548,
      "mean_token_accuracy": 0.23272415101528168,
      "num_tokens": 103978923.0,
      "step": 46580
    },
    {
      "entropy": 5.300690031051635,
      "epoch": 5.231931262986466,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026240791445136384,
      "loss": 4.7161,
      "mean_token_accuracy": 0.23674419522285461,
      "num_tokens": 103989176.0,
      "step": 46585
    },
    {
      "entropy": 5.363605546951294,
      "epoch": 5.232492839894424,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002623678288564647,
      "loss": 4.8513,
      "mean_token_accuracy": 0.23388962000608443,
      "num_tokens": 104000172.0,
      "step": 46590
    },
    {
      "entropy": 5.253322124481201,
      "epoch": 5.233054416802381,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002623277436637788,
      "loss": 4.7761,
      "mean_token_accuracy": 0.2382298856973648,
      "num_tokens": 104011883.0,
      "step": 46595
    },
    {
      "entropy": 5.323957872390747,
      "epoch": 5.233615993710338,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002622876588745825,
      "loss": 4.8067,
      "mean_token_accuracy": 0.23712734133005142,
      "num_tokens": 104023607.0,
      "step": 46600
    },
    {
      "entropy": 5.371814918518067,
      "epoch": 5.2341775706182965,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000262247574490152,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23181650936603546,
      "num_tokens": 104036424.0,
      "step": 46605
    },
    {
      "entropy": 5.370618057250977,
      "epoch": 5.234739147526254,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002622074905117637,
      "loss": 4.933,
      "mean_token_accuracy": 0.22846785187721252,
      "num_tokens": 104048076.0,
      "step": 46610
    },
    {
      "entropy": 5.396787166595459,
      "epoch": 5.235300724434211,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026216740694069387,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22326239496469497,
      "num_tokens": 104058662.0,
      "step": 46615
    },
    {
      "entropy": 5.3344708442687985,
      "epoch": 5.235862301342169,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002621273237782188,
      "loss": 4.7263,
      "mean_token_accuracy": 0.24438222646713256,
      "num_tokens": 104069525.0,
      "step": 46620
    },
    {
      "entropy": 5.274775886535645,
      "epoch": 5.236423878250126,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002620872410256146,
      "loss": 4.732,
      "mean_token_accuracy": 0.24117178171873094,
      "num_tokens": 104080491.0,
      "step": 46625
    },
    {
      "entropy": 5.3018981456756595,
      "epoch": 5.236985455158084,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002620471586841577,
      "loss": 4.794,
      "mean_token_accuracy": 0.24336625188589095,
      "num_tokens": 104091723.0,
      "step": 46630
    },
    {
      "entropy": 5.289374828338623,
      "epoch": 5.2375470320660416,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002620070767551244,
      "loss": 4.7447,
      "mean_token_accuracy": 0.24563791304826738,
      "num_tokens": 104103554.0,
      "step": 46635
    },
    {
      "entropy": 5.384687519073486,
      "epoch": 5.238108608973999,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002619669952397908,
      "loss": 4.8332,
      "mean_token_accuracy": 0.23145875036716462,
      "num_tokens": 104115228.0,
      "step": 46640
    },
    {
      "entropy": 5.355109691619873,
      "epoch": 5.238670185881957,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002619269141394331,
      "loss": 4.8766,
      "mean_token_accuracy": 0.23172284066677093,
      "num_tokens": 104127550.0,
      "step": 46645
    },
    {
      "entropy": 5.28333592414856,
      "epoch": 5.239231762789914,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002618868334553275,
      "loss": 4.8091,
      "mean_token_accuracy": 0.23967068493366242,
      "num_tokens": 104138567.0,
      "step": 46650
    },
    {
      "entropy": 5.294527196884156,
      "epoch": 5.239793339697871,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026184675318875023,
      "loss": 4.8604,
      "mean_token_accuracy": 0.2364148750901222,
      "num_tokens": 104150237.0,
      "step": 46655
    },
    {
      "entropy": 5.367743825912475,
      "epoch": 5.240354916605829,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026180667334097733,
      "loss": 4.8518,
      "mean_token_accuracy": 0.23355712890625,
      "num_tokens": 104161461.0,
      "step": 46660
    },
    {
      "entropy": 5.34079942703247,
      "epoch": 5.240916493513787,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002617665939132852,
      "loss": 4.7593,
      "mean_token_accuracy": 0.23976379185914992,
      "num_tokens": 104172225.0,
      "step": 46665
    },
    {
      "entropy": 5.278589344024658,
      "epoch": 5.241478070421744,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002617265149069498,
      "loss": 4.7467,
      "mean_token_accuracy": 0.24236240088939667,
      "num_tokens": 104182926.0,
      "step": 46670
    },
    {
      "entropy": 5.3144152641296385,
      "epoch": 5.242039647329702,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002616864363232474,
      "loss": 4.83,
      "mean_token_accuracy": 0.2432252749800682,
      "num_tokens": 104194044.0,
      "step": 46675
    },
    {
      "entropy": 5.315785551071167,
      "epoch": 5.242601224237659,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026164635816345395,
      "loss": 4.8261,
      "mean_token_accuracy": 0.24319904446601867,
      "num_tokens": 104205383.0,
      "step": 46680
    },
    {
      "entropy": 5.280603456497192,
      "epoch": 5.243162801145617,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002616062804288457,
      "loss": 4.7885,
      "mean_token_accuracy": 0.23710711896419526,
      "num_tokens": 104217423.0,
      "step": 46685
    },
    {
      "entropy": 5.311478185653686,
      "epoch": 5.2437243780535745,
      "grad_norm": 1.0,
      "learning_rate": 0.0002615662031206986,
      "loss": 4.7963,
      "mean_token_accuracy": 0.23360835611820222,
      "num_tokens": 104228551.0,
      "step": 46690
    },
    {
      "entropy": 5.371574020385742,
      "epoch": 5.244285954961532,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002615261262402887,
      "loss": 4.9223,
      "mean_token_accuracy": 0.23019634336233138,
      "num_tokens": 104239778.0,
      "step": 46695
    },
    {
      "entropy": 5.345068740844726,
      "epoch": 5.24484753186949,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026148604978889237,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23160379976034165,
      "num_tokens": 104250863.0,
      "step": 46700
    },
    {
      "entropy": 5.30193018913269,
      "epoch": 5.245409108777447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026144597376778533,
      "loss": 4.7456,
      "mean_token_accuracy": 0.24856564551591873,
      "num_tokens": 104261458.0,
      "step": 46705
    },
    {
      "entropy": 5.2212282657623295,
      "epoch": 5.245970685685404,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002614058981782438,
      "loss": 4.7191,
      "mean_token_accuracy": 0.245540651679039,
      "num_tokens": 104272426.0,
      "step": 46710
    },
    {
      "entropy": 5.270497131347656,
      "epoch": 5.246532262593362,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026136582302154377,
      "loss": 4.758,
      "mean_token_accuracy": 0.23975466787815095,
      "num_tokens": 104283616.0,
      "step": 46715
    },
    {
      "entropy": 5.3058741092681885,
      "epoch": 5.24709383950132,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026132574829896123,
      "loss": 4.8654,
      "mean_token_accuracy": 0.2355233609676361,
      "num_tokens": 104294016.0,
      "step": 46720
    },
    {
      "entropy": 5.401514768600464,
      "epoch": 5.247655416409277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002612856740117721,
      "loss": 4.8544,
      "mean_token_accuracy": 0.23238602876663209,
      "num_tokens": 104305004.0,
      "step": 46725
    },
    {
      "entropy": 5.393391752243042,
      "epoch": 5.248216993317235,
      "grad_norm": 0.96875,
      "learning_rate": 0.00026124560016125236,
      "loss": 4.8697,
      "mean_token_accuracy": 0.23866243660449982,
      "num_tokens": 104316789.0,
      "step": 46730
    },
    {
      "entropy": 5.2945366382598875,
      "epoch": 5.248778570225192,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002612055267486782,
      "loss": 4.7316,
      "mean_token_accuracy": 0.23730792552232743,
      "num_tokens": 104327701.0,
      "step": 46735
    },
    {
      "entropy": 5.298704385757446,
      "epoch": 5.24934014713315,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00026116545377532536,
      "loss": 4.7621,
      "mean_token_accuracy": 0.23683555871248246,
      "num_tokens": 104339242.0,
      "step": 46740
    },
    {
      "entropy": 5.380912590026855,
      "epoch": 5.249901724041107,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002611253812424698,
      "loss": 4.9913,
      "mean_token_accuracy": 0.2234647735953331,
      "num_tokens": 104349090.0,
      "step": 46745
    },
    {
      "entropy": 5.274644947052002,
      "epoch": 5.250463300949065,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002610853091513876,
      "loss": 4.7368,
      "mean_token_accuracy": 0.2385093942284584,
      "num_tokens": 104360850.0,
      "step": 46750
    },
    {
      "entropy": 5.30753378868103,
      "epoch": 5.251024877857023,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026104523750335445,
      "loss": 4.808,
      "mean_token_accuracy": 0.2369964137673378,
      "num_tokens": 104371829.0,
      "step": 46755
    },
    {
      "entropy": 5.385022735595703,
      "epoch": 5.25158645476498,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002610051662996464,
      "loss": 4.8822,
      "mean_token_accuracy": 0.23535809963941573,
      "num_tokens": 104383209.0,
      "step": 46760
    },
    {
      "entropy": 5.283401870727539,
      "epoch": 5.252148031672937,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026096509554153914,
      "loss": 4.761,
      "mean_token_accuracy": 0.2404167488217354,
      "num_tokens": 104394069.0,
      "step": 46765
    },
    {
      "entropy": 5.38891863822937,
      "epoch": 5.252709608580895,
      "grad_norm": 1.0,
      "learning_rate": 0.0002609250252303088,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22257187962532043,
      "num_tokens": 104406056.0,
      "step": 46770
    },
    {
      "entropy": 5.29359245300293,
      "epoch": 5.2532711854888525,
      "grad_norm": 1.125,
      "learning_rate": 0.0002608849553672311,
      "loss": 4.8693,
      "mean_token_accuracy": 0.23242984861135482,
      "num_tokens": 104416600.0,
      "step": 46775
    },
    {
      "entropy": 5.3562380313873295,
      "epoch": 5.253832762396811,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026084488595358197,
      "loss": 4.8969,
      "mean_token_accuracy": 0.23494188636541366,
      "num_tokens": 104428307.0,
      "step": 46780
    },
    {
      "entropy": 5.308892059326172,
      "epoch": 5.254394339304768,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002608048169906371,
      "loss": 4.7882,
      "mean_token_accuracy": 0.23528470396995543,
      "num_tokens": 104439691.0,
      "step": 46785
    },
    {
      "entropy": 5.377576780319214,
      "epoch": 5.254955916212725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026076474847967236,
      "loss": 4.924,
      "mean_token_accuracy": 0.23308319896459578,
      "num_tokens": 104450456.0,
      "step": 46790
    },
    {
      "entropy": 5.351230955123901,
      "epoch": 5.255517493120683,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026072468042196356,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2264087527990341,
      "num_tokens": 104462057.0,
      "step": 46795
    },
    {
      "entropy": 5.428287172317505,
      "epoch": 5.25607907002864,
      "grad_norm": 1.0,
      "learning_rate": 0.00026068461281878645,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22710398435592652,
      "num_tokens": 104473998.0,
      "step": 46800
    },
    {
      "entropy": 5.3246265888214115,
      "epoch": 5.256640646936598,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00026064454567141683,
      "loss": 4.7367,
      "mean_token_accuracy": 0.23714830428361894,
      "num_tokens": 104485487.0,
      "step": 46805
    },
    {
      "entropy": 5.374052286148071,
      "epoch": 5.257202223844556,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002606044789811304,
      "loss": 4.8794,
      "mean_token_accuracy": 0.2339703157544136,
      "num_tokens": 104496968.0,
      "step": 46810
    },
    {
      "entropy": 5.359351968765258,
      "epoch": 5.257763800752513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000260564412749203,
      "loss": 4.8765,
      "mean_token_accuracy": 0.23233067840337754,
      "num_tokens": 104508448.0,
      "step": 46815
    },
    {
      "entropy": 5.307772827148438,
      "epoch": 5.258325377660471,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026052434697691035,
      "loss": 4.7878,
      "mean_token_accuracy": 0.23097856044769288,
      "num_tokens": 104519706.0,
      "step": 46820
    },
    {
      "entropy": 5.198577785491944,
      "epoch": 5.258886954568428,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000260484281665528,
      "loss": 4.6783,
      "mean_token_accuracy": 0.24580665230751036,
      "num_tokens": 104530277.0,
      "step": 46825
    },
    {
      "entropy": 5.387057733535767,
      "epoch": 5.2594485314763855,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002604442168163318,
      "loss": 4.855,
      "mean_token_accuracy": 0.2332157760858536,
      "num_tokens": 104541577.0,
      "step": 46830
    },
    {
      "entropy": 5.316790962219239,
      "epoch": 5.260010108384344,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002604041524305974,
      "loss": 4.8464,
      "mean_token_accuracy": 0.23067595809698105,
      "num_tokens": 104553185.0,
      "step": 46835
    },
    {
      "entropy": 5.348239898681641,
      "epoch": 5.260571685292301,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002603640885096004,
      "loss": 4.8185,
      "mean_token_accuracy": 0.24327073693275453,
      "num_tokens": 104565221.0,
      "step": 46840
    },
    {
      "entropy": 5.233686351776123,
      "epoch": 5.261133262200258,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002603240250546165,
      "loss": 4.6677,
      "mean_token_accuracy": 0.24404096603393555,
      "num_tokens": 104577293.0,
      "step": 46845
    },
    {
      "entropy": 5.2941466808319095,
      "epoch": 5.261694839108216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026028396206692134,
      "loss": 4.7668,
      "mean_token_accuracy": 0.238493849337101,
      "num_tokens": 104587894.0,
      "step": 46850
    },
    {
      "entropy": 5.30857629776001,
      "epoch": 5.262256416016173,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026024389954779054,
      "loss": 4.824,
      "mean_token_accuracy": 0.23412288427352906,
      "num_tokens": 104598396.0,
      "step": 46855
    },
    {
      "entropy": 5.344823169708252,
      "epoch": 5.2628179929241306,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002602038374984997,
      "loss": 4.9095,
      "mean_token_accuracy": 0.23141019493341447,
      "num_tokens": 104608554.0,
      "step": 46860
    },
    {
      "entropy": 5.309149360656738,
      "epoch": 5.263379569832089,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026016377592032446,
      "loss": 4.8105,
      "mean_token_accuracy": 0.23867145776748658,
      "num_tokens": 104619562.0,
      "step": 46865
    },
    {
      "entropy": 5.28914852142334,
      "epoch": 5.263941146740046,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002601237148145403,
      "loss": 4.8132,
      "mean_token_accuracy": 0.23377913981676102,
      "num_tokens": 104629641.0,
      "step": 46870
    },
    {
      "entropy": 5.298944711685181,
      "epoch": 5.264502723648004,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026008365418242284,
      "loss": 4.7848,
      "mean_token_accuracy": 0.23894418478012086,
      "num_tokens": 104640303.0,
      "step": 46875
    },
    {
      "entropy": 5.307670927047729,
      "epoch": 5.265064300555961,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002600435940252476,
      "loss": 4.7557,
      "mean_token_accuracy": 0.23939779549837112,
      "num_tokens": 104650761.0,
      "step": 46880
    },
    {
      "entropy": 5.236644697189331,
      "epoch": 5.265625877463918,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002600035343442902,
      "loss": 4.6671,
      "mean_token_accuracy": 0.2504496738314629,
      "num_tokens": 104662159.0,
      "step": 46885
    },
    {
      "entropy": 5.297658109664917,
      "epoch": 5.2661874543718765,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025996347514082603,
      "loss": 4.7971,
      "mean_token_accuracy": 0.23789357095956803,
      "num_tokens": 104674209.0,
      "step": 46890
    },
    {
      "entropy": 5.300919008255005,
      "epoch": 5.266749031279834,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00025992341641613065,
      "loss": 4.7925,
      "mean_token_accuracy": 0.24119152426719664,
      "num_tokens": 104685315.0,
      "step": 46895
    },
    {
      "entropy": 5.378247880935669,
      "epoch": 5.267310608187791,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002598833581714795,
      "loss": 4.8454,
      "mean_token_accuracy": 0.2383030742406845,
      "num_tokens": 104697220.0,
      "step": 46900
    },
    {
      "entropy": 5.388354539871216,
      "epoch": 5.267872185095749,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025984330040814817,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23798543512821196,
      "num_tokens": 104709156.0,
      "step": 46905
    },
    {
      "entropy": 5.33437876701355,
      "epoch": 5.268433762003706,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002598032431274119,
      "loss": 4.805,
      "mean_token_accuracy": 0.23972556740045547,
      "num_tokens": 104721160.0,
      "step": 46910
    },
    {
      "entropy": 5.289506530761718,
      "epoch": 5.2689953389116635,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025976318633054636,
      "loss": 4.7914,
      "mean_token_accuracy": 0.2364952802658081,
      "num_tokens": 104732263.0,
      "step": 46915
    },
    {
      "entropy": 5.39394736289978,
      "epoch": 5.269556915819622,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002597231300188269,
      "loss": 4.8597,
      "mean_token_accuracy": 0.22910289019346236,
      "num_tokens": 104743756.0,
      "step": 46920
    },
    {
      "entropy": 5.3395710468292235,
      "epoch": 5.270118492727579,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00025968307419352884,
      "loss": 4.8609,
      "mean_token_accuracy": 0.2330589085817337,
      "num_tokens": 104754272.0,
      "step": 46925
    },
    {
      "entropy": 5.382916641235352,
      "epoch": 5.270680069635537,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025964301885592766,
      "loss": 4.9498,
      "mean_token_accuracy": 0.2290986582636833,
      "num_tokens": 104766366.0,
      "step": 46930
    },
    {
      "entropy": 5.37900743484497,
      "epoch": 5.271241646543494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002596029640072988,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22935412228107452,
      "num_tokens": 104778158.0,
      "step": 46935
    },
    {
      "entropy": 5.350668239593506,
      "epoch": 5.271803223451451,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002595629096489174,
      "loss": 4.8384,
      "mean_token_accuracy": 0.2398675262928009,
      "num_tokens": 104789759.0,
      "step": 46940
    },
    {
      "entropy": 5.3045759201049805,
      "epoch": 5.2723648003594095,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000259522855782059,
      "loss": 4.6971,
      "mean_token_accuracy": 0.25003596395254135,
      "num_tokens": 104800597.0,
      "step": 46945
    },
    {
      "entropy": 5.2118244647979735,
      "epoch": 5.272926377267367,
      "grad_norm": 1.0,
      "learning_rate": 0.0002594828024079989,
      "loss": 4.7441,
      "mean_token_accuracy": 0.24213664680719377,
      "num_tokens": 104810833.0,
      "step": 46950
    },
    {
      "entropy": 5.299561166763306,
      "epoch": 5.273487954175324,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025944274952801233,
      "loss": 4.7886,
      "mean_token_accuracy": 0.2428634449839592,
      "num_tokens": 104822015.0,
      "step": 46955
    },
    {
      "entropy": 5.388574838638306,
      "epoch": 5.274049531083282,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002594026971433747,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22842651158571242,
      "num_tokens": 104832462.0,
      "step": 46960
    },
    {
      "entropy": 5.404599046707153,
      "epoch": 5.274611107991239,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00025936264525536126,
      "loss": 4.8337,
      "mean_token_accuracy": 0.22951748073101044,
      "num_tokens": 104844386.0,
      "step": 46965
    },
    {
      "entropy": 5.374642276763916,
      "epoch": 5.275172684899197,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002593225938652472,
      "loss": 4.8546,
      "mean_token_accuracy": 0.228460593521595,
      "num_tokens": 104855569.0,
      "step": 46970
    },
    {
      "entropy": 5.3350084781646725,
      "epoch": 5.2757342618071545,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002592825429743079,
      "loss": 4.8374,
      "mean_token_accuracy": 0.23182704746723176,
      "num_tokens": 104866721.0,
      "step": 46975
    },
    {
      "entropy": 5.251456308364868,
      "epoch": 5.276295838715112,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002592424925838185,
      "loss": 4.7504,
      "mean_token_accuracy": 0.237536258995533,
      "num_tokens": 104877449.0,
      "step": 46980
    },
    {
      "entropy": 5.292297840118408,
      "epoch": 5.27685741562307,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002592024426950541,
      "loss": 4.7773,
      "mean_token_accuracy": 0.23880802392959594,
      "num_tokens": 104889050.0,
      "step": 46985
    },
    {
      "entropy": 5.340652132034302,
      "epoch": 5.277418992531027,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00025916239330929026,
      "loss": 4.8431,
      "mean_token_accuracy": 0.23764988780021667,
      "num_tokens": 104900231.0,
      "step": 46990
    },
    {
      "entropy": 5.434039735794068,
      "epoch": 5.277980569438984,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00025912234442780185,
      "loss": 4.9635,
      "mean_token_accuracy": 0.2242128074169159,
      "num_tokens": 104912886.0,
      "step": 46995
    },
    {
      "entropy": 5.418157291412354,
      "epoch": 5.278542146346942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025908229605186423,
      "loss": 4.8738,
      "mean_token_accuracy": 0.23258846551179885,
      "num_tokens": 104924177.0,
      "step": 47000
    },
    {
      "entropy": 5.4069623947143555,
      "epoch": 5.2791037232549,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025904224818275247,
      "loss": 4.8791,
      "mean_token_accuracy": 0.23878918588161469,
      "num_tokens": 104934503.0,
      "step": 47005
    },
    {
      "entropy": 5.3969005107879635,
      "epoch": 5.279665300162858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025900220082174165,
      "loss": 4.87,
      "mean_token_accuracy": 0.2260451063513756,
      "num_tokens": 104945643.0,
      "step": 47010
    },
    {
      "entropy": 5.315466213226318,
      "epoch": 5.280226877070815,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00025896215397010696,
      "loss": 4.802,
      "mean_token_accuracy": 0.22714556455612184,
      "num_tokens": 104957289.0,
      "step": 47015
    },
    {
      "entropy": 5.345183324813843,
      "epoch": 5.280788453978772,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002589221076291236,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2266632243990898,
      "num_tokens": 104967848.0,
      "step": 47020
    },
    {
      "entropy": 5.293126964569092,
      "epoch": 5.28135003088673,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025888206180006647,
      "loss": 4.7529,
      "mean_token_accuracy": 0.23837372213602065,
      "num_tokens": 104979036.0,
      "step": 47025
    },
    {
      "entropy": 5.330266904830933,
      "epoch": 5.2819116077946875,
      "grad_norm": 1.0,
      "learning_rate": 0.0002588420164842108,
      "loss": 4.7822,
      "mean_token_accuracy": 0.23795006722211837,
      "num_tokens": 104990399.0,
      "step": 47030
    },
    {
      "entropy": 5.344455337524414,
      "epoch": 5.282473184702645,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002588019716828316,
      "loss": 4.8346,
      "mean_token_accuracy": 0.24932210445404052,
      "num_tokens": 105001515.0,
      "step": 47035
    },
    {
      "entropy": 5.352672863006592,
      "epoch": 5.283034761610603,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025876192739720383,
      "loss": 4.8579,
      "mean_token_accuracy": 0.2314743235707283,
      "num_tokens": 105012609.0,
      "step": 47040
    },
    {
      "entropy": 5.413086318969727,
      "epoch": 5.28359633851856,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00025872188362860264,
      "loss": 4.8329,
      "mean_token_accuracy": 0.23158431351184844,
      "num_tokens": 105022120.0,
      "step": 47045
    },
    {
      "entropy": 5.343380641937256,
      "epoch": 5.284157915426517,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025868184037830305,
      "loss": 4.8282,
      "mean_token_accuracy": 0.23083868473768235,
      "num_tokens": 105033121.0,
      "step": 47050
    },
    {
      "entropy": 5.420180368423462,
      "epoch": 5.284719492334475,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002586417976475798,
      "loss": 4.7946,
      "mean_token_accuracy": 0.24607355296611785,
      "num_tokens": 105043279.0,
      "step": 47055
    },
    {
      "entropy": 5.254603338241577,
      "epoch": 5.285281069242433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025860175543770817,
      "loss": 4.7479,
      "mean_token_accuracy": 0.24171694815158845,
      "num_tokens": 105055252.0,
      "step": 47060
    },
    {
      "entropy": 5.347330522537232,
      "epoch": 5.285842646150391,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002585617137499631,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22596484571695327,
      "num_tokens": 105067292.0,
      "step": 47065
    },
    {
      "entropy": 5.4201775074005125,
      "epoch": 5.286404223058348,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002585216725856193,
      "loss": 4.8712,
      "mean_token_accuracy": 0.24164705127477645,
      "num_tokens": 105078125.0,
      "step": 47070
    },
    {
      "entropy": 5.374434614181519,
      "epoch": 5.286965799966305,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002584816319459519,
      "loss": 4.8611,
      "mean_token_accuracy": 0.23317748457193374,
      "num_tokens": 105090398.0,
      "step": 47075
    },
    {
      "entropy": 5.348397541046142,
      "epoch": 5.287527376874263,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002584415918322358,
      "loss": 4.8413,
      "mean_token_accuracy": 0.23167410641908645,
      "num_tokens": 105102182.0,
      "step": 47080
    },
    {
      "entropy": 5.2998247146606445,
      "epoch": 5.28808895378222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002584015522457458,
      "loss": 4.7106,
      "mean_token_accuracy": 0.2445102870464325,
      "num_tokens": 105112361.0,
      "step": 47085
    },
    {
      "entropy": 5.188279914855957,
      "epoch": 5.288650530690178,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00025836151318775675,
      "loss": 4.6867,
      "mean_token_accuracy": 0.24998184591531752,
      "num_tokens": 105124479.0,
      "step": 47090
    },
    {
      "entropy": 5.293795251846314,
      "epoch": 5.289212107598136,
      "grad_norm": 1.171875,
      "learning_rate": 0.00025832147465954365,
      "loss": 4.8175,
      "mean_token_accuracy": 0.24056858718395233,
      "num_tokens": 105136387.0,
      "step": 47095
    },
    {
      "entropy": 5.3327188968658445,
      "epoch": 5.289773684506093,
      "grad_norm": 1.0,
      "learning_rate": 0.00025828143666238124,
      "loss": 4.8186,
      "mean_token_accuracy": 0.23633965402841567,
      "num_tokens": 105148768.0,
      "step": 47100
    },
    {
      "entropy": 5.329674196243286,
      "epoch": 5.29033526141405,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00025824139919754436,
      "loss": 4.8328,
      "mean_token_accuracy": 0.23990897238254547,
      "num_tokens": 105161500.0,
      "step": 47105
    },
    {
      "entropy": 5.373843383789063,
      "epoch": 5.290896838322008,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002582013622663079,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23526697903871535,
      "num_tokens": 105173258.0,
      "step": 47110
    },
    {
      "entropy": 5.405191564559937,
      "epoch": 5.2914584152299655,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002581613258699465,
      "loss": 4.9056,
      "mean_token_accuracy": 0.23541885018348693,
      "num_tokens": 105184710.0,
      "step": 47115
    },
    {
      "entropy": 5.301182794570923,
      "epoch": 5.292019992137924,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000258121290009735,
      "loss": 4.7752,
      "mean_token_accuracy": 0.23481546640396117,
      "num_tokens": 105194804.0,
      "step": 47120
    },
    {
      "entropy": 5.3456045627594,
      "epoch": 5.292581569045881,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002580812546869481,
      "loss": 4.911,
      "mean_token_accuracy": 0.22760363221168517,
      "num_tokens": 105207074.0,
      "step": 47125
    },
    {
      "entropy": 5.380125284194946,
      "epoch": 5.293143145953838,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002580412199028607,
      "loss": 4.8342,
      "mean_token_accuracy": 0.23449423611164094,
      "num_tokens": 105218007.0,
      "step": 47130
    },
    {
      "entropy": 5.277137994766235,
      "epoch": 5.293704722861796,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025800118565874754,
      "loss": 4.7542,
      "mean_token_accuracy": 0.2402175024151802,
      "num_tokens": 105229299.0,
      "step": 47135
    },
    {
      "entropy": 5.322787237167359,
      "epoch": 5.294266299769753,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002579611519558831,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23408448249101638,
      "num_tokens": 105239482.0,
      "step": 47140
    },
    {
      "entropy": 5.373965549468994,
      "epoch": 5.294827876677711,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00025792111879554224,
      "loss": 4.891,
      "mean_token_accuracy": 0.22844665944576265,
      "num_tokens": 105250185.0,
      "step": 47145
    },
    {
      "entropy": 5.272731828689575,
      "epoch": 5.295389453585669,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002578810861789996,
      "loss": 4.6777,
      "mean_token_accuracy": 0.2420856773853302,
      "num_tokens": 105262992.0,
      "step": 47150
    },
    {
      "entropy": 5.377058315277099,
      "epoch": 5.295951030493626,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002578410541075298,
      "loss": 4.9181,
      "mean_token_accuracy": 0.23352336287498474,
      "num_tokens": 105274265.0,
      "step": 47155
    },
    {
      "entropy": 5.252666473388672,
      "epoch": 5.296512607401584,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00025780102258240745,
      "loss": 4.8366,
      "mean_token_accuracy": 0.23724357336759566,
      "num_tokens": 105285517.0,
      "step": 47160
    },
    {
      "entropy": 5.42457799911499,
      "epoch": 5.297074184309541,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002577609916049072,
      "loss": 4.8442,
      "mean_token_accuracy": 0.22679315656423568,
      "num_tokens": 105296052.0,
      "step": 47165
    },
    {
      "entropy": 5.413437175750732,
      "epoch": 5.2976357612174985,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002577209611763037,
      "loss": 4.8708,
      "mean_token_accuracy": 0.23428469002246857,
      "num_tokens": 105307728.0,
      "step": 47170
    },
    {
      "entropy": 5.381452226638794,
      "epoch": 5.298197338125457,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002576809312978715,
      "loss": 4.8732,
      "mean_token_accuracy": 0.22874194532632827,
      "num_tokens": 105319750.0,
      "step": 47175
    },
    {
      "entropy": 5.368034982681275,
      "epoch": 5.298758915033414,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002576409019708852,
      "loss": 4.8127,
      "mean_token_accuracy": 0.23716093003749847,
      "num_tokens": 105330457.0,
      "step": 47180
    },
    {
      "entropy": 5.290163850784301,
      "epoch": 5.299320491941371,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002576008731966193,
      "loss": 4.7443,
      "mean_token_accuracy": 0.2459518164396286,
      "num_tokens": 105341265.0,
      "step": 47185
    },
    {
      "entropy": 5.372845792770386,
      "epoch": 5.299882068849329,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025756084497634827,
      "loss": 4.9863,
      "mean_token_accuracy": 0.23329026401042938,
      "num_tokens": 105352353.0,
      "step": 47190
    },
    {
      "entropy": 5.360569477081299,
      "epoch": 5.300443645757286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002575208173113468,
      "loss": 4.814,
      "mean_token_accuracy": 0.2354424849152565,
      "num_tokens": 105363785.0,
      "step": 47195
    },
    {
      "entropy": 5.325840950012207,
      "epoch": 5.301005222665244,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002574807902028891,
      "loss": 4.7756,
      "mean_token_accuracy": 0.23452985882759095,
      "num_tokens": 105374185.0,
      "step": 47200
    },
    {
      "entropy": 5.281309938430786,
      "epoch": 5.301566799573202,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025744076365225,
      "loss": 4.7842,
      "mean_token_accuracy": 0.24457790851593017,
      "num_tokens": 105385698.0,
      "step": 47205
    },
    {
      "entropy": 5.220132160186767,
      "epoch": 5.302128376481159,
      "grad_norm": 0.953125,
      "learning_rate": 0.00025740073766070376,
      "loss": 4.7635,
      "mean_token_accuracy": 0.23649675846099855,
      "num_tokens": 105396630.0,
      "step": 47210
    },
    {
      "entropy": 5.29799861907959,
      "epoch": 5.302689953389117,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002573607122295249,
      "loss": 4.7485,
      "mean_token_accuracy": 0.24004548639059067,
      "num_tokens": 105408728.0,
      "step": 47215
    },
    {
      "entropy": 5.315957021713257,
      "epoch": 5.303251530297074,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002573206873599878,
      "loss": 4.7802,
      "mean_token_accuracy": 0.24149588495492935,
      "num_tokens": 105418545.0,
      "step": 47220
    },
    {
      "entropy": 5.348516035079956,
      "epoch": 5.303813107205031,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025728066305336685,
      "loss": 4.8706,
      "mean_token_accuracy": 0.229243466258049,
      "num_tokens": 105429748.0,
      "step": 47225
    },
    {
      "entropy": 5.275116491317749,
      "epoch": 5.3043746841129895,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002572406393109364,
      "loss": 4.7572,
      "mean_token_accuracy": 0.23763249218463897,
      "num_tokens": 105441298.0,
      "step": 47230
    },
    {
      "entropy": 5.377463293075562,
      "epoch": 5.304936261020947,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002572006161339709,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2384906753897667,
      "num_tokens": 105452511.0,
      "step": 47235
    },
    {
      "entropy": 5.328317594528198,
      "epoch": 5.305497837928904,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002571605935237448,
      "loss": 4.7579,
      "mean_token_accuracy": 0.24559869319200517,
      "num_tokens": 105464476.0,
      "step": 47240
    },
    {
      "entropy": 5.324857759475708,
      "epoch": 5.306059414836862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002571205714815322,
      "loss": 4.8834,
      "mean_token_accuracy": 0.23934174031019212,
      "num_tokens": 105475761.0,
      "step": 47245
    },
    {
      "entropy": 5.325757312774658,
      "epoch": 5.306620991744819,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002570805500086076,
      "loss": 4.8275,
      "mean_token_accuracy": 0.22887710481882095,
      "num_tokens": 105488162.0,
      "step": 47250
    },
    {
      "entropy": 5.318894481658935,
      "epoch": 5.307182568652777,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002570405291062452,
      "loss": 4.7664,
      "mean_token_accuracy": 0.23660466820001602,
      "num_tokens": 105498879.0,
      "step": 47255
    },
    {
      "entropy": 5.388209629058838,
      "epoch": 5.307744145560735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002570005087757193,
      "loss": 4.83,
      "mean_token_accuracy": 0.23380740135908126,
      "num_tokens": 105509052.0,
      "step": 47260
    },
    {
      "entropy": 5.359475755691529,
      "epoch": 5.308305722468692,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002569604890183043,
      "loss": 4.8615,
      "mean_token_accuracy": 0.23172544836997985,
      "num_tokens": 105520970.0,
      "step": 47265
    },
    {
      "entropy": 5.308510208129883,
      "epoch": 5.30886729937665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002569204698352741,
      "loss": 4.7786,
      "mean_token_accuracy": 0.24239733070135117,
      "num_tokens": 105532172.0,
      "step": 47270
    },
    {
      "entropy": 5.278017997741699,
      "epoch": 5.309428876284607,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002568804512279033,
      "loss": 4.7659,
      "mean_token_accuracy": 0.24613500237464905,
      "num_tokens": 105542849.0,
      "step": 47275
    },
    {
      "entropy": 5.316369724273682,
      "epoch": 5.309990453192564,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00025684043319746595,
      "loss": 4.8001,
      "mean_token_accuracy": 0.23683156669139863,
      "num_tokens": 105553488.0,
      "step": 47280
    },
    {
      "entropy": 5.259151983261108,
      "epoch": 5.3105520301005225,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025680041574523626,
      "loss": 4.7436,
      "mean_token_accuracy": 0.2340954691171646,
      "num_tokens": 105563900.0,
      "step": 47285
    },
    {
      "entropy": 5.270215034484863,
      "epoch": 5.31111360700848,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002567603988724884,
      "loss": 4.7269,
      "mean_token_accuracy": 0.24443728774785994,
      "num_tokens": 105575192.0,
      "step": 47290
    },
    {
      "entropy": 5.332947731018066,
      "epoch": 5.311675183916437,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002567203825804966,
      "loss": 4.8048,
      "mean_token_accuracy": 0.2414331093430519,
      "num_tokens": 105586614.0,
      "step": 47295
    },
    {
      "entropy": 5.360145473480225,
      "epoch": 5.312236760824395,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002566803668705347,
      "loss": 4.8065,
      "mean_token_accuracy": 0.23989985287189483,
      "num_tokens": 105597728.0,
      "step": 47300
    },
    {
      "entropy": 5.345130252838135,
      "epoch": 5.312798337732352,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002566403517438772,
      "loss": 4.9107,
      "mean_token_accuracy": 0.229011732339859,
      "num_tokens": 105608694.0,
      "step": 47305
    },
    {
      "entropy": 5.316271114349365,
      "epoch": 5.31335991464031,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025660033720179795,
      "loss": 4.7674,
      "mean_token_accuracy": 0.24549946039915085,
      "num_tokens": 105620337.0,
      "step": 47310
    },
    {
      "entropy": 5.307069492340088,
      "epoch": 5.3139214915482675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025656032324557107,
      "loss": 4.7472,
      "mean_token_accuracy": 0.24895475953817367,
      "num_tokens": 105630463.0,
      "step": 47315
    },
    {
      "entropy": 5.368904685974121,
      "epoch": 5.314483068456225,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025652030987647066,
      "loss": 4.9375,
      "mean_token_accuracy": 0.23399686217308044,
      "num_tokens": 105642236.0,
      "step": 47320
    },
    {
      "entropy": 5.337502098083496,
      "epoch": 5.315044645364183,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002564802970957708,
      "loss": 4.8205,
      "mean_token_accuracy": 0.23608140498399735,
      "num_tokens": 105652513.0,
      "step": 47325
    },
    {
      "entropy": 5.326572465896606,
      "epoch": 5.31560622227214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025644028490474544,
      "loss": 4.7884,
      "mean_token_accuracy": 0.24545741230249404,
      "num_tokens": 105665532.0,
      "step": 47330
    },
    {
      "entropy": 5.348112392425537,
      "epoch": 5.316167799180098,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025640027330466866,
      "loss": 4.8205,
      "mean_token_accuracy": 0.23306269496679305,
      "num_tokens": 105677826.0,
      "step": 47335
    },
    {
      "entropy": 5.282850027084351,
      "epoch": 5.316729376088055,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025636026229681433,
      "loss": 4.7935,
      "mean_token_accuracy": 0.2322593629360199,
      "num_tokens": 105689428.0,
      "step": 47340
    },
    {
      "entropy": 5.323724174499512,
      "epoch": 5.317290952996013,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002563202518824564,
      "loss": 4.7607,
      "mean_token_accuracy": 0.24096730798482896,
      "num_tokens": 105701423.0,
      "step": 47345
    },
    {
      "entropy": 5.250507354736328,
      "epoch": 5.317852529903971,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000256280242062869,
      "loss": 4.7599,
      "mean_token_accuracy": 0.2316095232963562,
      "num_tokens": 105712985.0,
      "step": 47350
    },
    {
      "entropy": 5.331543493270874,
      "epoch": 5.318414106811928,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002562402328393259,
      "loss": 4.8675,
      "mean_token_accuracy": 0.22923032194375992,
      "num_tokens": 105724283.0,
      "step": 47355
    },
    {
      "entropy": 5.3581830024719235,
      "epoch": 5.318975683719885,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00025620022421310115,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2237453803420067,
      "num_tokens": 105735250.0,
      "step": 47360
    },
    {
      "entropy": 5.396858167648316,
      "epoch": 5.319537260627843,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025616021618546847,
      "loss": 4.8303,
      "mean_token_accuracy": 0.23583883196115493,
      "num_tokens": 105747093.0,
      "step": 47365
    },
    {
      "entropy": 5.375815296173096,
      "epoch": 5.3200988375358005,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025612020875770186,
      "loss": 4.856,
      "mean_token_accuracy": 0.23857094943523408,
      "num_tokens": 105757906.0,
      "step": 47370
    },
    {
      "entropy": 5.307946443557739,
      "epoch": 5.320660414443758,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002560802019310751,
      "loss": 4.7593,
      "mean_token_accuracy": 0.24705451875925064,
      "num_tokens": 105768618.0,
      "step": 47375
    },
    {
      "entropy": 5.263268661499024,
      "epoch": 5.321221991351716,
      "grad_norm": 1.125,
      "learning_rate": 0.00025604019570686203,
      "loss": 4.759,
      "mean_token_accuracy": 0.24161226600408553,
      "num_tokens": 105779201.0,
      "step": 47380
    },
    {
      "entropy": 5.355069160461426,
      "epoch": 5.321783568259673,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002560001900863365,
      "loss": 4.7995,
      "mean_token_accuracy": 0.24010185450315474,
      "num_tokens": 105789796.0,
      "step": 47385
    },
    {
      "entropy": 5.36286883354187,
      "epoch": 5.322345145167631,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002559601850707723,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2379504844546318,
      "num_tokens": 105801576.0,
      "step": 47390
    },
    {
      "entropy": 5.320552635192871,
      "epoch": 5.322906722075588,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002559201806614432,
      "loss": 4.7908,
      "mean_token_accuracy": 0.23955017179250718,
      "num_tokens": 105812836.0,
      "step": 47395
    },
    {
      "entropy": 5.491349220275879,
      "epoch": 5.323468298983546,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025588017685962297,
      "loss": 4.9896,
      "mean_token_accuracy": 0.23090129792690278,
      "num_tokens": 105824408.0,
      "step": 47400
    },
    {
      "entropy": 5.2657098293304445,
      "epoch": 5.324029875891504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025584017366658525,
      "loss": 4.7211,
      "mean_token_accuracy": 0.24716109037399292,
      "num_tokens": 105834640.0,
      "step": 47405
    },
    {
      "entropy": 5.292738628387451,
      "epoch": 5.324591452799461,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025580017108360386,
      "loss": 4.7759,
      "mean_token_accuracy": 0.2348622277379036,
      "num_tokens": 105845941.0,
      "step": 47410
    },
    {
      "entropy": 5.321476125717163,
      "epoch": 5.325153029707418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002557601691119524,
      "loss": 4.8751,
      "mean_token_accuracy": 0.22688245475292207,
      "num_tokens": 105857025.0,
      "step": 47415
    },
    {
      "entropy": 5.338246488571167,
      "epoch": 5.325714606615376,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00025572016775290475,
      "loss": 4.8577,
      "mean_token_accuracy": 0.23243886083364487,
      "num_tokens": 105867501.0,
      "step": 47420
    },
    {
      "entropy": 5.2853399276733395,
      "epoch": 5.326276183523333,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025568016700773435,
      "loss": 4.7602,
      "mean_token_accuracy": 0.23947880566120147,
      "num_tokens": 105878153.0,
      "step": 47425
    },
    {
      "entropy": 5.30452790260315,
      "epoch": 5.326837760431291,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025564016687771495,
      "loss": 4.7131,
      "mean_token_accuracy": 0.24361384212970733,
      "num_tokens": 105888770.0,
      "step": 47430
    },
    {
      "entropy": 5.265668106079102,
      "epoch": 5.327399337339249,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00025560016736412015,
      "loss": 4.7354,
      "mean_token_accuracy": 0.24482796490192413,
      "num_tokens": 105899773.0,
      "step": 47435
    },
    {
      "entropy": 5.296401309967041,
      "epoch": 5.327960914247206,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025556016846822353,
      "loss": 4.8334,
      "mean_token_accuracy": 0.2437345027923584,
      "num_tokens": 105912608.0,
      "step": 47440
    },
    {
      "entropy": 5.328469848632812,
      "epoch": 5.328522491155164,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025552017019129874,
      "loss": 4.7935,
      "mean_token_accuracy": 0.2379808619618416,
      "num_tokens": 105922781.0,
      "step": 47445
    },
    {
      "entropy": 5.321211910247802,
      "epoch": 5.329084068063121,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025548017253461923,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23879847675561905,
      "num_tokens": 105934641.0,
      "step": 47450
    },
    {
      "entropy": 5.35409631729126,
      "epoch": 5.3296456449710785,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002554401754994586,
      "loss": 4.8761,
      "mean_token_accuracy": 0.23282630443573,
      "num_tokens": 105946722.0,
      "step": 47455
    },
    {
      "entropy": 5.311547946929932,
      "epoch": 5.330207221879037,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002554001790870904,
      "loss": 4.7977,
      "mean_token_accuracy": 0.23901815563440323,
      "num_tokens": 105957361.0,
      "step": 47460
    },
    {
      "entropy": 5.306861352920532,
      "epoch": 5.330768798786994,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025536018329878814,
      "loss": 4.7695,
      "mean_token_accuracy": 0.23651579320430755,
      "num_tokens": 105968314.0,
      "step": 47465
    },
    {
      "entropy": 5.32871527671814,
      "epoch": 5.331330375694951,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002553201881358252,
      "loss": 4.838,
      "mean_token_accuracy": 0.23680963963270188,
      "num_tokens": 105979324.0,
      "step": 47470
    },
    {
      "entropy": 5.335201454162598,
      "epoch": 5.331891952602909,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025528019359947514,
      "loss": 4.7926,
      "mean_token_accuracy": 0.23940346240997315,
      "num_tokens": 105989998.0,
      "step": 47475
    },
    {
      "entropy": 5.327815246582031,
      "epoch": 5.332453529510866,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002552401996910113,
      "loss": 4.8271,
      "mean_token_accuracy": 0.2347569152712822,
      "num_tokens": 106000389.0,
      "step": 47480
    },
    {
      "entropy": 5.368082475662232,
      "epoch": 5.333015106418824,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002552002064117072,
      "loss": 4.8703,
      "mean_token_accuracy": 0.2318431705236435,
      "num_tokens": 106010425.0,
      "step": 47485
    },
    {
      "entropy": 5.365311861038208,
      "epoch": 5.333576683326782,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002551602137628361,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2352878451347351,
      "num_tokens": 106022354.0,
      "step": 47490
    },
    {
      "entropy": 5.365330839157105,
      "epoch": 5.334138260234739,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025512022174567164,
      "loss": 4.8204,
      "mean_token_accuracy": 0.23885389119386674,
      "num_tokens": 106033651.0,
      "step": 47495
    },
    {
      "entropy": 5.261027050018311,
      "epoch": 5.334699837142697,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00025508023036148697,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23703691661357879,
      "num_tokens": 106046049.0,
      "step": 47500
    },
    {
      "entropy": 5.277173376083374,
      "epoch": 5.335261414050654,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025504023961155553,
      "loss": 4.7857,
      "mean_token_accuracy": 0.23693223744630815,
      "num_tokens": 106058196.0,
      "step": 47505
    },
    {
      "entropy": 5.312263774871826,
      "epoch": 5.3358229909586115,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025500024949715055,
      "loss": 4.7877,
      "mean_token_accuracy": 0.23667695224285126,
      "num_tokens": 106069553.0,
      "step": 47510
    },
    {
      "entropy": 5.358439588546753,
      "epoch": 5.33638456786657,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025496026001954543,
      "loss": 4.8013,
      "mean_token_accuracy": 0.2366901382803917,
      "num_tokens": 106080875.0,
      "step": 47515
    },
    {
      "entropy": 5.35123643875122,
      "epoch": 5.336946144774527,
      "grad_norm": 1.1875,
      "learning_rate": 0.00025492027118001345,
      "loss": 4.8664,
      "mean_token_accuracy": 0.23522239774465561,
      "num_tokens": 106093019.0,
      "step": 47520
    },
    {
      "entropy": 5.321600008010864,
      "epoch": 5.337507721682485,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002548802829798277,
      "loss": 4.8008,
      "mean_token_accuracy": 0.2392100304365158,
      "num_tokens": 106103572.0,
      "step": 47525
    },
    {
      "entropy": 5.2121936798095705,
      "epoch": 5.338069298590442,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002548402954202616,
      "loss": 4.6773,
      "mean_token_accuracy": 0.2440485268831253,
      "num_tokens": 106114633.0,
      "step": 47530
    },
    {
      "entropy": 5.288544416427612,
      "epoch": 5.338630875498399,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002548003085025883,
      "loss": 4.8374,
      "mean_token_accuracy": 0.23799049705266953,
      "num_tokens": 106126706.0,
      "step": 47535
    },
    {
      "entropy": 5.305491304397583,
      "epoch": 5.339192452406357,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025476032222808106,
      "loss": 4.8026,
      "mean_token_accuracy": 0.23320225030183792,
      "num_tokens": 106137180.0,
      "step": 47540
    },
    {
      "entropy": 5.274767160415649,
      "epoch": 5.339754029314315,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025472033659801297,
      "loss": 4.7545,
      "mean_token_accuracy": 0.24392879903316497,
      "num_tokens": 106148453.0,
      "step": 47545
    },
    {
      "entropy": 5.351220273971558,
      "epoch": 5.340315606222272,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002546803516136572,
      "loss": 4.9404,
      "mean_token_accuracy": 0.22069526016712188,
      "num_tokens": 106160262.0,
      "step": 47550
    },
    {
      "entropy": 5.320271778106689,
      "epoch": 5.34087718313023,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025464036727628695,
      "loss": 4.7679,
      "mean_token_accuracy": 0.23819841593503951,
      "num_tokens": 106170699.0,
      "step": 47555
    },
    {
      "entropy": 5.431132221221924,
      "epoch": 5.341438760038187,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025460038358717527,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23044902384281157,
      "num_tokens": 106181311.0,
      "step": 47560
    },
    {
      "entropy": 5.308819389343261,
      "epoch": 5.342000336946144,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002545604005475953,
      "loss": 4.8012,
      "mean_token_accuracy": 0.24007968306541444,
      "num_tokens": 106192651.0,
      "step": 47565
    },
    {
      "entropy": 5.319751310348511,
      "epoch": 5.3425619138541025,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002545204181588201,
      "loss": 4.825,
      "mean_token_accuracy": 0.24208389222621918,
      "num_tokens": 106203223.0,
      "step": 47570
    },
    {
      "entropy": 5.361058616638184,
      "epoch": 5.34312349076206,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00025448043642212274,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23403851091861724,
      "num_tokens": 106215403.0,
      "step": 47575
    },
    {
      "entropy": 5.366579294204712,
      "epoch": 5.343685067670018,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002544404553387763,
      "loss": 4.8605,
      "mean_token_accuracy": 0.22996347397565842,
      "num_tokens": 106226728.0,
      "step": 47580
    },
    {
      "entropy": 5.4105757713317875,
      "epoch": 5.344246644577975,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025440047491005367,
      "loss": 4.8307,
      "mean_token_accuracy": 0.2357742816209793,
      "num_tokens": 106238331.0,
      "step": 47585
    },
    {
      "entropy": 5.2663733959198,
      "epoch": 5.344808221485932,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025436049513722795,
      "loss": 4.7073,
      "mean_token_accuracy": 0.24326658993959427,
      "num_tokens": 106250095.0,
      "step": 47590
    },
    {
      "entropy": 5.271083879470825,
      "epoch": 5.34536979839389,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025432051602157203,
      "loss": 4.7289,
      "mean_token_accuracy": 0.24909279495477676,
      "num_tokens": 106260318.0,
      "step": 47595
    },
    {
      "entropy": 5.255571365356445,
      "epoch": 5.345931375301848,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002542805375643589,
      "loss": 4.7854,
      "mean_token_accuracy": 0.23965156078338623,
      "num_tokens": 106271110.0,
      "step": 47600
    },
    {
      "entropy": 5.448188447952271,
      "epoch": 5.346492952209805,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002542405597668615,
      "loss": 4.8917,
      "mean_token_accuracy": 0.23102816045284272,
      "num_tokens": 106283052.0,
      "step": 47605
    },
    {
      "entropy": 5.3275505065917965,
      "epoch": 5.347054529117763,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025420058263035276,
      "loss": 4.8081,
      "mean_token_accuracy": 0.23917210400104522,
      "num_tokens": 106293641.0,
      "step": 47610
    },
    {
      "entropy": 5.255698013305664,
      "epoch": 5.34761610602572,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002541606061561055,
      "loss": 4.8047,
      "mean_token_accuracy": 0.23552166521549225,
      "num_tokens": 106305230.0,
      "step": 47615
    },
    {
      "entropy": 5.347999048233032,
      "epoch": 5.348177682933677,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002541206303453927,
      "loss": 4.8982,
      "mean_token_accuracy": 0.23400607109069824,
      "num_tokens": 106316700.0,
      "step": 47620
    },
    {
      "entropy": 5.3459385395050045,
      "epoch": 5.3487392598416355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000254080655199487,
      "loss": 4.8339,
      "mean_token_accuracy": 0.23970895260572433,
      "num_tokens": 106327318.0,
      "step": 47625
    },
    {
      "entropy": 5.407559156417847,
      "epoch": 5.349300836749593,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025404068071966133,
      "loss": 4.8287,
      "mean_token_accuracy": 0.24170265793800355,
      "num_tokens": 106338134.0,
      "step": 47630
    },
    {
      "entropy": 5.323448562622071,
      "epoch": 5.349862413657551,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025400070690718865,
      "loss": 4.8542,
      "mean_token_accuracy": 0.2305162012577057,
      "num_tokens": 106348903.0,
      "step": 47635
    },
    {
      "entropy": 5.38914532661438,
      "epoch": 5.350423990565508,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025396073376334155,
      "loss": 4.8705,
      "mean_token_accuracy": 0.23939420580863952,
      "num_tokens": 106360124.0,
      "step": 47640
    },
    {
      "entropy": 5.356241893768311,
      "epoch": 5.350985567473465,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025392076128939283,
      "loss": 4.7838,
      "mean_token_accuracy": 0.23941846340894699,
      "num_tokens": 106371663.0,
      "step": 47645
    },
    {
      "entropy": 5.318000984191895,
      "epoch": 5.351547144381423,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002538807894866153,
      "loss": 4.8636,
      "mean_token_accuracy": 0.23861986696720122,
      "num_tokens": 106384337.0,
      "step": 47650
    },
    {
      "entropy": 5.3211921691894535,
      "epoch": 5.3521087212893805,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025384081835628163,
      "loss": 4.8457,
      "mean_token_accuracy": 0.2401798963546753,
      "num_tokens": 106395393.0,
      "step": 47655
    },
    {
      "entropy": 5.297837543487549,
      "epoch": 5.352670298197338,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002538008478996645,
      "loss": 4.7571,
      "mean_token_accuracy": 0.23714976757764816,
      "num_tokens": 106405585.0,
      "step": 47660
    },
    {
      "entropy": 5.3356608867645265,
      "epoch": 5.353231875105296,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002537608781180366,
      "loss": 4.7889,
      "mean_token_accuracy": 0.24064831137657167,
      "num_tokens": 106415906.0,
      "step": 47665
    },
    {
      "entropy": 5.2751891136169435,
      "epoch": 5.353793452013253,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002537209090126704,
      "loss": 4.7114,
      "mean_token_accuracy": 0.24205243587493896,
      "num_tokens": 106426530.0,
      "step": 47670
    },
    {
      "entropy": 5.393040657043457,
      "epoch": 5.35435502892121,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025368094058483885,
      "loss": 4.8827,
      "mean_token_accuracy": 0.23715000003576278,
      "num_tokens": 106438138.0,
      "step": 47675
    },
    {
      "entropy": 5.1934346675872805,
      "epoch": 5.354916605829168,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002536409728358145,
      "loss": 4.6702,
      "mean_token_accuracy": 0.24412281662225724,
      "num_tokens": 106449072.0,
      "step": 47680
    },
    {
      "entropy": 5.26467490196228,
      "epoch": 5.355478182737126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002536010057668698,
      "loss": 4.7247,
      "mean_token_accuracy": 0.24329632818698882,
      "num_tokens": 106458528.0,
      "step": 47685
    },
    {
      "entropy": 5.304969930648804,
      "epoch": 5.356039759645084,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002535610393792773,
      "loss": 4.7785,
      "mean_token_accuracy": 0.2432912230491638,
      "num_tokens": 106470375.0,
      "step": 47690
    },
    {
      "entropy": 5.327494764328003,
      "epoch": 5.356601336553041,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00025352107367430975,
      "loss": 4.8065,
      "mean_token_accuracy": 0.23267156928777694,
      "num_tokens": 106481994.0,
      "step": 47695
    },
    {
      "entropy": 5.408557271957397,
      "epoch": 5.357162913460998,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025348110865323946,
      "loss": 4.8528,
      "mean_token_accuracy": 0.23862131386995317,
      "num_tokens": 106492491.0,
      "step": 47700
    },
    {
      "entropy": 5.26609435081482,
      "epoch": 5.357724490368956,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002534411443173389,
      "loss": 4.8335,
      "mean_token_accuracy": 0.23392152339220046,
      "num_tokens": 106504487.0,
      "step": 47705
    },
    {
      "entropy": 5.258625745773315,
      "epoch": 5.3582860672769135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002534011806678809,
      "loss": 4.8111,
      "mean_token_accuracy": 0.2357006251811981,
      "num_tokens": 106517049.0,
      "step": 47710
    },
    {
      "entropy": 5.4435821056365965,
      "epoch": 5.358847644184872,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002533612177061376,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22527101933956145,
      "num_tokens": 106528533.0,
      "step": 47715
    },
    {
      "entropy": 5.3340071678161625,
      "epoch": 5.359409221092829,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025332125543338146,
      "loss": 4.7758,
      "mean_token_accuracy": 0.24207032471895218,
      "num_tokens": 106539462.0,
      "step": 47720
    },
    {
      "entropy": 5.3368260860443115,
      "epoch": 5.359970798000786,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000253281293850885,
      "loss": 4.7481,
      "mean_token_accuracy": 0.24541267305612563,
      "num_tokens": 106549864.0,
      "step": 47725
    },
    {
      "entropy": 5.270118284225464,
      "epoch": 5.360532374908744,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00025324133295992056,
      "loss": 4.7825,
      "mean_token_accuracy": 0.24192671924829484,
      "num_tokens": 106561990.0,
      "step": 47730
    },
    {
      "entropy": 5.271847152709961,
      "epoch": 5.361093951816701,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002532013727617605,
      "loss": 4.7114,
      "mean_token_accuracy": 0.24620699733495713,
      "num_tokens": 106572995.0,
      "step": 47735
    },
    {
      "entropy": 5.3785804271697994,
      "epoch": 5.361655528724659,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002531614132576772,
      "loss": 4.8613,
      "mean_token_accuracy": 0.23692122399806975,
      "num_tokens": 106583368.0,
      "step": 47740
    },
    {
      "entropy": 5.226902151107788,
      "epoch": 5.362217105632617,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000253121454448943,
      "loss": 4.7677,
      "mean_token_accuracy": 0.24133924543857574,
      "num_tokens": 106595515.0,
      "step": 47745
    },
    {
      "entropy": 5.328099155426026,
      "epoch": 5.362778682540574,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025308149633683014,
      "loss": 4.777,
      "mean_token_accuracy": 0.23965131491422653,
      "num_tokens": 106606662.0,
      "step": 47750
    },
    {
      "entropy": 5.333357286453247,
      "epoch": 5.363340259448531,
      "grad_norm": 1.125,
      "learning_rate": 0.000253041538922611,
      "loss": 4.8114,
      "mean_token_accuracy": 0.23325524032115935,
      "num_tokens": 106616750.0,
      "step": 47755
    },
    {
      "entropy": 5.35966944694519,
      "epoch": 5.363901836356489,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025300158220755776,
      "loss": 4.8158,
      "mean_token_accuracy": 0.2495961844921112,
      "num_tokens": 106630952.0,
      "step": 47760
    },
    {
      "entropy": 5.346912956237793,
      "epoch": 5.364463413264446,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025296162619294267,
      "loss": 4.8541,
      "mean_token_accuracy": 0.23015479296445845,
      "num_tokens": 106641158.0,
      "step": 47765
    },
    {
      "entropy": 5.278714179992676,
      "epoch": 5.3650249901724045,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002529216708800379,
      "loss": 4.7093,
      "mean_token_accuracy": 0.24825299680233,
      "num_tokens": 106651463.0,
      "step": 47770
    },
    {
      "entropy": 5.389921712875366,
      "epoch": 5.365586567080362,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025288171627011564,
      "loss": 4.9659,
      "mean_token_accuracy": 0.23136436939239502,
      "num_tokens": 106663896.0,
      "step": 47775
    },
    {
      "entropy": 5.268519067764283,
      "epoch": 5.366148143988319,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00025284176236444817,
      "loss": 4.6752,
      "mean_token_accuracy": 0.24296614676713943,
      "num_tokens": 106675805.0,
      "step": 47780
    },
    {
      "entropy": 5.315757989883423,
      "epoch": 5.366709720896277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025280180916430764,
      "loss": 4.8465,
      "mean_token_accuracy": 0.2337662771344185,
      "num_tokens": 106686392.0,
      "step": 47785
    },
    {
      "entropy": 5.348355436325074,
      "epoch": 5.367271297804234,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00025276185667096605,
      "loss": 4.8404,
      "mean_token_accuracy": 0.23706826716661453,
      "num_tokens": 106698724.0,
      "step": 47790
    },
    {
      "entropy": 5.281580829620362,
      "epoch": 5.3678328747121915,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025272190488569564,
      "loss": 4.698,
      "mean_token_accuracy": 0.24780855178833008,
      "num_tokens": 106710632.0,
      "step": 47795
    },
    {
      "entropy": 5.285517692565918,
      "epoch": 5.36839445162015,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025268195380976835,
      "loss": 4.7756,
      "mean_token_accuracy": 0.23960219025611879,
      "num_tokens": 106722476.0,
      "step": 47800
    },
    {
      "entropy": 5.246679306030273,
      "epoch": 5.368956028528107,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002526420034444564,
      "loss": 4.7544,
      "mean_token_accuracy": 0.24378267973661422,
      "num_tokens": 106734399.0,
      "step": 47805
    },
    {
      "entropy": 5.360498142242432,
      "epoch": 5.369517605436064,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025260205379103157,
      "loss": 4.8486,
      "mean_token_accuracy": 0.23830356001853942,
      "num_tokens": 106747873.0,
      "step": 47810
    },
    {
      "entropy": 5.34823169708252,
      "epoch": 5.370079182344022,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025256210485076616,
      "loss": 4.801,
      "mean_token_accuracy": 0.24208822697401047,
      "num_tokens": 106758382.0,
      "step": 47815
    },
    {
      "entropy": 5.330787229537964,
      "epoch": 5.370640759251979,
      "grad_norm": 1.078125,
      "learning_rate": 0.000252522156624932,
      "loss": 4.7863,
      "mean_token_accuracy": 0.23996385484933852,
      "num_tokens": 106768699.0,
      "step": 47820
    },
    {
      "entropy": 5.428933620452881,
      "epoch": 5.3712023361599375,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025248220911480105,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22548712491989137,
      "num_tokens": 106779926.0,
      "step": 47825
    },
    {
      "entropy": 5.360793972015381,
      "epoch": 5.371763913067895,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025244226232164533,
      "loss": 4.8384,
      "mean_token_accuracy": 0.23186196982860566,
      "num_tokens": 106791421.0,
      "step": 47830
    },
    {
      "entropy": 5.264261722564697,
      "epoch": 5.372325489975852,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025240231624673674,
      "loss": 4.7719,
      "mean_token_accuracy": 0.2383972227573395,
      "num_tokens": 106802074.0,
      "step": 47835
    },
    {
      "entropy": 5.291747665405273,
      "epoch": 5.37288706688381,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002523623708913472,
      "loss": 4.8243,
      "mean_token_accuracy": 0.2307315096259117,
      "num_tokens": 106812525.0,
      "step": 47840
    },
    {
      "entropy": 5.395515012741089,
      "epoch": 5.373448643791767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002523224262567485,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2326488196849823,
      "num_tokens": 106824248.0,
      "step": 47845
    },
    {
      "entropy": 5.431639051437378,
      "epoch": 5.3740102206997244,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025228248234421247,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23430620431900023,
      "num_tokens": 106834998.0,
      "step": 47850
    },
    {
      "entropy": 5.468851804733276,
      "epoch": 5.3745717976076826,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002522425391550111,
      "loss": 5.009,
      "mean_token_accuracy": 0.2272447884082794,
      "num_tokens": 106847251.0,
      "step": 47855
    },
    {
      "entropy": 5.324404239654541,
      "epoch": 5.37513337451564,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025220259669041613,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23712380230426788,
      "num_tokens": 106858761.0,
      "step": 47860
    },
    {
      "entropy": 5.229291152954102,
      "epoch": 5.375694951423597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002521626549516993,
      "loss": 4.7507,
      "mean_token_accuracy": 0.24566058367490767,
      "num_tokens": 106868874.0,
      "step": 47865
    },
    {
      "entropy": 5.41117148399353,
      "epoch": 5.376256528331555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002521227139401324,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2330926090478897,
      "num_tokens": 106881446.0,
      "step": 47870
    },
    {
      "entropy": 5.285684823989868,
      "epoch": 5.376818105239512,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00025208277365698715,
      "loss": 4.7593,
      "mean_token_accuracy": 0.24519579261541366,
      "num_tokens": 106893492.0,
      "step": 47875
    },
    {
      "entropy": 5.4346246242523195,
      "epoch": 5.37737968214747,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002520428341035353,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22349176406860352,
      "num_tokens": 106905295.0,
      "step": 47880
    },
    {
      "entropy": 5.341614723205566,
      "epoch": 5.377941259055428,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002520028952810485,
      "loss": 4.798,
      "mean_token_accuracy": 0.23807974457740783,
      "num_tokens": 106915330.0,
      "step": 47885
    },
    {
      "entropy": 5.252960252761841,
      "epoch": 5.378502835963385,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00025196295719079835,
      "loss": 4.7262,
      "mean_token_accuracy": 0.2461970940232277,
      "num_tokens": 106926703.0,
      "step": 47890
    },
    {
      "entropy": 5.329194450378418,
      "epoch": 5.379064412871343,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002519230198340567,
      "loss": 4.839,
      "mean_token_accuracy": 0.2304550975561142,
      "num_tokens": 106938132.0,
      "step": 47895
    },
    {
      "entropy": 5.324065923690796,
      "epoch": 5.3796259897793,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025188308321209504,
      "loss": 4.7887,
      "mean_token_accuracy": 0.23526858240365983,
      "num_tokens": 106948546.0,
      "step": 47900
    },
    {
      "entropy": 5.31822190284729,
      "epoch": 5.380187566687258,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025184314732618496,
      "loss": 4.8071,
      "mean_token_accuracy": 0.2350893050432205,
      "num_tokens": 106958208.0,
      "step": 47905
    },
    {
      "entropy": 5.279502296447754,
      "epoch": 5.3807491435952155,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000251803212177598,
      "loss": 4.7529,
      "mean_token_accuracy": 0.2486165001988411,
      "num_tokens": 106969057.0,
      "step": 47910
    },
    {
      "entropy": 5.291154575347901,
      "epoch": 5.381310720503173,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002517632777676058,
      "loss": 4.8357,
      "mean_token_accuracy": 0.2393472373485565,
      "num_tokens": 106980503.0,
      "step": 47915
    },
    {
      "entropy": 5.365461254119873,
      "epoch": 5.381872297411131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025172334409747976,
      "loss": 4.8454,
      "mean_token_accuracy": 0.23828313052654265,
      "num_tokens": 106992027.0,
      "step": 47920
    },
    {
      "entropy": 5.383753156661987,
      "epoch": 5.382433874319088,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002516834111684915,
      "loss": 4.8576,
      "mean_token_accuracy": 0.22911339104175568,
      "num_tokens": 107002873.0,
      "step": 47925
    },
    {
      "entropy": 5.262319421768188,
      "epoch": 5.382995451227045,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002516434789819126,
      "loss": 4.7519,
      "mean_token_accuracy": 0.2351854681968689,
      "num_tokens": 107014940.0,
      "step": 47930
    },
    {
      "entropy": 5.269494295120239,
      "epoch": 5.383557028135003,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002516035475390143,
      "loss": 4.7511,
      "mean_token_accuracy": 0.24307684749364852,
      "num_tokens": 107025750.0,
      "step": 47935
    },
    {
      "entropy": 5.273420000076294,
      "epoch": 5.384118605042961,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025156361684106814,
      "loss": 4.8044,
      "mean_token_accuracy": 0.2389841392636299,
      "num_tokens": 107035931.0,
      "step": 47940
    },
    {
      "entropy": 5.29899525642395,
      "epoch": 5.384680181950918,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002515236868893455,
      "loss": 4.8062,
      "mean_token_accuracy": 0.24495109915733337,
      "num_tokens": 107047799.0,
      "step": 47945
    },
    {
      "entropy": 5.3670509338378904,
      "epoch": 5.385241758858876,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00025148375768511776,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23430469036102294,
      "num_tokens": 107058434.0,
      "step": 47950
    },
    {
      "entropy": 5.416207981109619,
      "epoch": 5.385803335766833,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002514438292296563,
      "loss": 4.9891,
      "mean_token_accuracy": 0.23083456009626388,
      "num_tokens": 107070912.0,
      "step": 47955
    },
    {
      "entropy": 5.330361938476562,
      "epoch": 5.386364912674791,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025140390152423247,
      "loss": 4.8422,
      "mean_token_accuracy": 0.23689731508493422,
      "num_tokens": 107081885.0,
      "step": 47960
    },
    {
      "entropy": 5.456743574142456,
      "epoch": 5.3869264895827484,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002513639745701176,
      "loss": 4.8528,
      "mean_token_accuracy": 0.23894565999507905,
      "num_tokens": 107093859.0,
      "step": 47965
    },
    {
      "entropy": 5.248056125640869,
      "epoch": 5.387488066490706,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025132404836858294,
      "loss": 4.7651,
      "mean_token_accuracy": 0.23859720379114152,
      "num_tokens": 107104966.0,
      "step": 47970
    },
    {
      "entropy": 5.302087736129761,
      "epoch": 5.388049643398664,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002512841229208997,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23724612444639206,
      "num_tokens": 107116274.0,
      "step": 47975
    },
    {
      "entropy": 5.308991003036499,
      "epoch": 5.388611220306621,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002512441982283392,
      "loss": 4.7269,
      "mean_token_accuracy": 0.24343357384204864,
      "num_tokens": 107126366.0,
      "step": 47980
    },
    {
      "entropy": 5.40924186706543,
      "epoch": 5.389172797214578,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002512042742921727,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22660405784845353,
      "num_tokens": 107137557.0,
      "step": 47985
    },
    {
      "entropy": 5.438967990875244,
      "epoch": 5.389734374122536,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002511643511136712,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22508950978517533,
      "num_tokens": 107149034.0,
      "step": 47990
    },
    {
      "entropy": 5.386489343643189,
      "epoch": 5.3902959510304935,
      "grad_norm": 0.953125,
      "learning_rate": 0.00025112442869410607,
      "loss": 4.8378,
      "mean_token_accuracy": 0.23104346692562103,
      "num_tokens": 107161151.0,
      "step": 47995
    },
    {
      "entropy": 5.356549263000488,
      "epoch": 5.390857527938451,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025108450703474844,
      "loss": 4.7927,
      "mean_token_accuracy": 0.2407902255654335,
      "num_tokens": 107172423.0,
      "step": 48000
    },
    {
      "epoch": 5.390857527938451,
      "eval_entropy": 5.140413817591909,
      "eval_loss": 5.010140895843506,
      "eval_mean_token_accuracy": 0.23330395977952784,
      "eval_num_tokens": 107172423.0,
      "eval_runtime": 23.8979,
      "eval_samples_per_second": 1297.73,
      "eval_steps_per_second": 162.232,
      "step": 48000
    },
    {
      "entropy": 5.335368633270264,
      "epoch": 5.391419104846409,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025104458613686936,
      "loss": 4.7737,
      "mean_token_accuracy": 0.23998275846242906,
      "num_tokens": 107183609.0,
      "step": 48005
    },
    {
      "entropy": 5.340459680557251,
      "epoch": 5.391980681754366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025100466600174,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2314401775598526,
      "num_tokens": 107195225.0,
      "step": 48010
    },
    {
      "entropy": 5.35895562171936,
      "epoch": 5.392542258662324,
      "grad_norm": 1.125,
      "learning_rate": 0.0002509647466306313,
      "loss": 4.8782,
      "mean_token_accuracy": 0.23695301115512848,
      "num_tokens": 107206399.0,
      "step": 48015
    },
    {
      "entropy": 5.304448080062866,
      "epoch": 5.393103835570281,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025092482802481443,
      "loss": 4.7861,
      "mean_token_accuracy": 0.23737318813800812,
      "num_tokens": 107217099.0,
      "step": 48020
    },
    {
      "entropy": 5.346317338943481,
      "epoch": 5.393665412478239,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002508849101855604,
      "loss": 4.8937,
      "mean_token_accuracy": 0.22852680087089539,
      "num_tokens": 107228463.0,
      "step": 48025
    },
    {
      "entropy": 5.3188355445861815,
      "epoch": 5.394226989386197,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002508449931141401,
      "loss": 4.7544,
      "mean_token_accuracy": 0.24686156511306762,
      "num_tokens": 107238118.0,
      "step": 48030
    },
    {
      "entropy": 5.342896747589111,
      "epoch": 5.394788566294154,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025080507681182464,
      "loss": 4.8091,
      "mean_token_accuracy": 0.23939220160245894,
      "num_tokens": 107249537.0,
      "step": 48035
    },
    {
      "entropy": 5.375359201431275,
      "epoch": 5.395350143202111,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002507651612798849,
      "loss": 4.873,
      "mean_token_accuracy": 0.23002255260944365,
      "num_tokens": 107261005.0,
      "step": 48040
    },
    {
      "entropy": 5.278658246994018,
      "epoch": 5.395911720110069,
      "grad_norm": 0.96875,
      "learning_rate": 0.00025072524651959183,
      "loss": 4.7331,
      "mean_token_accuracy": 0.24016711711883545,
      "num_tokens": 107272099.0,
      "step": 48045
    },
    {
      "entropy": 5.277265167236328,
      "epoch": 5.3964732970180265,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002506853325322164,
      "loss": 4.7085,
      "mean_token_accuracy": 0.2416931137442589,
      "num_tokens": 107283290.0,
      "step": 48050
    },
    {
      "entropy": 5.267444229125976,
      "epoch": 5.397034873925985,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002506454193190293,
      "loss": 4.8482,
      "mean_token_accuracy": 0.23336421847343444,
      "num_tokens": 107295028.0,
      "step": 48055
    },
    {
      "entropy": 5.321864175796509,
      "epoch": 5.397596450833942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025060550688130157,
      "loss": 4.7909,
      "mean_token_accuracy": 0.2426442950963974,
      "num_tokens": 107306374.0,
      "step": 48060
    },
    {
      "entropy": 5.350157594680786,
      "epoch": 5.398158027741899,
      "grad_norm": 1.0,
      "learning_rate": 0.00025056559522030383,
      "loss": 4.8652,
      "mean_token_accuracy": 0.23519549369812012,
      "num_tokens": 107318241.0,
      "step": 48065
    },
    {
      "entropy": 5.378100967407226,
      "epoch": 5.398719604649857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002505256843373071,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23076984286308289,
      "num_tokens": 107330252.0,
      "step": 48070
    },
    {
      "entropy": 5.366825437545776,
      "epoch": 5.399281181557814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002504857742335821,
      "loss": 4.8999,
      "mean_token_accuracy": 0.23679858893156053,
      "num_tokens": 107341771.0,
      "step": 48075
    },
    {
      "entropy": 5.262651777267456,
      "epoch": 5.3998427584657716,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025044586491039956,
      "loss": 4.7623,
      "mean_token_accuracy": 0.23433341979980468,
      "num_tokens": 107352963.0,
      "step": 48080
    },
    {
      "entropy": 5.261600923538208,
      "epoch": 5.40040433537373,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002504059563690302,
      "loss": 4.7498,
      "mean_token_accuracy": 0.24079796969890593,
      "num_tokens": 107363901.0,
      "step": 48085
    },
    {
      "entropy": 5.288555765151978,
      "epoch": 5.400965912281687,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002503660486107447,
      "loss": 4.7823,
      "mean_token_accuracy": 0.24232788681983947,
      "num_tokens": 107376033.0,
      "step": 48090
    },
    {
      "entropy": 5.320987606048584,
      "epoch": 5.401527489189645,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002503261416368138,
      "loss": 4.786,
      "mean_token_accuracy": 0.24333046972751618,
      "num_tokens": 107387134.0,
      "step": 48095
    },
    {
      "entropy": 5.292643117904663,
      "epoch": 5.402089066097602,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002502862354485081,
      "loss": 4.7383,
      "mean_token_accuracy": 0.24711241126060485,
      "num_tokens": 107397186.0,
      "step": 48100
    },
    {
      "entropy": 5.276294851303101,
      "epoch": 5.402650643005559,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025024633004709824,
      "loss": 4.7672,
      "mean_token_accuracy": 0.24313766360282899,
      "num_tokens": 107407841.0,
      "step": 48105
    },
    {
      "entropy": 5.285129117965698,
      "epoch": 5.4032122199135175,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025020642543385486,
      "loss": 4.6922,
      "mean_token_accuracy": 0.2507660076022148,
      "num_tokens": 107418530.0,
      "step": 48110
    },
    {
      "entropy": 5.342335319519043,
      "epoch": 5.403773796821475,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025016652161004847,
      "loss": 4.8826,
      "mean_token_accuracy": 0.22995602637529372,
      "num_tokens": 107429424.0,
      "step": 48115
    },
    {
      "entropy": 5.329483413696289,
      "epoch": 5.404335373729432,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002501266185769497,
      "loss": 4.7877,
      "mean_token_accuracy": 0.23626340180635452,
      "num_tokens": 107440323.0,
      "step": 48120
    },
    {
      "entropy": 5.360688543319702,
      "epoch": 5.40489695063739,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025008671633582904,
      "loss": 4.8002,
      "mean_token_accuracy": 0.24705660492181777,
      "num_tokens": 107451449.0,
      "step": 48125
    },
    {
      "entropy": 5.4204802989959715,
      "epoch": 5.405458527545347,
      "grad_norm": 1.0625,
      "learning_rate": 0.000250046814887957,
      "loss": 5.0,
      "mean_token_accuracy": 0.22388266772031784,
      "num_tokens": 107463605.0,
      "step": 48130
    },
    {
      "entropy": 5.34502911567688,
      "epoch": 5.4060201044533045,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002500069142346039,
      "loss": 4.8651,
      "mean_token_accuracy": 0.23374599665403367,
      "num_tokens": 107475767.0,
      "step": 48135
    },
    {
      "entropy": 5.293302297592163,
      "epoch": 5.406581681361263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002499670143770405,
      "loss": 4.7461,
      "mean_token_accuracy": 0.24396583586931228,
      "num_tokens": 107486039.0,
      "step": 48140
    },
    {
      "entropy": 5.302363634109497,
      "epoch": 5.40714325826922,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002499271153165371,
      "loss": 4.801,
      "mean_token_accuracy": 0.23384447991847992,
      "num_tokens": 107497029.0,
      "step": 48145
    },
    {
      "entropy": 5.3703282356262205,
      "epoch": 5.407704835177178,
      "grad_norm": 1.125,
      "learning_rate": 0.000249887217054364,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22903691977262497,
      "num_tokens": 107508605.0,
      "step": 48150
    },
    {
      "entropy": 5.312437582015991,
      "epoch": 5.408266412085135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002498473195917917,
      "loss": 4.7911,
      "mean_token_accuracy": 0.23885067999362947,
      "num_tokens": 107519688.0,
      "step": 48155
    },
    {
      "entropy": 5.236100530624389,
      "epoch": 5.408827988993092,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002498074229300905,
      "loss": 4.7003,
      "mean_token_accuracy": 0.24818704426288604,
      "num_tokens": 107531291.0,
      "step": 48160
    },
    {
      "entropy": 5.1776265621185305,
      "epoch": 5.4093895659010505,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002497675270705307,
      "loss": 4.6571,
      "mean_token_accuracy": 0.24140879213809968,
      "num_tokens": 107541581.0,
      "step": 48165
    },
    {
      "entropy": 5.3026731491088865,
      "epoch": 5.409951142809008,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002497276320143826,
      "loss": 4.8555,
      "mean_token_accuracy": 0.2361392304301262,
      "num_tokens": 107552038.0,
      "step": 48170
    },
    {
      "entropy": 5.418278598785401,
      "epoch": 5.410512719716965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024968773776291647,
      "loss": 4.8262,
      "mean_token_accuracy": 0.23506632596254348,
      "num_tokens": 107563669.0,
      "step": 48175
    },
    {
      "entropy": 5.216590070724488,
      "epoch": 5.411074296624923,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002496478443174026,
      "loss": 4.7262,
      "mean_token_accuracy": 0.2494857758283615,
      "num_tokens": 107576509.0,
      "step": 48180
    },
    {
      "entropy": 5.341580486297607,
      "epoch": 5.41163587353288,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002496079516791113,
      "loss": 4.8979,
      "mean_token_accuracy": 0.23061861395835875,
      "num_tokens": 107586601.0,
      "step": 48185
    },
    {
      "entropy": 5.378854894638062,
      "epoch": 5.412197450440837,
      "grad_norm": 1.125,
      "learning_rate": 0.0002495680598493126,
      "loss": 4.8331,
      "mean_token_accuracy": 0.23445071280002594,
      "num_tokens": 107597357.0,
      "step": 48190
    },
    {
      "entropy": 5.353949356079101,
      "epoch": 5.4127590273487955,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002495281688292768,
      "loss": 4.7629,
      "mean_token_accuracy": 0.23308364748954774,
      "num_tokens": 107607727.0,
      "step": 48195
    },
    {
      "entropy": 5.404633903503418,
      "epoch": 5.413320604256753,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024948827862027385,
      "loss": 4.8694,
      "mean_token_accuracy": 0.2384641319513321,
      "num_tokens": 107617973.0,
      "step": 48200
    },
    {
      "entropy": 5.360018110275268,
      "epoch": 5.413882181164711,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002494483892235741,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22808065116405488,
      "num_tokens": 107628845.0,
      "step": 48205
    },
    {
      "entropy": 5.31061201095581,
      "epoch": 5.414443758072668,
      "grad_norm": 1.0,
      "learning_rate": 0.0002494085006404474,
      "loss": 4.7339,
      "mean_token_accuracy": 0.2434367686510086,
      "num_tokens": 107639702.0,
      "step": 48210
    },
    {
      "entropy": 5.365789127349854,
      "epoch": 5.415005334980625,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002493686128721641,
      "loss": 4.8397,
      "mean_token_accuracy": 0.24060240983963013,
      "num_tokens": 107651754.0,
      "step": 48215
    },
    {
      "entropy": 5.429797601699829,
      "epoch": 5.415566911888583,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00024932872591999414,
      "loss": 4.9975,
      "mean_token_accuracy": 0.22708537429571152,
      "num_tokens": 107664271.0,
      "step": 48220
    },
    {
      "entropy": 5.334953641891479,
      "epoch": 5.416128488796541,
      "grad_norm": 0.9375,
      "learning_rate": 0.00024928883978520743,
      "loss": 4.7835,
      "mean_token_accuracy": 0.242853944003582,
      "num_tokens": 107675316.0,
      "step": 48225
    },
    {
      "entropy": 5.345213747024536,
      "epoch": 5.416690065704498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024924895446907407,
      "loss": 4.8549,
      "mean_token_accuracy": 0.23540873378515242,
      "num_tokens": 107686806.0,
      "step": 48230
    },
    {
      "entropy": 5.357829761505127,
      "epoch": 5.417251642612456,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000249209069972864,
      "loss": 4.7755,
      "mean_token_accuracy": 0.2364482581615448,
      "num_tokens": 107697097.0,
      "step": 48235
    },
    {
      "entropy": 5.303747844696045,
      "epoch": 5.417813219520413,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024916918629784714,
      "loss": 4.7804,
      "mean_token_accuracy": 0.23974881917238236,
      "num_tokens": 107708117.0,
      "step": 48240
    },
    {
      "entropy": 5.410014247894287,
      "epoch": 5.418374796428371,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002491293034452933,
      "loss": 4.8768,
      "mean_token_accuracy": 0.228361177444458,
      "num_tokens": 107718927.0,
      "step": 48245
    },
    {
      "entropy": 5.316575956344605,
      "epoch": 5.4189363733363285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002490894214164726,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23911189436912536,
      "num_tokens": 107730362.0,
      "step": 48250
    },
    {
      "entropy": 5.283337783813477,
      "epoch": 5.419497950244286,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002490495402126548,
      "loss": 4.8372,
      "mean_token_accuracy": 0.23491731584072112,
      "num_tokens": 107741148.0,
      "step": 48255
    },
    {
      "entropy": 5.28989052772522,
      "epoch": 5.420059527152244,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002490096598351096,
      "loss": 4.7491,
      "mean_token_accuracy": 0.2360307678580284,
      "num_tokens": 107752023.0,
      "step": 48260
    },
    {
      "entropy": 5.299873685836792,
      "epoch": 5.420621104060201,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024896978028510696,
      "loss": 4.7397,
      "mean_token_accuracy": 0.24120659232139588,
      "num_tokens": 107763136.0,
      "step": 48265
    },
    {
      "entropy": 5.339196681976318,
      "epoch": 5.421182680968158,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002489299015639167,
      "loss": 4.8632,
      "mean_token_accuracy": 0.24057164639234543,
      "num_tokens": 107774372.0,
      "step": 48270
    },
    {
      "entropy": 5.290726089477539,
      "epoch": 5.421744257876116,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002488900236728083,
      "loss": 4.7779,
      "mean_token_accuracy": 0.23183539807796477,
      "num_tokens": 107785042.0,
      "step": 48275
    },
    {
      "entropy": 5.361528444290161,
      "epoch": 5.422305834784074,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002488501466130517,
      "loss": 4.7918,
      "mean_token_accuracy": 0.24187275618314744,
      "num_tokens": 107795476.0,
      "step": 48280
    },
    {
      "entropy": 5.28851580619812,
      "epoch": 5.422867411692032,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024881027038591664,
      "loss": 4.7843,
      "mean_token_accuracy": 0.2346884489059448,
      "num_tokens": 107806016.0,
      "step": 48285
    },
    {
      "entropy": 5.309561109542846,
      "epoch": 5.423428988599989,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00024877039499267273,
      "loss": 4.7925,
      "mean_token_accuracy": 0.23707161992788314,
      "num_tokens": 107815988.0,
      "step": 48290
    },
    {
      "entropy": 5.281680679321289,
      "epoch": 5.423990565507946,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002487305204345896,
      "loss": 4.7564,
      "mean_token_accuracy": 0.24033102095127107,
      "num_tokens": 107826321.0,
      "step": 48295
    },
    {
      "entropy": 5.349907541275025,
      "epoch": 5.424552142415904,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024869064671293687,
      "loss": 4.7948,
      "mean_token_accuracy": 0.24109698235988616,
      "num_tokens": 107837031.0,
      "step": 48300
    },
    {
      "entropy": 5.40595269203186,
      "epoch": 5.425113719323861,
      "grad_norm": 1.125,
      "learning_rate": 0.00024865077382898423,
      "loss": 4.9073,
      "mean_token_accuracy": 0.23471041023731232,
      "num_tokens": 107846739.0,
      "step": 48305
    },
    {
      "entropy": 5.372014427185059,
      "epoch": 5.425675296231819,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024861090178400106,
      "loss": 4.8312,
      "mean_token_accuracy": 0.23821086883544923,
      "num_tokens": 107858640.0,
      "step": 48310
    },
    {
      "entropy": 5.3566468238830565,
      "epoch": 5.426236873139777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024857103057925707,
      "loss": 4.8504,
      "mean_token_accuracy": 0.23501634746789932,
      "num_tokens": 107871586.0,
      "step": 48315
    },
    {
      "entropy": 5.3845624923706055,
      "epoch": 5.426798450047734,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024853116021602173,
      "loss": 4.8358,
      "mean_token_accuracy": 0.23712316751480103,
      "num_tokens": 107881979.0,
      "step": 48320
    },
    {
      "entropy": 5.3541182518005375,
      "epoch": 5.427360026955691,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002484912906955645,
      "loss": 4.9169,
      "mean_token_accuracy": 0.23002346009016036,
      "num_tokens": 107893899.0,
      "step": 48325
    },
    {
      "entropy": 5.360014629364014,
      "epoch": 5.427921603863649,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002484514220191549,
      "loss": 4.8031,
      "mean_token_accuracy": 0.23989544808864594,
      "num_tokens": 107904432.0,
      "step": 48330
    },
    {
      "entropy": 5.369151592254639,
      "epoch": 5.4284831807716065,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002484115541880623,
      "loss": 4.8935,
      "mean_token_accuracy": 0.23638660013675689,
      "num_tokens": 107915792.0,
      "step": 48335
    },
    {
      "entropy": 5.272096347808838,
      "epoch": 5.429044757679565,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024837168720355605,
      "loss": 4.6869,
      "mean_token_accuracy": 0.25365591496229173,
      "num_tokens": 107927546.0,
      "step": 48340
    },
    {
      "entropy": 5.273371458053589,
      "epoch": 5.429606334587522,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002483318210669057,
      "loss": 4.7468,
      "mean_token_accuracy": 0.24341964423656465,
      "num_tokens": 107937965.0,
      "step": 48345
    },
    {
      "entropy": 5.411024856567383,
      "epoch": 5.430167911495479,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002482919557793805,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2345738723874092,
      "num_tokens": 107949581.0,
      "step": 48350
    },
    {
      "entropy": 5.362826871871948,
      "epoch": 5.430729488403437,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00024825209134224965,
      "loss": 4.8227,
      "mean_token_accuracy": 0.2354622572660446,
      "num_tokens": 107961636.0,
      "step": 48355
    },
    {
      "entropy": 5.395572233200073,
      "epoch": 5.431291065311394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002482122277567827,
      "loss": 4.8663,
      "mean_token_accuracy": 0.24542443901300431,
      "num_tokens": 107972488.0,
      "step": 48360
    },
    {
      "entropy": 5.336085081100464,
      "epoch": 5.431852642219352,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024817236502424885,
      "loss": 4.8361,
      "mean_token_accuracy": 0.235746169090271,
      "num_tokens": 107984071.0,
      "step": 48365
    },
    {
      "entropy": 5.278129243850708,
      "epoch": 5.43241421912731,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002481325031459172,
      "loss": 4.7799,
      "mean_token_accuracy": 0.23761677592992783,
      "num_tokens": 107994816.0,
      "step": 48370
    },
    {
      "entropy": 5.291623973846436,
      "epoch": 5.432975796035267,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00024809264212305714,
      "loss": 4.747,
      "mean_token_accuracy": 0.24205498546361923,
      "num_tokens": 108005737.0,
      "step": 48375
    },
    {
      "entropy": 5.314851236343384,
      "epoch": 5.433537372943224,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002480527819569378,
      "loss": 4.7747,
      "mean_token_accuracy": 0.24083291441202165,
      "num_tokens": 108017519.0,
      "step": 48380
    },
    {
      "entropy": 5.391932964324951,
      "epoch": 5.434098949851182,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002480129226488283,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23004532605409622,
      "num_tokens": 108029318.0,
      "step": 48385
    },
    {
      "entropy": 5.264876890182495,
      "epoch": 5.4346605267591395,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024797306419999777,
      "loss": 4.7788,
      "mean_token_accuracy": 0.23958482295274736,
      "num_tokens": 108040844.0,
      "step": 48390
    },
    {
      "entropy": 5.287418413162231,
      "epoch": 5.435222103667098,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002479332066117154,
      "loss": 4.818,
      "mean_token_accuracy": 0.23970102816820144,
      "num_tokens": 108052558.0,
      "step": 48395
    },
    {
      "entropy": 5.392337083816528,
      "epoch": 5.435783680575055,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002478933498852503,
      "loss": 4.8853,
      "mean_token_accuracy": 0.2348636880517006,
      "num_tokens": 108063619.0,
      "step": 48400
    },
    {
      "entropy": 5.322489976882935,
      "epoch": 5.436345257483012,
      "grad_norm": 1.0,
      "learning_rate": 0.0002478534940218714,
      "loss": 4.7795,
      "mean_token_accuracy": 0.24001667648553848,
      "num_tokens": 108074942.0,
      "step": 48405
    },
    {
      "entropy": 5.34663143157959,
      "epoch": 5.43690683439097,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024781363902284776,
      "loss": 4.848,
      "mean_token_accuracy": 0.2299207016825676,
      "num_tokens": 108086245.0,
      "step": 48410
    },
    {
      "entropy": 5.231891775131226,
      "epoch": 5.437468411298927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002477737848894484,
      "loss": 4.7249,
      "mean_token_accuracy": 0.24126538038253784,
      "num_tokens": 108096251.0,
      "step": 48415
    },
    {
      "entropy": 5.385582017898559,
      "epoch": 5.4380299882068845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002477339316229423,
      "loss": 4.8525,
      "mean_token_accuracy": 0.24366098046302795,
      "num_tokens": 108107494.0,
      "step": 48420
    },
    {
      "entropy": 5.3206957340240475,
      "epoch": 5.438591565114843,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002476940792245983,
      "loss": 4.815,
      "mean_token_accuracy": 0.23797137141227723,
      "num_tokens": 108119260.0,
      "step": 48425
    },
    {
      "entropy": 5.22964334487915,
      "epoch": 5.4391531420228,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024765422769568553,
      "loss": 4.7092,
      "mean_token_accuracy": 0.24271993190050126,
      "num_tokens": 108130125.0,
      "step": 48430
    },
    {
      "entropy": 5.380172538757324,
      "epoch": 5.439714718930758,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002476143770374727,
      "loss": 4.9516,
      "mean_token_accuracy": 0.22424257695674896,
      "num_tokens": 108141707.0,
      "step": 48435
    },
    {
      "entropy": 5.422353315353393,
      "epoch": 5.440276295838715,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024757452725122876,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23495023101568221,
      "num_tokens": 108154038.0,
      "step": 48440
    },
    {
      "entropy": 5.376693534851074,
      "epoch": 5.440837872746672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002475346783382225,
      "loss": 4.8533,
      "mean_token_accuracy": 0.23733012676239013,
      "num_tokens": 108165111.0,
      "step": 48445
    },
    {
      "entropy": 5.351035022735596,
      "epoch": 5.4413994496546305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024749483029972276,
      "loss": 4.8625,
      "mean_token_accuracy": 0.22563242465257644,
      "num_tokens": 108176139.0,
      "step": 48450
    },
    {
      "entropy": 5.430060291290284,
      "epoch": 5.441961026562588,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002474549831369983,
      "loss": 4.9643,
      "mean_token_accuracy": 0.23415762484073638,
      "num_tokens": 108187818.0,
      "step": 48455
    },
    {
      "entropy": 5.358787202835083,
      "epoch": 5.442522603470545,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024741513685131773,
      "loss": 4.8356,
      "mean_token_accuracy": 0.23578021973371505,
      "num_tokens": 108198153.0,
      "step": 48460
    },
    {
      "entropy": 5.30645170211792,
      "epoch": 5.443084180378503,
      "grad_norm": 1.0,
      "learning_rate": 0.00024737529144395,
      "loss": 4.7559,
      "mean_token_accuracy": 0.23851645290851592,
      "num_tokens": 108210334.0,
      "step": 48465
    },
    {
      "entropy": 5.283547973632812,
      "epoch": 5.44364575728646,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002473354469161637,
      "loss": 4.7968,
      "mean_token_accuracy": 0.2413155496120453,
      "num_tokens": 108222912.0,
      "step": 48470
    },
    {
      "entropy": 5.182802724838257,
      "epoch": 5.444207334194418,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002472956032692275,
      "loss": 4.6767,
      "mean_token_accuracy": 0.24550610333681105,
      "num_tokens": 108232911.0,
      "step": 48475
    },
    {
      "entropy": 5.315421628952026,
      "epoch": 5.444768911102376,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002472557605044101,
      "loss": 4.8381,
      "mean_token_accuracy": 0.23393265157938004,
      "num_tokens": 108244464.0,
      "step": 48480
    },
    {
      "entropy": 5.350901031494141,
      "epoch": 5.445330488010333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00024721591862297994,
      "loss": 4.7992,
      "mean_token_accuracy": 0.24357731342315675,
      "num_tokens": 108255609.0,
      "step": 48485
    },
    {
      "entropy": 5.373753929138184,
      "epoch": 5.445892064918291,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024717607762620575,
      "loss": 4.8351,
      "mean_token_accuracy": 0.23422912806272506,
      "num_tokens": 108266076.0,
      "step": 48490
    },
    {
      "entropy": 5.209778642654419,
      "epoch": 5.446453641826248,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002471362375153559,
      "loss": 4.6251,
      "mean_token_accuracy": 0.25387144535779954,
      "num_tokens": 108277091.0,
      "step": 48495
    },
    {
      "entropy": 5.243908643722534,
      "epoch": 5.447015218734205,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024709639829169924,
      "loss": 4.7543,
      "mean_token_accuracy": 0.2396395042538643,
      "num_tokens": 108288224.0,
      "step": 48500
    },
    {
      "entropy": 5.339846658706665,
      "epoch": 5.4475767956421635,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024705655995650407,
      "loss": 4.8495,
      "mean_token_accuracy": 0.23457620441913604,
      "num_tokens": 108298841.0,
      "step": 48505
    },
    {
      "entropy": 5.28974609375,
      "epoch": 5.448138372550121,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002470167225110388,
      "loss": 4.7265,
      "mean_token_accuracy": 0.24643294662237167,
      "num_tokens": 108309864.0,
      "step": 48510
    },
    {
      "entropy": 5.294095373153686,
      "epoch": 5.448699949458078,
      "grad_norm": 1.046875,
      "learning_rate": 0.000246976885956572,
      "loss": 4.7584,
      "mean_token_accuracy": 0.23491716384887695,
      "num_tokens": 108320097.0,
      "step": 48515
    },
    {
      "entropy": 5.292574834823609,
      "epoch": 5.449261526366036,
      "grad_norm": 1.03125,
      "learning_rate": 0.000246937050294372,
      "loss": 4.7734,
      "mean_token_accuracy": 0.23955602049827576,
      "num_tokens": 108330904.0,
      "step": 48520
    },
    {
      "entropy": 5.289916276931763,
      "epoch": 5.449823103273993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024689721552570714,
      "loss": 4.7926,
      "mean_token_accuracy": 0.24310359805822374,
      "num_tokens": 108340549.0,
      "step": 48525
    },
    {
      "entropy": 5.259102630615234,
      "epoch": 5.450384680181951,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002468573816518459,
      "loss": 4.7676,
      "mean_token_accuracy": 0.24304784089326859,
      "num_tokens": 108352427.0,
      "step": 48530
    },
    {
      "entropy": 5.344539022445678,
      "epoch": 5.4509462570899085,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002468175486740565,
      "loss": 4.8608,
      "mean_token_accuracy": 0.23490267544984816,
      "num_tokens": 108365228.0,
      "step": 48535
    },
    {
      "entropy": 5.329259538650513,
      "epoch": 5.451507833997866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002467777165936073,
      "loss": 4.7744,
      "mean_token_accuracy": 0.23860872834920882,
      "num_tokens": 108375584.0,
      "step": 48540
    },
    {
      "entropy": 5.391112852096557,
      "epoch": 5.452069410905824,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002467378854117666,
      "loss": 4.8307,
      "mean_token_accuracy": 0.23607154041528702,
      "num_tokens": 108386634.0,
      "step": 48545
    },
    {
      "entropy": 5.33463716506958,
      "epoch": 5.452630987813781,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002466980551298026,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2392538905143738,
      "num_tokens": 108397895.0,
      "step": 48550
    },
    {
      "entropy": 5.338215780258179,
      "epoch": 5.453192564721738,
      "grad_norm": 1.125,
      "learning_rate": 0.00024665822574898344,
      "loss": 4.8483,
      "mean_token_accuracy": 0.2361769884824753,
      "num_tokens": 108407576.0,
      "step": 48555
    },
    {
      "entropy": 5.353451251983643,
      "epoch": 5.453754141629696,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00024661839727057737,
      "loss": 4.8739,
      "mean_token_accuracy": 0.23070045709609985,
      "num_tokens": 108418751.0,
      "step": 48560
    },
    {
      "entropy": 5.409369325637817,
      "epoch": 5.454315718537654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024657856969585257,
      "loss": 4.8886,
      "mean_token_accuracy": 0.23030151724815368,
      "num_tokens": 108429997.0,
      "step": 48565
    },
    {
      "entropy": 5.410181474685669,
      "epoch": 5.454877295445611,
      "grad_norm": 1.0625,
      "learning_rate": 0.000246538743026077,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22826776802539825,
      "num_tokens": 108441380.0,
      "step": 48570
    },
    {
      "entropy": 5.324895906448364,
      "epoch": 5.455438872353569,
      "grad_norm": 1.03125,
      "learning_rate": 0.000246498917262519,
      "loss": 4.7739,
      "mean_token_accuracy": 0.24235834628343583,
      "num_tokens": 108452985.0,
      "step": 48575
    },
    {
      "entropy": 5.473445987701416,
      "epoch": 5.456000449261526,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002464590924064465,
      "loss": 5.0096,
      "mean_token_accuracy": 0.22290758043527603,
      "num_tokens": 108463624.0,
      "step": 48580
    },
    {
      "entropy": 5.359873962402344,
      "epoch": 5.456562026169484,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024641926845912765,
      "loss": 4.8107,
      "mean_token_accuracy": 0.23640419840812682,
      "num_tokens": 108475736.0,
      "step": 48585
    },
    {
      "entropy": 5.310965538024902,
      "epoch": 5.4571236030774415,
      "grad_norm": 0.96875,
      "learning_rate": 0.00024637944542183035,
      "loss": 4.7884,
      "mean_token_accuracy": 0.24931090772151948,
      "num_tokens": 108486238.0,
      "step": 48590
    },
    {
      "entropy": 5.248914051055908,
      "epoch": 5.457685179985399,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024633962329582255,
      "loss": 4.7505,
      "mean_token_accuracy": 0.24895514994859697,
      "num_tokens": 108498929.0,
      "step": 48595
    },
    {
      "entropy": 5.366007375717163,
      "epoch": 5.458246756893357,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002462998020823723,
      "loss": 4.87,
      "mean_token_accuracy": 0.24027867168188094,
      "num_tokens": 108511727.0,
      "step": 48600
    },
    {
      "entropy": 5.287298393249512,
      "epoch": 5.458808333801314,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024625998178274747,
      "loss": 4.7403,
      "mean_token_accuracy": 0.2482757031917572,
      "num_tokens": 108522794.0,
      "step": 48605
    },
    {
      "entropy": 5.403678131103516,
      "epoch": 5.459369910709271,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002462201623982159,
      "loss": 4.92,
      "mean_token_accuracy": 0.2319077342748642,
      "num_tokens": 108534021.0,
      "step": 48610
    },
    {
      "entropy": 5.372740125656128,
      "epoch": 5.459931487617229,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002461803439300456,
      "loss": 4.816,
      "mean_token_accuracy": 0.24020825326442719,
      "num_tokens": 108544509.0,
      "step": 48615
    },
    {
      "entropy": 5.301140069961548,
      "epoch": 5.460493064525187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002461405263795044,
      "loss": 4.8511,
      "mean_token_accuracy": 0.23806069642305375,
      "num_tokens": 108554718.0,
      "step": 48620
    },
    {
      "entropy": 5.305284452438355,
      "epoch": 5.461054641433145,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002461007097478599,
      "loss": 4.7618,
      "mean_token_accuracy": 0.2414195343852043,
      "num_tokens": 108566660.0,
      "step": 48625
    },
    {
      "entropy": 5.350522089004516,
      "epoch": 5.461616218341102,
      "grad_norm": 1.0,
      "learning_rate": 0.0002460608940363801,
      "loss": 4.8783,
      "mean_token_accuracy": 0.2351525530219078,
      "num_tokens": 108578187.0,
      "step": 48630
    },
    {
      "entropy": 5.295495796203613,
      "epoch": 5.462177795249059,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024602107924633266,
      "loss": 4.7807,
      "mean_token_accuracy": 0.2373659700155258,
      "num_tokens": 108590181.0,
      "step": 48635
    },
    {
      "entropy": 5.285640573501587,
      "epoch": 5.462739372157017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002459812653789851,
      "loss": 4.7542,
      "mean_token_accuracy": 0.24283964633941652,
      "num_tokens": 108600437.0,
      "step": 48640
    },
    {
      "entropy": 5.326772260665893,
      "epoch": 5.463300949064974,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002459414524356055,
      "loss": 4.8202,
      "mean_token_accuracy": 0.2341189667582512,
      "num_tokens": 108611302.0,
      "step": 48645
    },
    {
      "entropy": 5.371254014968872,
      "epoch": 5.463862525972932,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002459016404174613,
      "loss": 4.9295,
      "mean_token_accuracy": 0.23713070303201675,
      "num_tokens": 108622837.0,
      "step": 48650
    },
    {
      "entropy": 5.286772632598877,
      "epoch": 5.46442410288089,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00024586182932582014,
      "loss": 4.7873,
      "mean_token_accuracy": 0.23838101029396058,
      "num_tokens": 108635025.0,
      "step": 48655
    },
    {
      "entropy": 5.352306461334228,
      "epoch": 5.464985679788847,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0002458220191619496,
      "loss": 4.9161,
      "mean_token_accuracy": 0.23157168477773665,
      "num_tokens": 108647248.0,
      "step": 48660
    },
    {
      "entropy": 5.3630143165588375,
      "epoch": 5.465547256696805,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00024578220992711736,
      "loss": 4.8611,
      "mean_token_accuracy": 0.23149846792221068,
      "num_tokens": 108657677.0,
      "step": 48665
    },
    {
      "entropy": 5.327671051025391,
      "epoch": 5.466108833604762,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002457424016225909,
      "loss": 4.7306,
      "mean_token_accuracy": 0.23815231323242186,
      "num_tokens": 108668421.0,
      "step": 48670
    },
    {
      "entropy": 5.368958473205566,
      "epoch": 5.4666704105127195,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002457025942496376,
      "loss": 4.8378,
      "mean_token_accuracy": 0.23574099093675613,
      "num_tokens": 108680296.0,
      "step": 48675
    },
    {
      "entropy": 5.370146608352661,
      "epoch": 5.467231987420678,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024566278780952514,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2271747663617134,
      "num_tokens": 108691746.0,
      "step": 48680
    },
    {
      "entropy": 5.237767744064331,
      "epoch": 5.467793564328635,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002456229823035209,
      "loss": 4.7239,
      "mean_token_accuracy": 0.2435236319899559,
      "num_tokens": 108703420.0,
      "step": 48685
    },
    {
      "entropy": 5.335828733444214,
      "epoch": 5.468355141236592,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024558317773289227,
      "loss": 4.8176,
      "mean_token_accuracy": 0.2426401510834694,
      "num_tokens": 108714480.0,
      "step": 48690
    },
    {
      "entropy": 5.381621074676514,
      "epoch": 5.46891671814455,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002455433740989067,
      "loss": 4.8548,
      "mean_token_accuracy": 0.24158288687467575,
      "num_tokens": 108725004.0,
      "step": 48695
    },
    {
      "entropy": 5.3588653087615965,
      "epoch": 5.469478295052507,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002455035714028315,
      "loss": 4.8258,
      "mean_token_accuracy": 0.24195917546749116,
      "num_tokens": 108735526.0,
      "step": 48700
    },
    {
      "entropy": 5.3427516460418705,
      "epoch": 5.470039871960465,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002454637696459341,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22958477437496186,
      "num_tokens": 108746463.0,
      "step": 48705
    },
    {
      "entropy": 5.275126886367798,
      "epoch": 5.470601448868423,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024542396882948163,
      "loss": 4.7863,
      "mean_token_accuracy": 0.2359275460243225,
      "num_tokens": 108757485.0,
      "step": 48710
    },
    {
      "entropy": 5.308712530136108,
      "epoch": 5.47116302577638,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024538416895474144,
      "loss": 4.8337,
      "mean_token_accuracy": 0.23082361221313477,
      "num_tokens": 108768549.0,
      "step": 48715
    },
    {
      "entropy": 5.393625926971436,
      "epoch": 5.471724602684338,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024534437002298083,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2368304431438446,
      "num_tokens": 108778614.0,
      "step": 48720
    },
    {
      "entropy": 5.264206743240356,
      "epoch": 5.472286179592295,
      "grad_norm": 1.140625,
      "learning_rate": 0.000245304572035467,
      "loss": 4.7127,
      "mean_token_accuracy": 0.24635663032531738,
      "num_tokens": 108789719.0,
      "step": 48725
    },
    {
      "entropy": 5.282946825027466,
      "epoch": 5.4728477565002525,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002452647749934671,
      "loss": 4.8123,
      "mean_token_accuracy": 0.24256884902715684,
      "num_tokens": 108801202.0,
      "step": 48730
    },
    {
      "entropy": 5.299004602432251,
      "epoch": 5.473409333408211,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024522497889824837,
      "loss": 4.8646,
      "mean_token_accuracy": 0.23303425908088685,
      "num_tokens": 108812171.0,
      "step": 48735
    },
    {
      "entropy": 5.311777925491333,
      "epoch": 5.473970910316168,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002451851837510778,
      "loss": 4.7637,
      "mean_token_accuracy": 0.24300806522369384,
      "num_tokens": 108823213.0,
      "step": 48740
    },
    {
      "entropy": 5.33276743888855,
      "epoch": 5.474532487224125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002451453895532226,
      "loss": 4.8068,
      "mean_token_accuracy": 0.23879633098840714,
      "num_tokens": 108833817.0,
      "step": 48745
    },
    {
      "entropy": 5.411648178100586,
      "epoch": 5.475094064132083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002451055963059496,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2294970706105232,
      "num_tokens": 108845028.0,
      "step": 48750
    },
    {
      "entropy": 5.390277099609375,
      "epoch": 5.47565564104004,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024506580401052616,
      "loss": 4.8427,
      "mean_token_accuracy": 0.23343038707971572,
      "num_tokens": 108857483.0,
      "step": 48755
    },
    {
      "entropy": 5.278283548355103,
      "epoch": 5.4762172179479975,
      "grad_norm": 1.0,
      "learning_rate": 0.0002450260126682191,
      "loss": 4.7886,
      "mean_token_accuracy": 0.23112231194972993,
      "num_tokens": 108868264.0,
      "step": 48760
    },
    {
      "entropy": 5.343395090103149,
      "epoch": 5.476778794855956,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00024498622228029545,
      "loss": 4.8443,
      "mean_token_accuracy": 0.2352883443236351,
      "num_tokens": 108879053.0,
      "step": 48765
    },
    {
      "entropy": 5.380060768127441,
      "epoch": 5.477340371763913,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002449464328480221,
      "loss": 4.8304,
      "mean_token_accuracy": 0.24751342386007308,
      "num_tokens": 108889015.0,
      "step": 48770
    },
    {
      "entropy": 5.319025087356567,
      "epoch": 5.477901948671871,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000244906644372666,
      "loss": 4.8212,
      "mean_token_accuracy": 0.23443738073110582,
      "num_tokens": 108901171.0,
      "step": 48775
    },
    {
      "entropy": 5.314834690093994,
      "epoch": 5.478463525579828,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000244866856855494,
      "loss": 4.8106,
      "mean_token_accuracy": 0.24007904082536696,
      "num_tokens": 108912052.0,
      "step": 48780
    },
    {
      "entropy": 5.303533124923706,
      "epoch": 5.479025102487785,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024482707029777284,
      "loss": 4.7929,
      "mean_token_accuracy": 0.2346143364906311,
      "num_tokens": 108922348.0,
      "step": 48785
    },
    {
      "entropy": 5.3548253059387205,
      "epoch": 5.4795866793957435,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024478728470076963,
      "loss": 4.8666,
      "mean_token_accuracy": 0.2306789994239807,
      "num_tokens": 108933479.0,
      "step": 48790
    },
    {
      "entropy": 5.2497961044311525,
      "epoch": 5.480148256303701,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002447475000657509,
      "loss": 4.6809,
      "mean_token_accuracy": 0.24221818149089813,
      "num_tokens": 108944040.0,
      "step": 48795
    },
    {
      "entropy": 5.364357948303223,
      "epoch": 5.480709833211658,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002447077163939836,
      "loss": 4.8855,
      "mean_token_accuracy": 0.23175972700119019,
      "num_tokens": 108955379.0,
      "step": 48800
    },
    {
      "entropy": 5.374868392944336,
      "epoch": 5.481271410119616,
      "grad_norm": 1.125,
      "learning_rate": 0.00024466793368673427,
      "loss": 4.8768,
      "mean_token_accuracy": 0.23078415542840958,
      "num_tokens": 108966082.0,
      "step": 48805
    },
    {
      "entropy": 5.402729797363281,
      "epoch": 5.481832987027573,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002446281519452698,
      "loss": 4.857,
      "mean_token_accuracy": 0.24385218769311906,
      "num_tokens": 108976523.0,
      "step": 48810
    },
    {
      "entropy": 5.28530535697937,
      "epoch": 5.482394563935531,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002445883711708567,
      "loss": 4.8023,
      "mean_token_accuracy": 0.2395727425813675,
      "num_tokens": 108988149.0,
      "step": 48815
    },
    {
      "entropy": 5.332957649230957,
      "epoch": 5.482956140843489,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024454859136476155,
      "loss": 4.7979,
      "mean_token_accuracy": 0.24033304452896118,
      "num_tokens": 109000506.0,
      "step": 48820
    },
    {
      "entropy": 5.319605159759521,
      "epoch": 5.483517717751446,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024450881252825117,
      "loss": 4.748,
      "mean_token_accuracy": 0.24625588655471803,
      "num_tokens": 109011005.0,
      "step": 48825
    },
    {
      "entropy": 5.33412733078003,
      "epoch": 5.484079294659404,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000244469034662592,
      "loss": 4.7862,
      "mean_token_accuracy": 0.23482610285282135,
      "num_tokens": 109021005.0,
      "step": 48830
    },
    {
      "entropy": 5.339501810073853,
      "epoch": 5.484640871567361,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002444292577690506,
      "loss": 4.8181,
      "mean_token_accuracy": 0.2366468518972397,
      "num_tokens": 109032002.0,
      "step": 48835
    },
    {
      "entropy": 5.243677949905395,
      "epoch": 5.485202448475318,
      "grad_norm": 0.953125,
      "learning_rate": 0.00024438948184889354,
      "loss": 4.7384,
      "mean_token_accuracy": 0.24229494333267212,
      "num_tokens": 109043607.0,
      "step": 48840
    },
    {
      "entropy": 5.3579261779785154,
      "epoch": 5.4857640253832765,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002443497069033872,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22986946403980255,
      "num_tokens": 109055480.0,
      "step": 48845
    },
    {
      "entropy": 5.387786483764648,
      "epoch": 5.486325602291234,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002443099329337981,
      "loss": 4.9046,
      "mean_token_accuracy": 0.23804231286048888,
      "num_tokens": 109066684.0,
      "step": 48850
    },
    {
      "entropy": 5.339282703399658,
      "epoch": 5.486887179199192,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002442701599413927,
      "loss": 4.7469,
      "mean_token_accuracy": 0.24777586460113527,
      "num_tokens": 109075973.0,
      "step": 48855
    },
    {
      "entropy": 5.3363128185272215,
      "epoch": 5.487448756107149,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024423038792743715,
      "loss": 4.8165,
      "mean_token_accuracy": 0.23986100852489473,
      "num_tokens": 109087652.0,
      "step": 48860
    },
    {
      "entropy": 5.2809813022613525,
      "epoch": 5.488010333015106,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0002441906168931981,
      "loss": 4.7519,
      "mean_token_accuracy": 0.2430313929915428,
      "num_tokens": 109100315.0,
      "step": 48865
    },
    {
      "entropy": 5.301583290100098,
      "epoch": 5.488571909923064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002441508468399417,
      "loss": 4.8406,
      "mean_token_accuracy": 0.2409450814127922,
      "num_tokens": 109112005.0,
      "step": 48870
    },
    {
      "entropy": 5.328284502029419,
      "epoch": 5.4891334868310215,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024411107776893437,
      "loss": 4.8752,
      "mean_token_accuracy": 0.22999443411827086,
      "num_tokens": 109122216.0,
      "step": 48875
    },
    {
      "entropy": 5.3141122341156,
      "epoch": 5.489695063738979,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002440713096814423,
      "loss": 4.764,
      "mean_token_accuracy": 0.2359127327799797,
      "num_tokens": 109133073.0,
      "step": 48880
    },
    {
      "entropy": 5.3149340629577635,
      "epoch": 5.490256640646937,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024403154257873172,
      "loss": 4.7629,
      "mean_token_accuracy": 0.2432608962059021,
      "num_tokens": 109144097.0,
      "step": 48885
    },
    {
      "entropy": 5.345896434783936,
      "epoch": 5.490818217554894,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024399177646206884,
      "loss": 4.8463,
      "mean_token_accuracy": 0.23025868982076644,
      "num_tokens": 109155246.0,
      "step": 48890
    },
    {
      "entropy": 5.34710578918457,
      "epoch": 5.491379794462851,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00024395201133271978,
      "loss": 4.7801,
      "mean_token_accuracy": 0.24016969054937362,
      "num_tokens": 109165955.0,
      "step": 48895
    },
    {
      "entropy": 5.353817129135132,
      "epoch": 5.491941371370809,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024391224719195077,
      "loss": 4.8209,
      "mean_token_accuracy": 0.23630001842975618,
      "num_tokens": 109176471.0,
      "step": 48900
    },
    {
      "entropy": 5.341941976547242,
      "epoch": 5.492502948278767,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024387248404102779,
      "loss": 4.754,
      "mean_token_accuracy": 0.24674094021320342,
      "num_tokens": 109188613.0,
      "step": 48905
    },
    {
      "entropy": 5.427220344543457,
      "epoch": 5.493064525186725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024383272188121708,
      "loss": 4.8726,
      "mean_token_accuracy": 0.23631695806980133,
      "num_tokens": 109200652.0,
      "step": 48910
    },
    {
      "entropy": 5.322189569473267,
      "epoch": 5.493626102094682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002437929607137846,
      "loss": 4.806,
      "mean_token_accuracy": 0.2339221403002739,
      "num_tokens": 109210858.0,
      "step": 48915
    },
    {
      "entropy": 5.32694263458252,
      "epoch": 5.494187679002639,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00024375320053999627,
      "loss": 4.8035,
      "mean_token_accuracy": 0.236537866294384,
      "num_tokens": 109222058.0,
      "step": 48920
    },
    {
      "entropy": 5.315470266342163,
      "epoch": 5.494749255910597,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024371344136111823,
      "loss": 4.7743,
      "mean_token_accuracy": 0.23928722143173217,
      "num_tokens": 109232831.0,
      "step": 48925
    },
    {
      "entropy": 5.354319095611572,
      "epoch": 5.4953108328185545,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002436736831784162,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22688726037740709,
      "num_tokens": 109245203.0,
      "step": 48930
    },
    {
      "entropy": 5.282226467132569,
      "epoch": 5.495872409726512,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00024363392599315636,
      "loss": 4.7441,
      "mean_token_accuracy": 0.23904580175876616,
      "num_tokens": 109255738.0,
      "step": 48935
    },
    {
      "entropy": 5.390476989746094,
      "epoch": 5.49643398663447,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024359416980660454,
      "loss": 4.842,
      "mean_token_accuracy": 0.2378416731953621,
      "num_tokens": 109266567.0,
      "step": 48940
    },
    {
      "entropy": 5.332048940658569,
      "epoch": 5.496995563542427,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024355441462002648,
      "loss": 4.8547,
      "mean_token_accuracy": 0.2303945690393448,
      "num_tokens": 109277302.0,
      "step": 48945
    },
    {
      "entropy": 5.351855897903443,
      "epoch": 5.497557140450384,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024351466043468807,
      "loss": 4.8689,
      "mean_token_accuracy": 0.23205618113279342,
      "num_tokens": 109288336.0,
      "step": 48950
    },
    {
      "entropy": 5.313169527053833,
      "epoch": 5.498118717358342,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024347490725185506,
      "loss": 4.7477,
      "mean_token_accuracy": 0.23876751512289046,
      "num_tokens": 109298968.0,
      "step": 48955
    },
    {
      "entropy": 5.281483316421509,
      "epoch": 5.4986802942663,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00024343515507279325,
      "loss": 4.7333,
      "mean_token_accuracy": 0.24335992634296416,
      "num_tokens": 109308837.0,
      "step": 48960
    },
    {
      "entropy": 5.331518173217773,
      "epoch": 5.499241871174258,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002433954038987683,
      "loss": 4.8017,
      "mean_token_accuracy": 0.24409158080816268,
      "num_tokens": 109319464.0,
      "step": 48965
    },
    {
      "entropy": 5.321038389205933,
      "epoch": 5.499803448082215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024335565373104594,
      "loss": 4.7883,
      "mean_token_accuracy": 0.2379003331065178,
      "num_tokens": 109331126.0,
      "step": 48970
    },
    {
      "entropy": 5.274481010437012,
      "epoch": 5.500365024990172,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024331590457089193,
      "loss": 4.791,
      "mean_token_accuracy": 0.23924760520458221,
      "num_tokens": 109341298.0,
      "step": 48975
    },
    {
      "entropy": 5.341129112243652,
      "epoch": 5.50092660189813,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00024327615641957175,
      "loss": 4.8045,
      "mean_token_accuracy": 0.25015909522771834,
      "num_tokens": 109352817.0,
      "step": 48980
    },
    {
      "entropy": 5.311898136138916,
      "epoch": 5.501488178806087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024323640927835106,
      "loss": 4.7414,
      "mean_token_accuracy": 0.24486693292856215,
      "num_tokens": 109364004.0,
      "step": 48985
    },
    {
      "entropy": 5.281471490859985,
      "epoch": 5.5020497557140455,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00024319666314849542,
      "loss": 4.7342,
      "mean_token_accuracy": 0.2504726752638817,
      "num_tokens": 109375790.0,
      "step": 48990
    },
    {
      "entropy": 5.2982776165008545,
      "epoch": 5.502611332622003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024315691803127037,
      "loss": 4.8064,
      "mean_token_accuracy": 0.2361719638109207,
      "num_tokens": 109386521.0,
      "step": 48995
    },
    {
      "entropy": 5.346166133880615,
      "epoch": 5.50317290952996,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002431171739279413,
      "loss": 4.8478,
      "mean_token_accuracy": 0.23551374524831772,
      "num_tokens": 109397348.0,
      "step": 49000
    },
    {
      "entropy": 5.301602745056153,
      "epoch": 5.503734486437917,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024307743083977384,
      "loss": 4.7687,
      "mean_token_accuracy": 0.24105747789144516,
      "num_tokens": 109407636.0,
      "step": 49005
    },
    {
      "entropy": 5.3048927783966064,
      "epoch": 5.504296063345875,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024303768876803345,
      "loss": 4.8061,
      "mean_token_accuracy": 0.23726456761360168,
      "num_tokens": 109417822.0,
      "step": 49010
    },
    {
      "entropy": 5.3173072814941404,
      "epoch": 5.5048576402538325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024299794771398536,
      "loss": 4.7792,
      "mean_token_accuracy": 0.24135462939739227,
      "num_tokens": 109428602.0,
      "step": 49015
    },
    {
      "entropy": 5.330592250823974,
      "epoch": 5.505419217161791,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00024295820767889504,
      "loss": 4.7505,
      "mean_token_accuracy": 0.24390156716108322,
      "num_tokens": 109439042.0,
      "step": 49020
    },
    {
      "entropy": 5.4125463485717775,
      "epoch": 5.505980794069748,
      "grad_norm": 0.9375,
      "learning_rate": 0.00024291846866402784,
      "loss": 4.8294,
      "mean_token_accuracy": 0.23565893322229386,
      "num_tokens": 109450286.0,
      "step": 49025
    },
    {
      "entropy": 5.405391216278076,
      "epoch": 5.506542370977705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024287873067064904,
      "loss": 4.9021,
      "mean_token_accuracy": 0.2328319013118744,
      "num_tokens": 109461852.0,
      "step": 49030
    },
    {
      "entropy": 5.320379161834717,
      "epoch": 5.507103947885663,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00024283899370002393,
      "loss": 4.836,
      "mean_token_accuracy": 0.23690219968557358,
      "num_tokens": 109474340.0,
      "step": 49035
    },
    {
      "entropy": 5.335454416275025,
      "epoch": 5.50766552479362,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024279925775341765,
      "loss": 4.7873,
      "mean_token_accuracy": 0.23608896881341934,
      "num_tokens": 109485448.0,
      "step": 49040
    },
    {
      "entropy": 5.32970404624939,
      "epoch": 5.5082271017015785,
      "grad_norm": 1.1875,
      "learning_rate": 0.00024275952283209557,
      "loss": 4.728,
      "mean_token_accuracy": 0.24419861137866974,
      "num_tokens": 109494750.0,
      "step": 49045
    },
    {
      "entropy": 5.347968816757202,
      "epoch": 5.508788678609536,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024271978893732277,
      "loss": 4.7578,
      "mean_token_accuracy": 0.24270141869783401,
      "num_tokens": 109505372.0,
      "step": 49050
    },
    {
      "entropy": 5.34378080368042,
      "epoch": 5.509350255517493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00024268005607036442,
      "loss": 4.8527,
      "mean_token_accuracy": 0.23286802023649217,
      "num_tokens": 109516881.0,
      "step": 49055
    },
    {
      "entropy": 5.283629655838013,
      "epoch": 5.509911832425451,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024264032423248555,
      "loss": 4.7117,
      "mean_token_accuracy": 0.24494908452033998,
      "num_tokens": 109527285.0,
      "step": 49060
    },
    {
      "entropy": 5.309141206741333,
      "epoch": 5.510473409333408,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002426005934249514,
      "loss": 4.795,
      "mean_token_accuracy": 0.24027348011732103,
      "num_tokens": 109537635.0,
      "step": 49065
    },
    {
      "entropy": 5.268564987182617,
      "epoch": 5.5110349862413655,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002425608636490269,
      "loss": 4.7638,
      "mean_token_accuracy": 0.23503388166427613,
      "num_tokens": 109547932.0,
      "step": 49070
    },
    {
      "entropy": 5.2955889225006105,
      "epoch": 5.511596563149324,
      "grad_norm": 1.0,
      "learning_rate": 0.00024252113490597692,
      "loss": 4.7863,
      "mean_token_accuracy": 0.24987515956163406,
      "num_tokens": 109559087.0,
      "step": 49075
    },
    {
      "entropy": 5.364214706420898,
      "epoch": 5.512158140057281,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002424814071970668,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2345846876502037,
      "num_tokens": 109570926.0,
      "step": 49080
    },
    {
      "entropy": 5.25468430519104,
      "epoch": 5.512719716965238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024244168052356125,
      "loss": 4.6861,
      "mean_token_accuracy": 0.24319399744272233,
      "num_tokens": 109580988.0,
      "step": 49085
    },
    {
      "entropy": 5.404922389984131,
      "epoch": 5.513281293873196,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024240195488672516,
      "loss": 4.8885,
      "mean_token_accuracy": 0.23046984672546386,
      "num_tokens": 109592763.0,
      "step": 49090
    },
    {
      "entropy": 5.359957218170166,
      "epoch": 5.513842870781153,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024236223028782357,
      "loss": 4.8855,
      "mean_token_accuracy": 0.23193428218364714,
      "num_tokens": 109603786.0,
      "step": 49095
    },
    {
      "entropy": 5.32365312576294,
      "epoch": 5.514404447689111,
      "grad_norm": 1.125,
      "learning_rate": 0.00024232250672812116,
      "loss": 4.8002,
      "mean_token_accuracy": 0.25086279809474943,
      "num_tokens": 109615321.0,
      "step": 49100
    },
    {
      "entropy": 5.393411111831665,
      "epoch": 5.514966024597069,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00024228278420888283,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2372048020362854,
      "num_tokens": 109625758.0,
      "step": 49105
    },
    {
      "entropy": 5.308790588378907,
      "epoch": 5.515527601505026,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002422430627313733,
      "loss": 4.7448,
      "mean_token_accuracy": 0.23948934674263,
      "num_tokens": 109636725.0,
      "step": 49110
    },
    {
      "entropy": 5.309588289260864,
      "epoch": 5.516089178412984,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00024220334229685736,
      "loss": 4.7654,
      "mean_token_accuracy": 0.24142092764377593,
      "num_tokens": 109647179.0,
      "step": 49115
    },
    {
      "entropy": 5.309453010559082,
      "epoch": 5.516650755320941,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024216362290659977,
      "loss": 4.7953,
      "mean_token_accuracy": 0.2340933695435524,
      "num_tokens": 109658093.0,
      "step": 49120
    },
    {
      "entropy": 5.343444538116455,
      "epoch": 5.517212332228898,
      "grad_norm": 1.171875,
      "learning_rate": 0.00024212390456186512,
      "loss": 4.8111,
      "mean_token_accuracy": 0.24369694143533707,
      "num_tokens": 109668574.0,
      "step": 49125
    },
    {
      "entropy": 5.362764739990235,
      "epoch": 5.5177739091368565,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024208418726391813,
      "loss": 4.83,
      "mean_token_accuracy": 0.2351192504167557,
      "num_tokens": 109678640.0,
      "step": 49130
    },
    {
      "entropy": 5.343158102035522,
      "epoch": 5.518335486044814,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00024204447101402337,
      "loss": 4.8146,
      "mean_token_accuracy": 0.23371340185403824,
      "num_tokens": 109690183.0,
      "step": 49135
    },
    {
      "entropy": 5.282980442047119,
      "epoch": 5.518897062952771,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024200475581344545,
      "loss": 4.7573,
      "mean_token_accuracy": 0.24055927246809006,
      "num_tokens": 109700882.0,
      "step": 49140
    },
    {
      "entropy": 5.2862756729125975,
      "epoch": 5.519458639860729,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002419650416634488,
      "loss": 4.7809,
      "mean_token_accuracy": 0.23804033547639847,
      "num_tokens": 109712459.0,
      "step": 49145
    },
    {
      "entropy": 5.252402305603027,
      "epoch": 5.520020216768686,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024192532856529814,
      "loss": 4.7578,
      "mean_token_accuracy": 0.2422614425420761,
      "num_tokens": 109724116.0,
      "step": 49150
    },
    {
      "entropy": 5.2878961086273195,
      "epoch": 5.520581793676644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024188561652025787,
      "loss": 4.7933,
      "mean_token_accuracy": 0.23394181877374648,
      "num_tokens": 109735517.0,
      "step": 49155
    },
    {
      "entropy": 5.319989109039307,
      "epoch": 5.521143370584602,
      "grad_norm": 1.0,
      "learning_rate": 0.00024184590552959234,
      "loss": 4.8185,
      "mean_token_accuracy": 0.23697317242622376,
      "num_tokens": 109746762.0,
      "step": 49160
    },
    {
      "entropy": 5.280418539047242,
      "epoch": 5.521704947492559,
      "grad_norm": 1.0,
      "learning_rate": 0.00024180619559456608,
      "loss": 4.7453,
      "mean_token_accuracy": 0.23652812093496323,
      "num_tokens": 109758153.0,
      "step": 49165
    },
    {
      "entropy": 5.402957820892334,
      "epoch": 5.522266524400517,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002417664867164434,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22653261870145797,
      "num_tokens": 109770521.0,
      "step": 49170
    },
    {
      "entropy": 5.355305337905884,
      "epoch": 5.522828101308474,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002417267788964887,
      "loss": 4.8231,
      "mean_token_accuracy": 0.237220162153244,
      "num_tokens": 109781816.0,
      "step": 49175
    },
    {
      "entropy": 5.391757822036743,
      "epoch": 5.523389678216432,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024168707213596626,
      "loss": 4.8225,
      "mean_token_accuracy": 0.2336422771215439,
      "num_tokens": 109792558.0,
      "step": 49180
    },
    {
      "entropy": 5.277734327316284,
      "epoch": 5.5239512551243894,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00024164736643614033,
      "loss": 4.7098,
      "mean_token_accuracy": 0.23852664828300477,
      "num_tokens": 109803021.0,
      "step": 49185
    },
    {
      "entropy": 5.155351543426514,
      "epoch": 5.524512832032347,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002416076617982752,
      "loss": 4.727,
      "mean_token_accuracy": 0.2421427324414253,
      "num_tokens": 109813526.0,
      "step": 49190
    },
    {
      "entropy": 5.287590074539184,
      "epoch": 5.525074408940304,
      "grad_norm": 1.1875,
      "learning_rate": 0.00024156795822363508,
      "loss": 4.729,
      "mean_token_accuracy": 0.23774438947439194,
      "num_tokens": 109824411.0,
      "step": 49195
    },
    {
      "entropy": 5.340789031982422,
      "epoch": 5.525635985848262,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00024152825571348418,
      "loss": 4.7774,
      "mean_token_accuracy": 0.2443556934595108,
      "num_tokens": 109834908.0,
      "step": 49200
    },
    {
      "entropy": 5.220906019210815,
      "epoch": 5.526197562756219,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024148855426908656,
      "loss": 4.7511,
      "mean_token_accuracy": 0.240874020755291,
      "num_tokens": 109846332.0,
      "step": 49205
    },
    {
      "entropy": 5.333557939529419,
      "epoch": 5.526759139664177,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002414488538917064,
      "loss": 4.9087,
      "mean_token_accuracy": 0.23265471011400224,
      "num_tokens": 109857507.0,
      "step": 49210
    },
    {
      "entropy": 5.308560037612915,
      "epoch": 5.5273207165721345,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002414091545826077,
      "loss": 4.7702,
      "mean_token_accuracy": 0.2392491042613983,
      "num_tokens": 109869196.0,
      "step": 49215
    },
    {
      "entropy": 5.377552604675293,
      "epoch": 5.527882293480092,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024136945634305445,
      "loss": 4.8583,
      "mean_token_accuracy": 0.2277148887515068,
      "num_tokens": 109879894.0,
      "step": 49220
    },
    {
      "entropy": 5.199643945693969,
      "epoch": 5.52844387038805,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024132975917431088,
      "loss": 4.7061,
      "mean_token_accuracy": 0.24454743415117264,
      "num_tokens": 109890673.0,
      "step": 49225
    },
    {
      "entropy": 5.326724624633789,
      "epoch": 5.529005447296007,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024129006307764084,
      "loss": 4.8665,
      "mean_token_accuracy": 0.22981595546007155,
      "num_tokens": 109900931.0,
      "step": 49230
    },
    {
      "entropy": 5.346473312377929,
      "epoch": 5.529567024203965,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024125036805430829,
      "loss": 4.8238,
      "mean_token_accuracy": 0.24523911625146866,
      "num_tokens": 109913172.0,
      "step": 49235
    },
    {
      "entropy": 5.354193019866943,
      "epoch": 5.530128601111922,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000241210674105577,
      "loss": 4.8275,
      "mean_token_accuracy": 0.24880130290985109,
      "num_tokens": 109925116.0,
      "step": 49240
    },
    {
      "entropy": 5.255893516540527,
      "epoch": 5.53069017801988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024117098123271107,
      "loss": 4.6912,
      "mean_token_accuracy": 0.2533883646130562,
      "num_tokens": 109935201.0,
      "step": 49245
    },
    {
      "entropy": 5.3175910949707035,
      "epoch": 5.531251754927838,
      "grad_norm": 1.0,
      "learning_rate": 0.00024113128943697415,
      "loss": 4.8103,
      "mean_token_accuracy": 0.24125425666570663,
      "num_tokens": 109945679.0,
      "step": 49250
    },
    {
      "entropy": 5.334803581237793,
      "epoch": 5.531813331835795,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002410915987196301,
      "loss": 4.7978,
      "mean_token_accuracy": 0.23614946454763414,
      "num_tokens": 109956659.0,
      "step": 49255
    },
    {
      "entropy": 5.3123414516448975,
      "epoch": 5.532374908743752,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024105190908194275,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23542602658271788,
      "num_tokens": 109966371.0,
      "step": 49260
    },
    {
      "entropy": 5.311874532699585,
      "epoch": 5.53293648565171,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00024101222052517573,
      "loss": 4.8339,
      "mean_token_accuracy": 0.23439778834581376,
      "num_tokens": 109977630.0,
      "step": 49265
    },
    {
      "entropy": 5.252944850921631,
      "epoch": 5.5334980625596675,
      "grad_norm": 0.96875,
      "learning_rate": 0.00024097253305059287,
      "loss": 4.7459,
      "mean_token_accuracy": 0.23519934713840485,
      "num_tokens": 109988757.0,
      "step": 49270
    },
    {
      "entropy": 5.344314861297607,
      "epoch": 5.534059639467625,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002409328466594577,
      "loss": 4.8116,
      "mean_token_accuracy": 0.23522135019302368,
      "num_tokens": 109998992.0,
      "step": 49275
    },
    {
      "entropy": 5.278532981872559,
      "epoch": 5.534621216375583,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002408931613530339,
      "loss": 4.7095,
      "mean_token_accuracy": 0.2486620545387268,
      "num_tokens": 110009961.0,
      "step": 49280
    },
    {
      "entropy": 5.283925151824951,
      "epoch": 5.53518279328354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002408534771325851,
      "loss": 4.7719,
      "mean_token_accuracy": 0.24627807438373567,
      "num_tokens": 110020653.0,
      "step": 49285
    },
    {
      "entropy": 5.3315126419067385,
      "epoch": 5.535744370191498,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002408137939993747,
      "loss": 4.8163,
      "mean_token_accuracy": 0.23727014064788818,
      "num_tokens": 110030113.0,
      "step": 49290
    },
    {
      "entropy": 5.282347393035889,
      "epoch": 5.536305947099455,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024077411195466648,
      "loss": 4.8096,
      "mean_token_accuracy": 0.23876769840717316,
      "num_tokens": 110040952.0,
      "step": 49295
    },
    {
      "entropy": 5.253788757324219,
      "epoch": 5.5368675240074126,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002407344309997238,
      "loss": 4.7457,
      "mean_token_accuracy": 0.24085484445095062,
      "num_tokens": 110052861.0,
      "step": 49300
    },
    {
      "entropy": 5.305135011672974,
      "epoch": 5.537429100915371,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00024069475113581014,
      "loss": 4.8111,
      "mean_token_accuracy": 0.23482576310634612,
      "num_tokens": 110064974.0,
      "step": 49305
    },
    {
      "entropy": 5.304819822311401,
      "epoch": 5.537990677823328,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024065507236418888,
      "loss": 4.7897,
      "mean_token_accuracy": 0.24035324454307555,
      "num_tokens": 110076587.0,
      "step": 49310
    },
    {
      "entropy": 5.440578126907349,
      "epoch": 5.538552254731285,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00024061539468612348,
      "loss": 4.9047,
      "mean_token_accuracy": 0.23179208636283874,
      "num_tokens": 110089036.0,
      "step": 49315
    },
    {
      "entropy": 5.457334041595459,
      "epoch": 5.539113831639243,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024057571810287722,
      "loss": 4.9738,
      "mean_token_accuracy": 0.23648951053619385,
      "num_tokens": 110100253.0,
      "step": 49320
    },
    {
      "entropy": 5.351904249191284,
      "epoch": 5.5396754085472,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024053604261571338,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22847194969654083,
      "num_tokens": 110112567.0,
      "step": 49325
    },
    {
      "entropy": 5.373964023590088,
      "epoch": 5.540236985455158,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024049636822589537,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2392500653862953,
      "num_tokens": 110123623.0,
      "step": 49330
    },
    {
      "entropy": 5.41777720451355,
      "epoch": 5.540798562363116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002404566949346864,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23097027093172073,
      "num_tokens": 110134367.0,
      "step": 49335
    },
    {
      "entropy": 5.315137720108032,
      "epoch": 5.541360139271073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002404170227433497,
      "loss": 4.7895,
      "mean_token_accuracy": 0.23755075335502623,
      "num_tokens": 110144305.0,
      "step": 49340
    },
    {
      "entropy": 5.420220565795899,
      "epoch": 5.541921716179031,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00024037735165314834,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22901684939861297,
      "num_tokens": 110155682.0,
      "step": 49345
    },
    {
      "entropy": 5.348816633224487,
      "epoch": 5.542483293086988,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024033768166534553,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23065080493688583,
      "num_tokens": 110166175.0,
      "step": 49350
    },
    {
      "entropy": 5.297728729248047,
      "epoch": 5.5430448699949455,
      "grad_norm": 1.125,
      "learning_rate": 0.00024029801278120438,
      "loss": 4.8349,
      "mean_token_accuracy": 0.2380100294947624,
      "num_tokens": 110175930.0,
      "step": 49355
    },
    {
      "entropy": 5.213327693939209,
      "epoch": 5.543606446902904,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002402583450019879,
      "loss": 4.6717,
      "mean_token_accuracy": 0.24822961837053298,
      "num_tokens": 110187093.0,
      "step": 49360
    },
    {
      "entropy": 5.330848264694214,
      "epoch": 5.544168023810861,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024021867832895927,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2269832357764244,
      "num_tokens": 110197339.0,
      "step": 49365
    },
    {
      "entropy": 5.3579676151275635,
      "epoch": 5.544729600718819,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002401790127633814,
      "loss": 4.8456,
      "mean_token_accuracy": 0.22706105411052704,
      "num_tokens": 110207179.0,
      "step": 49370
    },
    {
      "entropy": 5.35638108253479,
      "epoch": 5.545291177626776,
      "grad_norm": 0.953125,
      "learning_rate": 0.00024013934830651728,
      "loss": 4.787,
      "mean_token_accuracy": 0.2369152694940567,
      "num_tokens": 110218150.0,
      "step": 49375
    },
    {
      "entropy": 5.454430007934571,
      "epoch": 5.545852754534733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024009968495962986,
      "loss": 4.9254,
      "mean_token_accuracy": 0.23092624098062514,
      "num_tokens": 110230454.0,
      "step": 49380
    },
    {
      "entropy": 5.290852832794189,
      "epoch": 5.546414331442691,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024006002272398199,
      "loss": 4.7643,
      "mean_token_accuracy": 0.2394581601023674,
      "num_tokens": 110241587.0,
      "step": 49385
    },
    {
      "entropy": 5.34937744140625,
      "epoch": 5.546975908350649,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002400203616008365,
      "loss": 4.8877,
      "mean_token_accuracy": 0.2328889861702919,
      "num_tokens": 110253858.0,
      "step": 49390
    },
    {
      "entropy": 5.266680002212524,
      "epoch": 5.547537485258606,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002399807015914563,
      "loss": 4.6992,
      "mean_token_accuracy": 0.24325504004955292,
      "num_tokens": 110265764.0,
      "step": 49395
    },
    {
      "entropy": 5.3679900646209715,
      "epoch": 5.548099062166564,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002399410426971041,
      "loss": 4.8452,
      "mean_token_accuracy": 0.23989685326814653,
      "num_tokens": 110277279.0,
      "step": 49400
    },
    {
      "entropy": 5.283869886398316,
      "epoch": 5.548660639074521,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023990138491904273,
      "loss": 4.7592,
      "mean_token_accuracy": 0.24054586589336396,
      "num_tokens": 110288509.0,
      "step": 49405
    },
    {
      "entropy": 5.360441446304321,
      "epoch": 5.5492222159824784,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023986172825853492,
      "loss": 4.8718,
      "mean_token_accuracy": 0.23751518577337266,
      "num_tokens": 110300678.0,
      "step": 49410
    },
    {
      "entropy": 5.323452806472778,
      "epoch": 5.5497837928904366,
      "grad_norm": 0.90625,
      "learning_rate": 0.00023982207271684326,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23554371744394303,
      "num_tokens": 110312785.0,
      "step": 49415
    },
    {
      "entropy": 5.316436672210694,
      "epoch": 5.550345369798394,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002397824182952305,
      "loss": 4.8204,
      "mean_token_accuracy": 0.24469601809978486,
      "num_tokens": 110324706.0,
      "step": 49420
    },
    {
      "entropy": 5.34657883644104,
      "epoch": 5.550906946706352,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023974276499495918,
      "loss": 4.824,
      "mean_token_accuracy": 0.2339731991291046,
      "num_tokens": 110335650.0,
      "step": 49425
    },
    {
      "entropy": 5.376606273651123,
      "epoch": 5.551468523614309,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023970311281729185,
      "loss": 4.8698,
      "mean_token_accuracy": 0.23344421833753587,
      "num_tokens": 110346416.0,
      "step": 49430
    },
    {
      "entropy": 5.226602697372437,
      "epoch": 5.552030100522266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000239663461763491,
      "loss": 4.6335,
      "mean_token_accuracy": 0.2556192547082901,
      "num_tokens": 110356681.0,
      "step": 49435
    },
    {
      "entropy": 5.278290033340454,
      "epoch": 5.552591677430224,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002396238118348194,
      "loss": 4.7066,
      "mean_token_accuracy": 0.2411363571882248,
      "num_tokens": 110368865.0,
      "step": 49440
    },
    {
      "entropy": 5.273737192153931,
      "epoch": 5.553153254338182,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023958416303253933,
      "loss": 4.7184,
      "mean_token_accuracy": 0.23952889293432236,
      "num_tokens": 110378960.0,
      "step": 49445
    },
    {
      "entropy": 5.366466760635376,
      "epoch": 5.553714831246139,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00023954451535791323,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23172518014907836,
      "num_tokens": 110390006.0,
      "step": 49450
    },
    {
      "entropy": 5.256264543533325,
      "epoch": 5.554276408154097,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023950486881220353,
      "loss": 4.7491,
      "mean_token_accuracy": 0.23354048728942872,
      "num_tokens": 110401789.0,
      "step": 49455
    },
    {
      "entropy": 5.2916006565094,
      "epoch": 5.554837985062054,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023946522339667254,
      "loss": 4.7232,
      "mean_token_accuracy": 0.24264327436685562,
      "num_tokens": 110412469.0,
      "step": 49460
    },
    {
      "entropy": 5.331110191345215,
      "epoch": 5.555399561970011,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002394255791125826,
      "loss": 4.8544,
      "mean_token_accuracy": 0.2410809651017189,
      "num_tokens": 110423791.0,
      "step": 49465
    },
    {
      "entropy": 5.30524697303772,
      "epoch": 5.5559611388779695,
      "grad_norm": 0.984375,
      "learning_rate": 0.000239385935961196,
      "loss": 4.8075,
      "mean_token_accuracy": 0.23374055325984955,
      "num_tokens": 110434393.0,
      "step": 49470
    },
    {
      "entropy": 5.380861711502075,
      "epoch": 5.556522715785927,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023934629394377512,
      "loss": 4.8993,
      "mean_token_accuracy": 0.23241067230701445,
      "num_tokens": 110445207.0,
      "step": 49475
    },
    {
      "entropy": 5.431865501403808,
      "epoch": 5.557084292693885,
      "grad_norm": 1.0,
      "learning_rate": 0.000239306653061582,
      "loss": 4.789,
      "mean_token_accuracy": 0.24713585674762725,
      "num_tokens": 110457090.0,
      "step": 49480
    },
    {
      "entropy": 5.361879587173462,
      "epoch": 5.557645869601842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023926701331587885,
      "loss": 4.8535,
      "mean_token_accuracy": 0.23658012002706527,
      "num_tokens": 110468527.0,
      "step": 49485
    },
    {
      "entropy": 5.314646768569946,
      "epoch": 5.558207446509799,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00023922737470792794,
      "loss": 4.7816,
      "mean_token_accuracy": 0.24477563798427582,
      "num_tokens": 110478449.0,
      "step": 49490
    },
    {
      "entropy": 5.346352291107178,
      "epoch": 5.558769023417757,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002391877372389913,
      "loss": 4.8714,
      "mean_token_accuracy": 0.24028194099664688,
      "num_tokens": 110488126.0,
      "step": 49495
    },
    {
      "entropy": 5.2712665557861325,
      "epoch": 5.559330600325715,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023914810091033096,
      "loss": 4.7503,
      "mean_token_accuracy": 0.2486584559082985,
      "num_tokens": 110499593.0,
      "step": 49500
    },
    {
      "entropy": 5.310043478012085,
      "epoch": 5.559892177233672,
      "grad_norm": 1.0,
      "learning_rate": 0.0002391084657232089,
      "loss": 4.8094,
      "mean_token_accuracy": 0.23556599617004395,
      "num_tokens": 110511612.0,
      "step": 49505
    },
    {
      "entropy": 5.393037557601929,
      "epoch": 5.56045375414163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023906883167888727,
      "loss": 4.8702,
      "mean_token_accuracy": 0.2298495277762413,
      "num_tokens": 110524040.0,
      "step": 49510
    },
    {
      "entropy": 5.364755535125733,
      "epoch": 5.561015331049587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000239029198778628,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22401278614997863,
      "num_tokens": 110535475.0,
      "step": 49515
    },
    {
      "entropy": 5.34985842704773,
      "epoch": 5.561576907957544,
      "grad_norm": 0.984375,
      "learning_rate": 0.000238989567023693,
      "loss": 4.838,
      "mean_token_accuracy": 0.23736632019281387,
      "num_tokens": 110547149.0,
      "step": 49520
    },
    {
      "entropy": 5.38202748298645,
      "epoch": 5.562138484865502,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023894993641534418,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23934174478054046,
      "num_tokens": 110559378.0,
      "step": 49525
    },
    {
      "entropy": 5.310741567611695,
      "epoch": 5.56270006177346,
      "grad_norm": 1.125,
      "learning_rate": 0.00023891030695484333,
      "loss": 4.7991,
      "mean_token_accuracy": 0.243086838722229,
      "num_tokens": 110570537.0,
      "step": 49530
    },
    {
      "entropy": 5.311999225616455,
      "epoch": 5.563261638681418,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002388706786434523,
      "loss": 4.7882,
      "mean_token_accuracy": 0.23863174617290497,
      "num_tokens": 110583073.0,
      "step": 49535
    },
    {
      "entropy": 5.26105809211731,
      "epoch": 5.563823215589375,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002388310514824329,
      "loss": 4.74,
      "mean_token_accuracy": 0.24237012416124343,
      "num_tokens": 110594711.0,
      "step": 49540
    },
    {
      "entropy": 5.316193056106568,
      "epoch": 5.564384792497332,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00023879142547304677,
      "loss": 4.8258,
      "mean_token_accuracy": 0.23756435215473176,
      "num_tokens": 110605187.0,
      "step": 49545
    },
    {
      "entropy": 5.310779666900634,
      "epoch": 5.56494636940529,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023875180061655576,
      "loss": 4.8215,
      "mean_token_accuracy": 0.22946882396936416,
      "num_tokens": 110615623.0,
      "step": 49550
    },
    {
      "entropy": 5.249540567398071,
      "epoch": 5.5655079463132475,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023871217691422147,
      "loss": 4.7478,
      "mean_token_accuracy": 0.24620334506034852,
      "num_tokens": 110627476.0,
      "step": 49555
    },
    {
      "entropy": 5.286901426315308,
      "epoch": 5.566069523221206,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023867255436730555,
      "loss": 4.8115,
      "mean_token_accuracy": 0.23468062281608582,
      "num_tokens": 110640016.0,
      "step": 49560
    },
    {
      "entropy": 5.350117015838623,
      "epoch": 5.566631100129163,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002386329329770696,
      "loss": 4.7647,
      "mean_token_accuracy": 0.24277010709047317,
      "num_tokens": 110650577.0,
      "step": 49565
    },
    {
      "entropy": 5.399088191986084,
      "epoch": 5.56719267703712,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00023859331274477514,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2401089683175087,
      "num_tokens": 110661657.0,
      "step": 49570
    },
    {
      "entropy": 5.301662015914917,
      "epoch": 5.567754253945077,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023855369367168378,
      "loss": 4.7828,
      "mean_token_accuracy": 0.24306727796792985,
      "num_tokens": 110673116.0,
      "step": 49575
    },
    {
      "entropy": 5.355267763137817,
      "epoch": 5.568315830853035,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023851407575905682,
      "loss": 4.8789,
      "mean_token_accuracy": 0.23933662921190263,
      "num_tokens": 110683422.0,
      "step": 49580
    },
    {
      "entropy": 5.278424024581909,
      "epoch": 5.568877407760993,
      "grad_norm": 1.125,
      "learning_rate": 0.0002384744590081559,
      "loss": 4.7628,
      "mean_token_accuracy": 0.24453894793987274,
      "num_tokens": 110693407.0,
      "step": 49585
    },
    {
      "entropy": 5.343217849731445,
      "epoch": 5.569438984668951,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002384348434202424,
      "loss": 4.8395,
      "mean_token_accuracy": 0.2288582593202591,
      "num_tokens": 110704105.0,
      "step": 49590
    },
    {
      "entropy": 5.304334306716919,
      "epoch": 5.570000561576908,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023839522899657767,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2395038440823555,
      "num_tokens": 110714745.0,
      "step": 49595
    },
    {
      "entropy": 5.246079683303833,
      "epoch": 5.570562138484865,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023835561573842302,
      "loss": 4.6813,
      "mean_token_accuracy": 0.2532370090484619,
      "num_tokens": 110725955.0,
      "step": 49600
    },
    {
      "entropy": 5.346138668060303,
      "epoch": 5.571123715392823,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002383160036470398,
      "loss": 4.9154,
      "mean_token_accuracy": 0.232114277780056,
      "num_tokens": 110737023.0,
      "step": 49605
    },
    {
      "entropy": 5.343060302734375,
      "epoch": 5.5716852923007805,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023827639272368923,
      "loss": 4.7834,
      "mean_token_accuracy": 0.2304174691438675,
      "num_tokens": 110747558.0,
      "step": 49610
    },
    {
      "entropy": 5.318275213241577,
      "epoch": 5.572246869208739,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023823678296963253,
      "loss": 4.6949,
      "mean_token_accuracy": 0.2500362530350685,
      "num_tokens": 110757529.0,
      "step": 49615
    },
    {
      "entropy": 5.292996025085449,
      "epoch": 5.572808446116696,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023819717438613097,
      "loss": 4.926,
      "mean_token_accuracy": 0.23145831823349,
      "num_tokens": 110770180.0,
      "step": 49620
    },
    {
      "entropy": 5.3808411121368405,
      "epoch": 5.573370023024653,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023815756697444568,
      "loss": 4.9052,
      "mean_token_accuracy": 0.2330790191888809,
      "num_tokens": 110781845.0,
      "step": 49625
    },
    {
      "entropy": 5.35908522605896,
      "epoch": 5.573931599932611,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023811796073583774,
      "loss": 4.7825,
      "mean_token_accuracy": 0.2351295009255409,
      "num_tokens": 110792862.0,
      "step": 49630
    },
    {
      "entropy": 5.299463701248169,
      "epoch": 5.574493176840568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023807835567156815,
      "loss": 4.7666,
      "mean_token_accuracy": 0.24024985283613204,
      "num_tokens": 110803423.0,
      "step": 49635
    },
    {
      "entropy": 5.328565692901611,
      "epoch": 5.5750547537485255,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023803875178289813,
      "loss": 4.8092,
      "mean_token_accuracy": 0.2388436421751976,
      "num_tokens": 110814785.0,
      "step": 49640
    },
    {
      "entropy": 5.2144591331481935,
      "epoch": 5.575616330656484,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023799914907108856,
      "loss": 4.6967,
      "mean_token_accuracy": 0.24436908960342407,
      "num_tokens": 110827384.0,
      "step": 49645
    },
    {
      "entropy": 5.288430023193359,
      "epoch": 5.576177907564441,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023795954753740034,
      "loss": 4.7523,
      "mean_token_accuracy": 0.23878295123577117,
      "num_tokens": 110837819.0,
      "step": 49650
    },
    {
      "entropy": 5.2407067775726315,
      "epoch": 5.576739484472398,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023791994718309455,
      "loss": 4.7346,
      "mean_token_accuracy": 0.24089981317520143,
      "num_tokens": 110849302.0,
      "step": 49655
    },
    {
      "entropy": 5.265918159484864,
      "epoch": 5.577301061380356,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023788034800943203,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24372402131557463,
      "num_tokens": 110860009.0,
      "step": 49660
    },
    {
      "entropy": 5.255563497543335,
      "epoch": 5.577862638288313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023784075001767369,
      "loss": 4.7296,
      "mean_token_accuracy": 0.24097576290369033,
      "num_tokens": 110870232.0,
      "step": 49665
    },
    {
      "entropy": 5.332714653015136,
      "epoch": 5.5784242151962715,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00023780115320908024,
      "loss": 4.825,
      "mean_token_accuracy": 0.23312304466962813,
      "num_tokens": 110882339.0,
      "step": 49670
    },
    {
      "entropy": 5.206681442260742,
      "epoch": 5.578985792104229,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023776155758491253,
      "loss": 4.6094,
      "mean_token_accuracy": 0.26004977375268934,
      "num_tokens": 110893190.0,
      "step": 49675
    },
    {
      "entropy": 5.310211515426635,
      "epoch": 5.579547369012186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002377219631464312,
      "loss": 4.861,
      "mean_token_accuracy": 0.2340739443898201,
      "num_tokens": 110903353.0,
      "step": 49680
    },
    {
      "entropy": 5.318776607513428,
      "epoch": 5.580108945920144,
      "grad_norm": 1.078125,
      "learning_rate": 0.000237682369894897,
      "loss": 4.6762,
      "mean_token_accuracy": 0.2555518507957458,
      "num_tokens": 110913434.0,
      "step": 49685
    },
    {
      "entropy": 5.406667041778564,
      "epoch": 5.580670522828101,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023764277783157068,
      "loss": 4.862,
      "mean_token_accuracy": 0.23687906414270402,
      "num_tokens": 110924313.0,
      "step": 49690
    },
    {
      "entropy": 5.299753332138062,
      "epoch": 5.5812320997360585,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023760318695771282,
      "loss": 4.7646,
      "mean_token_accuracy": 0.23626088947057725,
      "num_tokens": 110935080.0,
      "step": 49695
    },
    {
      "entropy": 5.346971273422241,
      "epoch": 5.581793676644017,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023756359727458397,
      "loss": 4.8287,
      "mean_token_accuracy": 0.23872870355844497,
      "num_tokens": 110945759.0,
      "step": 49700
    },
    {
      "entropy": 5.359305000305175,
      "epoch": 5.582355253551974,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002375240087834447,
      "loss": 4.8975,
      "mean_token_accuracy": 0.2349878132343292,
      "num_tokens": 110957366.0,
      "step": 49705
    },
    {
      "entropy": 5.3496050357818605,
      "epoch": 5.582916830459931,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023748442148555554,
      "loss": 4.7661,
      "mean_token_accuracy": 0.24256251603364945,
      "num_tokens": 110968802.0,
      "step": 49710
    },
    {
      "entropy": 5.3203315258026125,
      "epoch": 5.583478407367889,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00023744483538217694,
      "loss": 4.7713,
      "mean_token_accuracy": 0.24997451901435852,
      "num_tokens": 110979149.0,
      "step": 49715
    },
    {
      "entropy": 5.279441499710083,
      "epoch": 5.584039984275846,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023740525047456936,
      "loss": 4.8108,
      "mean_token_accuracy": 0.23515270352363588,
      "num_tokens": 110989763.0,
      "step": 49720
    },
    {
      "entropy": 5.269330215454102,
      "epoch": 5.5846015611838045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00023736566676399307,
      "loss": 4.7508,
      "mean_token_accuracy": 0.24597535878419877,
      "num_tokens": 111002725.0,
      "step": 49725
    },
    {
      "entropy": 5.372385931015015,
      "epoch": 5.585163138091762,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002373260842517086,
      "loss": 4.8744,
      "mean_token_accuracy": 0.23326727598905564,
      "num_tokens": 111012594.0,
      "step": 49730
    },
    {
      "entropy": 5.329184198379517,
      "epoch": 5.585724714999719,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002372865029389763,
      "loss": 4.7381,
      "mean_token_accuracy": 0.23379893004894256,
      "num_tokens": 111023715.0,
      "step": 49735
    },
    {
      "entropy": 5.3332297801971436,
      "epoch": 5.586286291907677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023724692282705635,
      "loss": 4.8656,
      "mean_token_accuracy": 0.23298469483852385,
      "num_tokens": 111033483.0,
      "step": 49740
    },
    {
      "entropy": 5.289475727081299,
      "epoch": 5.586847868815634,
      "grad_norm": 1.140625,
      "learning_rate": 0.000237207343917209,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2330700382590294,
      "num_tokens": 111043947.0,
      "step": 49745
    },
    {
      "entropy": 5.348280191421509,
      "epoch": 5.587409445723592,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023716776621069446,
      "loss": 4.7921,
      "mean_token_accuracy": 0.24430675506591798,
      "num_tokens": 111054846.0,
      "step": 49750
    },
    {
      "entropy": 5.351887893676758,
      "epoch": 5.5879710226315495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002371281897087729,
      "loss": 4.7697,
      "mean_token_accuracy": 0.24037100970745087,
      "num_tokens": 111065629.0,
      "step": 49755
    },
    {
      "entropy": 5.3933525562286375,
      "epoch": 5.588532599539507,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023708861441270452,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2298197105526924,
      "num_tokens": 111075948.0,
      "step": 49760
    },
    {
      "entropy": 5.3321852684021,
      "epoch": 5.589094176447465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023704904032374936,
      "loss": 4.7974,
      "mean_token_accuracy": 0.23858806639909744,
      "num_tokens": 111087202.0,
      "step": 49765
    },
    {
      "entropy": 5.3726685523986815,
      "epoch": 5.589655753355422,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023700946744316747,
      "loss": 4.7974,
      "mean_token_accuracy": 0.23890706151723862,
      "num_tokens": 111098362.0,
      "step": 49770
    },
    {
      "entropy": 5.29086651802063,
      "epoch": 5.590217330263379,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002369698957722189,
      "loss": 4.75,
      "mean_token_accuracy": 0.24195538461208344,
      "num_tokens": 111110285.0,
      "step": 49775
    },
    {
      "entropy": 5.353623580932617,
      "epoch": 5.590778907171337,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023693032531216352,
      "loss": 4.8571,
      "mean_token_accuracy": 0.2439577117562294,
      "num_tokens": 111120778.0,
      "step": 49780
    },
    {
      "entropy": 5.283048677444458,
      "epoch": 5.591340484079295,
      "grad_norm": 1.0,
      "learning_rate": 0.00023689075606426143,
      "loss": 4.7259,
      "mean_token_accuracy": 0.24740778058767318,
      "num_tokens": 111133448.0,
      "step": 49785
    },
    {
      "entropy": 5.41764087677002,
      "epoch": 5.591902060987252,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002368511880297724,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2281196415424347,
      "num_tokens": 111144707.0,
      "step": 49790
    },
    {
      "entropy": 5.267609071731568,
      "epoch": 5.59246363789521,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023681162120995624,
      "loss": 4.7772,
      "mean_token_accuracy": 0.23908395916223527,
      "num_tokens": 111155314.0,
      "step": 49795
    },
    {
      "entropy": 5.264981842041015,
      "epoch": 5.593025214803167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023677205560607294,
      "loss": 4.8087,
      "mean_token_accuracy": 0.23505190312862395,
      "num_tokens": 111166642.0,
      "step": 49800
    },
    {
      "entropy": 5.286869096755981,
      "epoch": 5.593586791711125,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023673249121938228,
      "loss": 4.7395,
      "mean_token_accuracy": 0.23502422273159027,
      "num_tokens": 111176970.0,
      "step": 49805
    },
    {
      "entropy": 5.279078102111816,
      "epoch": 5.5941483686190825,
      "grad_norm": 0.96875,
      "learning_rate": 0.00023669292805114395,
      "loss": 4.76,
      "mean_token_accuracy": 0.24023348242044448,
      "num_tokens": 111189126.0,
      "step": 49810
    },
    {
      "entropy": 5.221080636978149,
      "epoch": 5.59470994552704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023665336610261762,
      "loss": 4.6639,
      "mean_token_accuracy": 0.24819029718637467,
      "num_tokens": 111199767.0,
      "step": 49815
    },
    {
      "entropy": 5.35191068649292,
      "epoch": 5.595271522434998,
      "grad_norm": 1.046875,
      "learning_rate": 0.000236613805375063,
      "loss": 4.9072,
      "mean_token_accuracy": 0.23030041754245759,
      "num_tokens": 111212117.0,
      "step": 49820
    },
    {
      "entropy": 5.299631404876709,
      "epoch": 5.595833099342955,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002365742458697397,
      "loss": 4.7538,
      "mean_token_accuracy": 0.23554390370845796,
      "num_tokens": 111223035.0,
      "step": 49825
    },
    {
      "entropy": 5.290033054351807,
      "epoch": 5.596394676250912,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023653468758790726,
      "loss": 4.7483,
      "mean_token_accuracy": 0.24961043149232864,
      "num_tokens": 111234858.0,
      "step": 49830
    },
    {
      "entropy": 5.261278820037842,
      "epoch": 5.59695625315887,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023649513053082528,
      "loss": 4.7414,
      "mean_token_accuracy": 0.2457188531756401,
      "num_tokens": 111246429.0,
      "step": 49835
    },
    {
      "entropy": 5.363491201400757,
      "epoch": 5.597517830066828,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002364555746997533,
      "loss": 4.846,
      "mean_token_accuracy": 0.2318970575928688,
      "num_tokens": 111257575.0,
      "step": 49840
    },
    {
      "entropy": 5.366553401947021,
      "epoch": 5.598079406974785,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023641602009595087,
      "loss": 4.8198,
      "mean_token_accuracy": 0.23420167863368987,
      "num_tokens": 111267542.0,
      "step": 49845
    },
    {
      "entropy": 5.2887451171875,
      "epoch": 5.598640983882743,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002363764667206772,
      "loss": 4.7621,
      "mean_token_accuracy": 0.24240659326314926,
      "num_tokens": 111277791.0,
      "step": 49850
    },
    {
      "entropy": 5.342336416244507,
      "epoch": 5.5992025607907,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023633691457519191,
      "loss": 4.8489,
      "mean_token_accuracy": 0.24252943992614745,
      "num_tokens": 111288997.0,
      "step": 49855
    },
    {
      "entropy": 5.384472608566284,
      "epoch": 5.599764137698658,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023629736366075416,
      "loss": 4.8497,
      "mean_token_accuracy": 0.23190672993659972,
      "num_tokens": 111299350.0,
      "step": 49860
    },
    {
      "entropy": 5.316676950454712,
      "epoch": 5.600325714606615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023625781397862333,
      "loss": 4.7756,
      "mean_token_accuracy": 0.2417615070939064,
      "num_tokens": 111310355.0,
      "step": 49865
    },
    {
      "entropy": 5.345535469055176,
      "epoch": 5.600887291514573,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023621826553005883,
      "loss": 4.7853,
      "mean_token_accuracy": 0.2321890264749527,
      "num_tokens": 111321822.0,
      "step": 49870
    },
    {
      "entropy": 5.209042978286743,
      "epoch": 5.601448868422531,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023617871831631976,
      "loss": 4.6765,
      "mean_token_accuracy": 0.25015122443437576,
      "num_tokens": 111333753.0,
      "step": 49875
    },
    {
      "entropy": 5.262098598480224,
      "epoch": 5.602010445330488,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023613917233866538,
      "loss": 4.8069,
      "mean_token_accuracy": 0.23463741689920425,
      "num_tokens": 111343933.0,
      "step": 49880
    },
    {
      "entropy": 5.243875789642334,
      "epoch": 5.602572022238445,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002360996275983548,
      "loss": 4.7002,
      "mean_token_accuracy": 0.24533117711544036,
      "num_tokens": 111353829.0,
      "step": 49885
    },
    {
      "entropy": 5.332877588272095,
      "epoch": 5.603133599146403,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002360600840966472,
      "loss": 4.881,
      "mean_token_accuracy": 0.2287111446261406,
      "num_tokens": 111365973.0,
      "step": 49890
    },
    {
      "entropy": 5.333840417861938,
      "epoch": 5.6036951760543605,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002360205418348016,
      "loss": 4.8127,
      "mean_token_accuracy": 0.24085693061351776,
      "num_tokens": 111377753.0,
      "step": 49895
    },
    {
      "entropy": 5.383044958114624,
      "epoch": 5.604256752962318,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023598100081407704,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2252856269478798,
      "num_tokens": 111389156.0,
      "step": 49900
    },
    {
      "entropy": 5.32713475227356,
      "epoch": 5.604818329870276,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023594146103573255,
      "loss": 4.7892,
      "mean_token_accuracy": 0.23620528131723403,
      "num_tokens": 111400387.0,
      "step": 49905
    },
    {
      "entropy": 5.241427898406982,
      "epoch": 5.605379906778233,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002359019225010271,
      "loss": 4.7162,
      "mean_token_accuracy": 0.24929002225399016,
      "num_tokens": 111410879.0,
      "step": 49910
    },
    {
      "entropy": 5.328026485443115,
      "epoch": 5.605941483686191,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002358623852112196,
      "loss": 4.8232,
      "mean_token_accuracy": 0.24401530176401137,
      "num_tokens": 111421463.0,
      "step": 49915
    },
    {
      "entropy": 5.314103507995606,
      "epoch": 5.606503060594148,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023582284916756898,
      "loss": 4.7758,
      "mean_token_accuracy": 0.24582781940698623,
      "num_tokens": 111432434.0,
      "step": 49920
    },
    {
      "entropy": 5.300924253463745,
      "epoch": 5.607064637502106,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000235783314371334,
      "loss": 4.823,
      "mean_token_accuracy": 0.23440522700548172,
      "num_tokens": 111443957.0,
      "step": 49925
    },
    {
      "entropy": 5.250681591033936,
      "epoch": 5.607626214410064,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00023574378082377351,
      "loss": 4.7167,
      "mean_token_accuracy": 0.24513275921344757,
      "num_tokens": 111456359.0,
      "step": 49930
    },
    {
      "entropy": 5.331561470031739,
      "epoch": 5.608187791318021,
      "grad_norm": 1.0,
      "learning_rate": 0.0002357042485261463,
      "loss": 4.807,
      "mean_token_accuracy": 0.24165308624505996,
      "num_tokens": 111467546.0,
      "step": 49935
    },
    {
      "entropy": 5.306063461303711,
      "epoch": 5.608749368225979,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002356647174797109,
      "loss": 4.8456,
      "mean_token_accuracy": 0.2352576047182083,
      "num_tokens": 111478151.0,
      "step": 49940
    },
    {
      "entropy": 5.302238845825196,
      "epoch": 5.609310945133936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023562518768572628,
      "loss": 4.8086,
      "mean_token_accuracy": 0.2396385118365288,
      "num_tokens": 111488553.0,
      "step": 49945
    },
    {
      "entropy": 5.334729242324829,
      "epoch": 5.6098725220418935,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023558565914545094,
      "loss": 4.8297,
      "mean_token_accuracy": 0.23153970241546631,
      "num_tokens": 111500406.0,
      "step": 49950
    },
    {
      "entropy": 5.305220651626587,
      "epoch": 5.610434098949852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023554613186014352,
      "loss": 4.7773,
      "mean_token_accuracy": 0.241000697016716,
      "num_tokens": 111512512.0,
      "step": 49955
    },
    {
      "entropy": 5.351428747177124,
      "epoch": 5.610995675857809,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023550660583106253,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23227994441986083,
      "num_tokens": 111523732.0,
      "step": 49960
    },
    {
      "entropy": 5.228313302993774,
      "epoch": 5.611557252765766,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023546708105946657,
      "loss": 4.7466,
      "mean_token_accuracy": 0.24349550753831864,
      "num_tokens": 111534531.0,
      "step": 49965
    },
    {
      "entropy": 5.369480228424072,
      "epoch": 5.612118829673724,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023542755754661404,
      "loss": 4.8637,
      "mean_token_accuracy": 0.22384159564971923,
      "num_tokens": 111545912.0,
      "step": 49970
    },
    {
      "entropy": 5.304588508605957,
      "epoch": 5.612680406581681,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002353880352937634,
      "loss": 4.7526,
      "mean_token_accuracy": 0.24084495455026628,
      "num_tokens": 111557720.0,
      "step": 49975
    },
    {
      "entropy": 5.258032751083374,
      "epoch": 5.6132419834896385,
      "grad_norm": 1.171875,
      "learning_rate": 0.00023534851430217314,
      "loss": 4.7152,
      "mean_token_accuracy": 0.24197059869766235,
      "num_tokens": 111568495.0,
      "step": 49980
    },
    {
      "entropy": 5.23774733543396,
      "epoch": 5.613803560397597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023530899457310157,
      "loss": 4.6964,
      "mean_token_accuracy": 0.24533956497907639,
      "num_tokens": 111578681.0,
      "step": 49985
    },
    {
      "entropy": 5.272009181976318,
      "epoch": 5.614365137305554,
      "grad_norm": 1.0,
      "learning_rate": 0.00023526947610780696,
      "loss": 4.8061,
      "mean_token_accuracy": 0.24793111830949782,
      "num_tokens": 111591130.0,
      "step": 49990
    },
    {
      "entropy": 5.218097591400147,
      "epoch": 5.614926714213512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023522995890754768,
      "loss": 4.6691,
      "mean_token_accuracy": 0.2594942539930344,
      "num_tokens": 111602233.0,
      "step": 49995
    },
    {
      "entropy": 5.264907693862915,
      "epoch": 5.615488291121469,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002351904429735819,
      "loss": 4.6755,
      "mean_token_accuracy": 0.2515743374824524,
      "num_tokens": 111613849.0,
      "step": 50000
    },
    {
      "entropy": 5.4169165134429935,
      "epoch": 5.616049868029426,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002351509283071679,
      "loss": 4.926,
      "mean_token_accuracy": 0.23416833579540253,
      "num_tokens": 111625070.0,
      "step": 50005
    },
    {
      "entropy": 5.3235053539276125,
      "epoch": 5.6166114449373845,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002351114149095636,
      "loss": 4.8233,
      "mean_token_accuracy": 0.2411249175667763,
      "num_tokens": 111635490.0,
      "step": 50010
    },
    {
      "entropy": 5.387363481521606,
      "epoch": 5.617173021845342,
      "grad_norm": 0.953125,
      "learning_rate": 0.00023507190278202755,
      "loss": 4.9299,
      "mean_token_accuracy": 0.234357587993145,
      "num_tokens": 111646985.0,
      "step": 50015
    },
    {
      "entropy": 5.332396793365478,
      "epoch": 5.617734598753299,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023503239192581755,
      "loss": 4.7885,
      "mean_token_accuracy": 0.23604532927274705,
      "num_tokens": 111658013.0,
      "step": 50020
    },
    {
      "entropy": 5.394826650619507,
      "epoch": 5.618296175661257,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002349928823421917,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22885941714048386,
      "num_tokens": 111668774.0,
      "step": 50025
    },
    {
      "entropy": 5.2444798946380615,
      "epoch": 5.618857752569214,
      "grad_norm": 1.0,
      "learning_rate": 0.00023495337403240795,
      "loss": 4.6475,
      "mean_token_accuracy": 0.2504767805337906,
      "num_tokens": 111679529.0,
      "step": 50030
    },
    {
      "entropy": 5.3151788234710695,
      "epoch": 5.6194193294771715,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023491386699772432,
      "loss": 4.8302,
      "mean_token_accuracy": 0.23455745726823807,
      "num_tokens": 111689528.0,
      "step": 50035
    },
    {
      "entropy": 5.258044099807739,
      "epoch": 5.61998090638513,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002348743612393987,
      "loss": 4.7468,
      "mean_token_accuracy": 0.23780842870473862,
      "num_tokens": 111700127.0,
      "step": 50040
    },
    {
      "entropy": 5.2800469398498535,
      "epoch": 5.620542483293087,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023483485675868899,
      "loss": 4.8104,
      "mean_token_accuracy": 0.23599256128072738,
      "num_tokens": 111711235.0,
      "step": 50045
    },
    {
      "entropy": 5.420039939880371,
      "epoch": 5.621104060201045,
      "grad_norm": 1.0,
      "learning_rate": 0.00023479535355685306,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23612986803054808,
      "num_tokens": 111722815.0,
      "step": 50050
    },
    {
      "entropy": 5.369415426254273,
      "epoch": 5.621665637109002,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023475585163514862,
      "loss": 4.8068,
      "mean_token_accuracy": 0.24198617190122604,
      "num_tokens": 111733690.0,
      "step": 50055
    },
    {
      "entropy": 5.258489894866943,
      "epoch": 5.622227214016959,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002347163509948335,
      "loss": 4.7559,
      "mean_token_accuracy": 0.24393816590309142,
      "num_tokens": 111744300.0,
      "step": 50060
    },
    {
      "entropy": 5.3482362747192385,
      "epoch": 5.6227887909249175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023467685163716534,
      "loss": 4.8872,
      "mean_token_accuracy": 0.2332092985510826,
      "num_tokens": 111757526.0,
      "step": 50065
    },
    {
      "entropy": 5.428538131713867,
      "epoch": 5.623350367832875,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023463735356340194,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22608986347913743,
      "num_tokens": 111769211.0,
      "step": 50070
    },
    {
      "entropy": 5.336868143081665,
      "epoch": 5.623911944740832,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023459785677480078,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23970758467912673,
      "num_tokens": 111780652.0,
      "step": 50075
    },
    {
      "entropy": 5.376889038085937,
      "epoch": 5.62447352164879,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002345583612726196,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23729144185781478,
      "num_tokens": 111791749.0,
      "step": 50080
    },
    {
      "entropy": 5.305047035217285,
      "epoch": 5.625035098556747,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023451886705811576,
      "loss": 4.8125,
      "mean_token_accuracy": 0.233553746342659,
      "num_tokens": 111802879.0,
      "step": 50085
    },
    {
      "entropy": 5.21832308769226,
      "epoch": 5.625596675464704,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000234479374132547,
      "loss": 4.7827,
      "mean_token_accuracy": 0.23702092319726945,
      "num_tokens": 111813768.0,
      "step": 50090
    },
    {
      "entropy": 5.2972578525543215,
      "epoch": 5.6261582523726625,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023443988249717063,
      "loss": 4.7897,
      "mean_token_accuracy": 0.23812369108200074,
      "num_tokens": 111824667.0,
      "step": 50095
    },
    {
      "entropy": 5.296841621398926,
      "epoch": 5.62671982928062,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00023440039215324415,
      "loss": 4.7448,
      "mean_token_accuracy": 0.246409872174263,
      "num_tokens": 111835969.0,
      "step": 50100
    },
    {
      "entropy": 5.296240425109863,
      "epoch": 5.627281406188578,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023436090310202497,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22913775593042374,
      "num_tokens": 111846991.0,
      "step": 50105
    },
    {
      "entropy": 5.321786165237427,
      "epoch": 5.627842983096535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023432141534477036,
      "loss": 4.8244,
      "mean_token_accuracy": 0.2395289123058319,
      "num_tokens": 111858160.0,
      "step": 50110
    },
    {
      "entropy": 5.289545297622681,
      "epoch": 5.628404560004492,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00023428192888273768,
      "loss": 4.7378,
      "mean_token_accuracy": 0.2444293588399887,
      "num_tokens": 111869382.0,
      "step": 50115
    },
    {
      "entropy": 5.3944765567779545,
      "epoch": 5.62896613691245,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023424244371718418,
      "loss": 4.8416,
      "mean_token_accuracy": 0.2415391817688942,
      "num_tokens": 111879538.0,
      "step": 50120
    },
    {
      "entropy": 5.300840568542481,
      "epoch": 5.629527713820408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023420295984936707,
      "loss": 4.825,
      "mean_token_accuracy": 0.23382407873868943,
      "num_tokens": 111890342.0,
      "step": 50125
    },
    {
      "entropy": 5.354081296920777,
      "epoch": 5.630089290728366,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023416347728054355,
      "loss": 4.8239,
      "mean_token_accuracy": 0.23479485362768174,
      "num_tokens": 111902009.0,
      "step": 50130
    },
    {
      "entropy": 5.472142887115479,
      "epoch": 5.630650867636323,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023412399601197072,
      "loss": 4.9691,
      "mean_token_accuracy": 0.23491252064704896,
      "num_tokens": 111915409.0,
      "step": 50135
    },
    {
      "entropy": 5.367962121963501,
      "epoch": 5.63121244454428,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023408451604490577,
      "loss": 4.7931,
      "mean_token_accuracy": 0.23825278580188752,
      "num_tokens": 111926355.0,
      "step": 50140
    },
    {
      "entropy": 5.306478357315063,
      "epoch": 5.631774021452238,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023404503738060566,
      "loss": 4.7732,
      "mean_token_accuracy": 0.24004885107278823,
      "num_tokens": 111936733.0,
      "step": 50145
    },
    {
      "entropy": 5.323543834686279,
      "epoch": 5.6323355983601955,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023400556002032747,
      "loss": 4.8526,
      "mean_token_accuracy": 0.2386070594191551,
      "num_tokens": 111947707.0,
      "step": 50150
    },
    {
      "entropy": 5.288376665115356,
      "epoch": 5.632897175268153,
      "grad_norm": 1.015625,
      "learning_rate": 0.000233966083965328,
      "loss": 4.7954,
      "mean_token_accuracy": 0.24198931753635405,
      "num_tokens": 111958666.0,
      "step": 50155
    },
    {
      "entropy": 5.3213966369628904,
      "epoch": 5.633458752176111,
      "grad_norm": 1.0,
      "learning_rate": 0.00023392660921686448,
      "loss": 4.8058,
      "mean_token_accuracy": 0.23493824601173402,
      "num_tokens": 111970476.0,
      "step": 50160
    },
    {
      "entropy": 5.394649076461792,
      "epoch": 5.634020329084068,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023388713577619357,
      "loss": 4.9138,
      "mean_token_accuracy": 0.23239467442035674,
      "num_tokens": 111981617.0,
      "step": 50165
    },
    {
      "entropy": 5.272856903076172,
      "epoch": 5.634581905992025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023384766364457227,
      "loss": 4.7438,
      "mean_token_accuracy": 0.24254267960786818,
      "num_tokens": 111993370.0,
      "step": 50170
    },
    {
      "entropy": 5.315666961669922,
      "epoch": 5.635143482899983,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023380819282325732,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23991773426532745,
      "num_tokens": 112004832.0,
      "step": 50175
    },
    {
      "entropy": 5.341872692108154,
      "epoch": 5.635705059807941,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023376872331350553,
      "loss": 4.8057,
      "mean_token_accuracy": 0.24645335078239441,
      "num_tokens": 112017388.0,
      "step": 50180
    },
    {
      "entropy": 5.2631066799163815,
      "epoch": 5.636266636715899,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002337292551165735,
      "loss": 4.7233,
      "mean_token_accuracy": 0.2391225203871727,
      "num_tokens": 112027960.0,
      "step": 50185
    },
    {
      "entropy": 5.2514289855957035,
      "epoch": 5.636828213623856,
      "grad_norm": 1.046875,
      "learning_rate": 0.000233689788233718,
      "loss": 4.8044,
      "mean_token_accuracy": 0.23843126893043518,
      "num_tokens": 112039209.0,
      "step": 50190
    },
    {
      "entropy": 5.288823413848877,
      "epoch": 5.637389790531813,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023365032266619567,
      "loss": 4.8243,
      "mean_token_accuracy": 0.2317632958292961,
      "num_tokens": 112050857.0,
      "step": 50195
    },
    {
      "entropy": 5.329683542251587,
      "epoch": 5.637951367439771,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023361085841526314,
      "loss": 4.7527,
      "mean_token_accuracy": 0.249263134598732,
      "num_tokens": 112062071.0,
      "step": 50200
    },
    {
      "entropy": 5.427093505859375,
      "epoch": 5.638512944347728,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023357139548217692,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2312479868531227,
      "num_tokens": 112074548.0,
      "step": 50205
    },
    {
      "entropy": 5.383727359771728,
      "epoch": 5.639074521255686,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023353193386819353,
      "loss": 4.8496,
      "mean_token_accuracy": 0.2308310478925705,
      "num_tokens": 112084494.0,
      "step": 50210
    },
    {
      "entropy": 5.316034317016602,
      "epoch": 5.639636098163644,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023349247357456942,
      "loss": 4.8305,
      "mean_token_accuracy": 0.2345963254570961,
      "num_tokens": 112095551.0,
      "step": 50215
    },
    {
      "entropy": 5.412002897262573,
      "epoch": 5.640197675071601,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023345301460256113,
      "loss": 4.9083,
      "mean_token_accuracy": 0.23743947893381118,
      "num_tokens": 112107571.0,
      "step": 50220
    },
    {
      "entropy": 5.393449687957764,
      "epoch": 5.640759251979558,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002334135569534248,
      "loss": 4.8704,
      "mean_token_accuracy": 0.23376665264368057,
      "num_tokens": 112118136.0,
      "step": 50225
    },
    {
      "entropy": 5.323233222961425,
      "epoch": 5.641320828887516,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023337410062841706,
      "loss": 4.788,
      "mean_token_accuracy": 0.23755892366170883,
      "num_tokens": 112129611.0,
      "step": 50230
    },
    {
      "entropy": 5.395784664154053,
      "epoch": 5.6418824057954735,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023333464562879408,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23660051375627517,
      "num_tokens": 112140097.0,
      "step": 50235
    },
    {
      "entropy": 5.405519342422485,
      "epoch": 5.642443982703432,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00023329519195581218,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2319188177585602,
      "num_tokens": 112151117.0,
      "step": 50240
    },
    {
      "entropy": 5.310891246795654,
      "epoch": 5.643005559611389,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023325573961072754,
      "loss": 4.7807,
      "mean_token_accuracy": 0.24309643656015395,
      "num_tokens": 112162126.0,
      "step": 50245
    },
    {
      "entropy": 5.3603898048400875,
      "epoch": 5.643567136519346,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023321628859479632,
      "loss": 4.8563,
      "mean_token_accuracy": 0.23365887254476547,
      "num_tokens": 112172473.0,
      "step": 50250
    },
    {
      "entropy": 5.254723978042603,
      "epoch": 5.644128713427304,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002331768389092747,
      "loss": 4.6912,
      "mean_token_accuracy": 0.23849423080682755,
      "num_tokens": 112182718.0,
      "step": 50255
    },
    {
      "entropy": 5.284247732162475,
      "epoch": 5.644690290335261,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023313739055541878,
      "loss": 4.7218,
      "mean_token_accuracy": 0.24317964911460876,
      "num_tokens": 112193349.0,
      "step": 50260
    },
    {
      "entropy": 5.281680727005005,
      "epoch": 5.645251867243219,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002330979435344845,
      "loss": 4.8106,
      "mean_token_accuracy": 0.2382815510034561,
      "num_tokens": 112203674.0,
      "step": 50265
    },
    {
      "entropy": 5.34605975151062,
      "epoch": 5.645813444151177,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023305849784772799,
      "loss": 4.8266,
      "mean_token_accuracy": 0.23596908897161484,
      "num_tokens": 112216601.0,
      "step": 50270
    },
    {
      "entropy": 5.34045000076294,
      "epoch": 5.646375021059134,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002330190534964051,
      "loss": 4.7471,
      "mean_token_accuracy": 0.2421631395816803,
      "num_tokens": 112227594.0,
      "step": 50275
    },
    {
      "entropy": 5.408769035339356,
      "epoch": 5.646936597967091,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00023297961048177195,
      "loss": 4.8304,
      "mean_token_accuracy": 0.2398897036910057,
      "num_tokens": 112238706.0,
      "step": 50280
    },
    {
      "entropy": 5.273558282852173,
      "epoch": 5.647498174875049,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023294016880508428,
      "loss": 4.7706,
      "mean_token_accuracy": 0.242348313331604,
      "num_tokens": 112249248.0,
      "step": 50285
    },
    {
      "entropy": 5.304581642150879,
      "epoch": 5.6480597517830065,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023290072846759784,
      "loss": 4.8352,
      "mean_token_accuracy": 0.2401745155453682,
      "num_tokens": 112259188.0,
      "step": 50290
    },
    {
      "entropy": 5.277836179733276,
      "epoch": 5.648621328690965,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023286128947056867,
      "loss": 4.7349,
      "mean_token_accuracy": 0.2413178116083145,
      "num_tokens": 112270275.0,
      "step": 50295
    },
    {
      "entropy": 5.3264806270599365,
      "epoch": 5.649182905598922,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023282185181525223,
      "loss": 4.8209,
      "mean_token_accuracy": 0.2327048659324646,
      "num_tokens": 112282133.0,
      "step": 50300
    },
    {
      "entropy": 5.293250703811646,
      "epoch": 5.649744482506879,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00023278241550290453,
      "loss": 4.7406,
      "mean_token_accuracy": 0.2394850581884384,
      "num_tokens": 112293448.0,
      "step": 50305
    },
    {
      "entropy": 5.313910055160522,
      "epoch": 5.650306059414837,
      "grad_norm": 1.125,
      "learning_rate": 0.00023274298053478106,
      "loss": 4.7743,
      "mean_token_accuracy": 0.24056325554847718,
      "num_tokens": 112304979.0,
      "step": 50310
    },
    {
      "entropy": 5.341831016540527,
      "epoch": 5.650867636322794,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00023270354691213742,
      "loss": 4.7513,
      "mean_token_accuracy": 0.238727305829525,
      "num_tokens": 112314690.0,
      "step": 50315
    },
    {
      "entropy": 5.23231315612793,
      "epoch": 5.651429213230752,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023266411463622927,
      "loss": 4.6899,
      "mean_token_accuracy": 0.24260217249393462,
      "num_tokens": 112325153.0,
      "step": 50320
    },
    {
      "entropy": 5.236805200576782,
      "epoch": 5.65199079013871,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002326246837083122,
      "loss": 4.7468,
      "mean_token_accuracy": 0.23671069741249084,
      "num_tokens": 112336848.0,
      "step": 50325
    },
    {
      "entropy": 5.338944005966186,
      "epoch": 5.652552367046667,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002325852541296416,
      "loss": 4.8991,
      "mean_token_accuracy": 0.23050131648778915,
      "num_tokens": 112348849.0,
      "step": 50330
    },
    {
      "entropy": 5.303166103363037,
      "epoch": 5.653113943954625,
      "grad_norm": 0.96875,
      "learning_rate": 0.00023254582590147288,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23731044232845305,
      "num_tokens": 112360408.0,
      "step": 50335
    },
    {
      "entropy": 5.329484462738037,
      "epoch": 5.653675520862582,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002325063990250616,
      "loss": 4.8354,
      "mean_token_accuracy": 0.23300303518772125,
      "num_tokens": 112372633.0,
      "step": 50340
    },
    {
      "entropy": 5.412740325927734,
      "epoch": 5.654237097770539,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023246697350166306,
      "loss": 4.8402,
      "mean_token_accuracy": 0.23597698360681535,
      "num_tokens": 112383269.0,
      "step": 50345
    },
    {
      "entropy": 5.350608873367309,
      "epoch": 5.6547986746784975,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023242754933253256,
      "loss": 4.856,
      "mean_token_accuracy": 0.2412739172577858,
      "num_tokens": 112395079.0,
      "step": 50350
    },
    {
      "entropy": 5.295021486282349,
      "epoch": 5.655360251586455,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002323881265189255,
      "loss": 4.7286,
      "mean_token_accuracy": 0.24483977854251862,
      "num_tokens": 112406674.0,
      "step": 50355
    },
    {
      "entropy": 5.277003335952759,
      "epoch": 5.655921828494412,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023234870506209693,
      "loss": 4.7111,
      "mean_token_accuracy": 0.24847178012132645,
      "num_tokens": 112417684.0,
      "step": 50360
    },
    {
      "entropy": 5.278333854675293,
      "epoch": 5.65648340540237,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002323092849633021,
      "loss": 4.751,
      "mean_token_accuracy": 0.2426846817135811,
      "num_tokens": 112427986.0,
      "step": 50365
    },
    {
      "entropy": 5.262766313552857,
      "epoch": 5.657044982310327,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023226986622379615,
      "loss": 4.7334,
      "mean_token_accuracy": 0.246213661134243,
      "num_tokens": 112438527.0,
      "step": 50370
    },
    {
      "entropy": 5.253688764572144,
      "epoch": 5.657606559218285,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023223044884483436,
      "loss": 4.7373,
      "mean_token_accuracy": 0.2424413874745369,
      "num_tokens": 112448432.0,
      "step": 50375
    },
    {
      "entropy": 5.301943826675415,
      "epoch": 5.658168136126243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023219103282767157,
      "loss": 4.7742,
      "mean_token_accuracy": 0.23988452404737473,
      "num_tokens": 112459793.0,
      "step": 50380
    },
    {
      "entropy": 5.271808385848999,
      "epoch": 5.6587297130342,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000232151618173563,
      "loss": 4.6948,
      "mean_token_accuracy": 0.25057790279388426,
      "num_tokens": 112470868.0,
      "step": 50385
    },
    {
      "entropy": 5.308812952041626,
      "epoch": 5.659291289942158,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002321122048837635,
      "loss": 4.8018,
      "mean_token_accuracy": 0.2371307447552681,
      "num_tokens": 112481362.0,
      "step": 50390
    },
    {
      "entropy": 5.438502168655395,
      "epoch": 5.659852866850115,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023207279295952798,
      "loss": 4.9371,
      "mean_token_accuracy": 0.22647445797920226,
      "num_tokens": 112492421.0,
      "step": 50395
    },
    {
      "entropy": 5.33788423538208,
      "epoch": 5.660414443758072,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023203338240211142,
      "loss": 4.8179,
      "mean_token_accuracy": 0.24389825910329818,
      "num_tokens": 112503327.0,
      "step": 50400
    },
    {
      "entropy": 5.329284143447876,
      "epoch": 5.6609760206660305,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002319939732127686,
      "loss": 4.7823,
      "mean_token_accuracy": 0.24604029804468155,
      "num_tokens": 112514424.0,
      "step": 50405
    },
    {
      "entropy": 5.313203525543213,
      "epoch": 5.661537597573988,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00023195456539275433,
      "loss": 4.7819,
      "mean_token_accuracy": 0.24753837287425995,
      "num_tokens": 112526366.0,
      "step": 50410
    },
    {
      "entropy": 5.2862719058990475,
      "epoch": 5.662099174481945,
      "grad_norm": 0.953125,
      "learning_rate": 0.00023191515894332345,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23780821710824968,
      "num_tokens": 112538712.0,
      "step": 50415
    },
    {
      "entropy": 5.3067668914794925,
      "epoch": 5.662660751389903,
      "grad_norm": 0.96875,
      "learning_rate": 0.00023187575386573057,
      "loss": 4.824,
      "mean_token_accuracy": 0.23691004663705825,
      "num_tokens": 112549696.0,
      "step": 50420
    },
    {
      "entropy": 5.308401298522949,
      "epoch": 5.66322232829786,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023183635016123044,
      "loss": 4.7248,
      "mean_token_accuracy": 0.2467487037181854,
      "num_tokens": 112561056.0,
      "step": 50425
    },
    {
      "entropy": 5.302348470687866,
      "epoch": 5.663783905205818,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002317969478310777,
      "loss": 4.7988,
      "mean_token_accuracy": 0.2349467843770981,
      "num_tokens": 112572537.0,
      "step": 50430
    },
    {
      "entropy": 5.328875970840454,
      "epoch": 5.6643454821137755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023175754687652684,
      "loss": 4.7835,
      "mean_token_accuracy": 0.2431969165802002,
      "num_tokens": 112582795.0,
      "step": 50435
    },
    {
      "entropy": 5.349136924743652,
      "epoch": 5.664907059021733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002317181472988325,
      "loss": 4.8653,
      "mean_token_accuracy": 0.2317551776766777,
      "num_tokens": 112595187.0,
      "step": 50440
    },
    {
      "entropy": 5.3060108661651615,
      "epoch": 5.665468635929691,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023167874909924902,
      "loss": 4.7832,
      "mean_token_accuracy": 0.24428419321775435,
      "num_tokens": 112605739.0,
      "step": 50445
    },
    {
      "entropy": 5.344893884658814,
      "epoch": 5.666030212837648,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002316393522790311,
      "loss": 4.8415,
      "mean_token_accuracy": 0.23465729951858522,
      "num_tokens": 112616542.0,
      "step": 50450
    },
    {
      "entropy": 5.360758590698242,
      "epoch": 5.666591789745605,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023159995683943297,
      "loss": 4.8523,
      "mean_token_accuracy": 0.23193674236536027,
      "num_tokens": 112627438.0,
      "step": 50455
    },
    {
      "entropy": 5.324874496459961,
      "epoch": 5.667153366653563,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023156056278170905,
      "loss": 4.7934,
      "mean_token_accuracy": 0.2431514784693718,
      "num_tokens": 112639535.0,
      "step": 50460
    },
    {
      "entropy": 5.34074296951294,
      "epoch": 5.667714943561521,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023152117010711372,
      "loss": 4.8403,
      "mean_token_accuracy": 0.23114818781614305,
      "num_tokens": 112650084.0,
      "step": 50465
    },
    {
      "entropy": 5.3014488697052,
      "epoch": 5.668276520469478,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023148177881690112,
      "loss": 4.8416,
      "mean_token_accuracy": 0.2440393254160881,
      "num_tokens": 112661199.0,
      "step": 50470
    },
    {
      "entropy": 5.309929513931275,
      "epoch": 5.668838097377436,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002314423889123256,
      "loss": 4.7635,
      "mean_token_accuracy": 0.24268273115158082,
      "num_tokens": 112672494.0,
      "step": 50475
    },
    {
      "entropy": 5.316413307189942,
      "epoch": 5.669399674285393,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023140300039464125,
      "loss": 4.8269,
      "mean_token_accuracy": 0.23067785501480104,
      "num_tokens": 112683148.0,
      "step": 50480
    },
    {
      "entropy": 5.256842756271363,
      "epoch": 5.669961251193351,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023136361326510236,
      "loss": 4.777,
      "mean_token_accuracy": 0.23688191920518875,
      "num_tokens": 112692927.0,
      "step": 50485
    },
    {
      "entropy": 5.272323083877564,
      "epoch": 5.6705228281013085,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002313242275249629,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23903842270374298,
      "num_tokens": 112704481.0,
      "step": 50490
    },
    {
      "entropy": 5.318587112426758,
      "epoch": 5.671084405009266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023128484317547698,
      "loss": 4.823,
      "mean_token_accuracy": 0.23526267558336258,
      "num_tokens": 112716167.0,
      "step": 50495
    },
    {
      "entropy": 5.3239830493927,
      "epoch": 5.671645981917224,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023124546021789866,
      "loss": 4.7615,
      "mean_token_accuracy": 0.24323577284812928,
      "num_tokens": 112726639.0,
      "step": 50500
    },
    {
      "entropy": 5.321477746963501,
      "epoch": 5.672207558825181,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002312060786534818,
      "loss": 4.8332,
      "mean_token_accuracy": 0.23055649548768997,
      "num_tokens": 112737151.0,
      "step": 50505
    },
    {
      "entropy": 5.363887357711792,
      "epoch": 5.672769135733139,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023116669848348043,
      "loss": 4.8517,
      "mean_token_accuracy": 0.23334273397922517,
      "num_tokens": 112748507.0,
      "step": 50510
    },
    {
      "entropy": 5.367174482345581,
      "epoch": 5.673330712641096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023112731970914823,
      "loss": 4.8852,
      "mean_token_accuracy": 0.22842475175857543,
      "num_tokens": 112760042.0,
      "step": 50515
    },
    {
      "entropy": 5.429375886917114,
      "epoch": 5.673892289549054,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002310879423317393,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22588659673929215,
      "num_tokens": 112771734.0,
      "step": 50520
    },
    {
      "entropy": 5.352873849868774,
      "epoch": 5.674453866457012,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023104856635250732,
      "loss": 4.8078,
      "mean_token_accuracy": 0.2395622417330742,
      "num_tokens": 112784551.0,
      "step": 50525
    },
    {
      "entropy": 5.372788524627685,
      "epoch": 5.675015443364969,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023100919177270607,
      "loss": 4.8816,
      "mean_token_accuracy": 0.2405443698167801,
      "num_tokens": 112795622.0,
      "step": 50530
    },
    {
      "entropy": 5.335906171798706,
      "epoch": 5.675577020272926,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023096981859358918,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2366152063012123,
      "num_tokens": 112806765.0,
      "step": 50535
    },
    {
      "entropy": 5.312664127349853,
      "epoch": 5.676138597180884,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023093044681641035,
      "loss": 4.8049,
      "mean_token_accuracy": 0.24394844025373458,
      "num_tokens": 112816672.0,
      "step": 50540
    },
    {
      "entropy": 5.338701295852661,
      "epoch": 5.676700174088841,
      "grad_norm": 1.125,
      "learning_rate": 0.0002308910764424232,
      "loss": 4.8863,
      "mean_token_accuracy": 0.23739567548036575,
      "num_tokens": 112827384.0,
      "step": 50545
    },
    {
      "entropy": 5.356695699691772,
      "epoch": 5.677261750996799,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002308517074728812,
      "loss": 4.8337,
      "mean_token_accuracy": 0.23628831058740615,
      "num_tokens": 112839666.0,
      "step": 50550
    },
    {
      "entropy": 5.386043691635132,
      "epoch": 5.677823327904757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023081233990903798,
      "loss": 4.845,
      "mean_token_accuracy": 0.2386646255850792,
      "num_tokens": 112851644.0,
      "step": 50555
    },
    {
      "entropy": 5.351182270050049,
      "epoch": 5.678384904812714,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023077297375214708,
      "loss": 4.7942,
      "mean_token_accuracy": 0.2430309996008873,
      "num_tokens": 112862931.0,
      "step": 50560
    },
    {
      "entropy": 5.347780656814575,
      "epoch": 5.678946481720672,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002307336090034618,
      "loss": 4.8504,
      "mean_token_accuracy": 0.23886551409959794,
      "num_tokens": 112874363.0,
      "step": 50565
    },
    {
      "entropy": 5.406192779541016,
      "epoch": 5.679508058628629,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023069424566423557,
      "loss": 4.9072,
      "mean_token_accuracy": 0.23655912280082703,
      "num_tokens": 112885099.0,
      "step": 50570
    },
    {
      "entropy": 5.373861646652221,
      "epoch": 5.6800696355365865,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023065488373572173,
      "loss": 4.8775,
      "mean_token_accuracy": 0.23291868418455125,
      "num_tokens": 112898018.0,
      "step": 50575
    },
    {
      "entropy": 5.46699013710022,
      "epoch": 5.680631212444545,
      "grad_norm": 0.953125,
      "learning_rate": 0.00023061552321917368,
      "loss": 4.945,
      "mean_token_accuracy": 0.2327926516532898,
      "num_tokens": 112910235.0,
      "step": 50580
    },
    {
      "entropy": 5.33443055152893,
      "epoch": 5.681192789352502,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023057616411584453,
      "loss": 4.7732,
      "mean_token_accuracy": 0.24201924949884415,
      "num_tokens": 112921278.0,
      "step": 50585
    },
    {
      "entropy": 5.285068798065185,
      "epoch": 5.681754366260459,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00023053680642698743,
      "loss": 4.7511,
      "mean_token_accuracy": 0.24415156990289688,
      "num_tokens": 112931898.0,
      "step": 50590
    },
    {
      "entropy": 5.3394897937774655,
      "epoch": 5.682315943168417,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023049745015385576,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23727932423353196,
      "num_tokens": 112942916.0,
      "step": 50595
    },
    {
      "entropy": 5.345683526992798,
      "epoch": 5.682877520076374,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023045809529770252,
      "loss": 4.7718,
      "mean_token_accuracy": 0.23371665328741073,
      "num_tokens": 112953687.0,
      "step": 50600
    },
    {
      "entropy": 5.4508086204528805,
      "epoch": 5.683439096984332,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023041874185978086,
      "loss": 4.9516,
      "mean_token_accuracy": 0.22772231549024582,
      "num_tokens": 112965358.0,
      "step": 50605
    },
    {
      "entropy": 5.340061283111572,
      "epoch": 5.68400067389229,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002303793898413437,
      "loss": 4.7974,
      "mean_token_accuracy": 0.23974230736494065,
      "num_tokens": 112977503.0,
      "step": 50610
    },
    {
      "entropy": 5.381912660598755,
      "epoch": 5.684562250800247,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002303400392436441,
      "loss": 4.8648,
      "mean_token_accuracy": 0.23822314143180848,
      "num_tokens": 112988920.0,
      "step": 50615
    },
    {
      "entropy": 5.3739248752594,
      "epoch": 5.685123827708205,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002303006900679349,
      "loss": 4.841,
      "mean_token_accuracy": 0.23464004695415497,
      "num_tokens": 113000204.0,
      "step": 50620
    },
    {
      "entropy": 5.254441261291504,
      "epoch": 5.685685404616162,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00023026134231546907,
      "loss": 4.7469,
      "mean_token_accuracy": 0.24086756110191346,
      "num_tokens": 113013002.0,
      "step": 50625
    },
    {
      "entropy": 5.2904867172241214,
      "epoch": 5.6862469815241194,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023022199598749955,
      "loss": 4.735,
      "mean_token_accuracy": 0.24280466735363007,
      "num_tokens": 113025155.0,
      "step": 50630
    },
    {
      "entropy": 5.280267381668091,
      "epoch": 5.6868085584320776,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00023018265108527903,
      "loss": 4.7782,
      "mean_token_accuracy": 0.24775807112455367,
      "num_tokens": 113035512.0,
      "step": 50635
    },
    {
      "entropy": 5.378650236129761,
      "epoch": 5.687370135340035,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023014330761006018,
      "loss": 4.8514,
      "mean_token_accuracy": 0.2352456718683243,
      "num_tokens": 113046842.0,
      "step": 50640
    },
    {
      "entropy": 5.415841579437256,
      "epoch": 5.687931712247992,
      "grad_norm": 1.171875,
      "learning_rate": 0.00023010396556309592,
      "loss": 4.8159,
      "mean_token_accuracy": 0.24644220620393753,
      "num_tokens": 113057608.0,
      "step": 50645
    },
    {
      "entropy": 5.304312086105346,
      "epoch": 5.68849328915595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002300646249456388,
      "loss": 4.733,
      "mean_token_accuracy": 0.2428860992193222,
      "num_tokens": 113068540.0,
      "step": 50650
    },
    {
      "entropy": 5.333523082733154,
      "epoch": 5.689054866063907,
      "grad_norm": 1.0,
      "learning_rate": 0.00023002528575894138,
      "loss": 4.8274,
      "mean_token_accuracy": 0.23288071304559707,
      "num_tokens": 113080085.0,
      "step": 50655
    },
    {
      "entropy": 5.3003747940063475,
      "epoch": 5.6896164429718645,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022998594800425627,
      "loss": 4.7469,
      "mean_token_accuracy": 0.2462264746427536,
      "num_tokens": 113091539.0,
      "step": 50660
    },
    {
      "entropy": 5.346331071853638,
      "epoch": 5.690178019879823,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022994661168283608,
      "loss": 4.8101,
      "mean_token_accuracy": 0.23819625973701478,
      "num_tokens": 113101017.0,
      "step": 50665
    },
    {
      "entropy": 5.259552764892578,
      "epoch": 5.69073959678778,
      "grad_norm": 0.9375,
      "learning_rate": 0.00022990727679593332,
      "loss": 4.7769,
      "mean_token_accuracy": 0.23396315425634384,
      "num_tokens": 113112583.0,
      "step": 50670
    },
    {
      "entropy": 5.330195617675781,
      "epoch": 5.691301173695738,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022986794334480033,
      "loss": 4.8158,
      "mean_token_accuracy": 0.23189981281757355,
      "num_tokens": 113123689.0,
      "step": 50675
    },
    {
      "entropy": 5.41085467338562,
      "epoch": 5.691862750603695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002298286113306895,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22999204695224762,
      "num_tokens": 113135466.0,
      "step": 50680
    },
    {
      "entropy": 5.3285387516021725,
      "epoch": 5.692424327511652,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022978928075485322,
      "loss": 4.7576,
      "mean_token_accuracy": 0.2352118268609047,
      "num_tokens": 113146892.0,
      "step": 50685
    },
    {
      "entropy": 5.34357647895813,
      "epoch": 5.6929859044196105,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002297499516185438,
      "loss": 4.8055,
      "mean_token_accuracy": 0.23437814116477967,
      "num_tokens": 113157448.0,
      "step": 50690
    },
    {
      "entropy": 5.3635648727417,
      "epoch": 5.693547481327568,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00022971062392301335,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23511876463890075,
      "num_tokens": 113168455.0,
      "step": 50695
    },
    {
      "entropy": 5.377234172821045,
      "epoch": 5.694109058235526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022967129766951429,
      "loss": 4.8112,
      "mean_token_accuracy": 0.23832148611545562,
      "num_tokens": 113178612.0,
      "step": 50700
    },
    {
      "entropy": 5.327203941345215,
      "epoch": 5.694670635143483,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002296319728592987,
      "loss": 4.8554,
      "mean_token_accuracy": 0.23466818183660507,
      "num_tokens": 113190444.0,
      "step": 50705
    },
    {
      "entropy": 5.346721124649048,
      "epoch": 5.69523221205144,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002295926494936187,
      "loss": 4.8283,
      "mean_token_accuracy": 0.23772380650043487,
      "num_tokens": 113200994.0,
      "step": 50710
    },
    {
      "entropy": 5.40915961265564,
      "epoch": 5.695793788959398,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022955332757372633,
      "loss": 4.8704,
      "mean_token_accuracy": 0.23765342235565184,
      "num_tokens": 113212238.0,
      "step": 50715
    },
    {
      "entropy": 5.397620725631714,
      "epoch": 5.696355365867356,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022951400710087368,
      "loss": 4.8776,
      "mean_token_accuracy": 0.23589127361774445,
      "num_tokens": 113222693.0,
      "step": 50720
    },
    {
      "entropy": 5.340759038925171,
      "epoch": 5.696916942775313,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002294746880763126,
      "loss": 4.8347,
      "mean_token_accuracy": 0.23840138912200928,
      "num_tokens": 113233238.0,
      "step": 50725
    },
    {
      "entropy": 5.283716297149658,
      "epoch": 5.697478519683271,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022943537050129504,
      "loss": 4.665,
      "mean_token_accuracy": 0.24747377634048462,
      "num_tokens": 113244224.0,
      "step": 50730
    },
    {
      "entropy": 5.331808853149414,
      "epoch": 5.698040096591228,
      "grad_norm": 1.0,
      "learning_rate": 0.0002293960543770731,
      "loss": 4.8594,
      "mean_token_accuracy": 0.24174703508615494,
      "num_tokens": 113255892.0,
      "step": 50735
    },
    {
      "entropy": 5.268104219436646,
      "epoch": 5.698601673499185,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022935673970489847,
      "loss": 4.8095,
      "mean_token_accuracy": 0.23891334384679794,
      "num_tokens": 113268301.0,
      "step": 50740
    },
    {
      "entropy": 5.333758115768433,
      "epoch": 5.699163250407143,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022931742648602294,
      "loss": 4.8301,
      "mean_token_accuracy": 0.2351090803742409,
      "num_tokens": 113280521.0,
      "step": 50745
    },
    {
      "entropy": 5.3543061256408695,
      "epoch": 5.699724827315101,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022927811472169823,
      "loss": 4.7795,
      "mean_token_accuracy": 0.24883992969989777,
      "num_tokens": 113291004.0,
      "step": 50750
    },
    {
      "entropy": 5.459448337554932,
      "epoch": 5.700286404223059,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022923880441317606,
      "loss": 4.9598,
      "mean_token_accuracy": 0.22426485717296601,
      "num_tokens": 113301718.0,
      "step": 50755
    },
    {
      "entropy": 5.3613087177276615,
      "epoch": 5.700847981131016,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002291994955617082,
      "loss": 4.8515,
      "mean_token_accuracy": 0.2291525721549988,
      "num_tokens": 113312848.0,
      "step": 50760
    },
    {
      "entropy": 5.3625829219818115,
      "epoch": 5.701409558038973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002291601881685461,
      "loss": 4.8745,
      "mean_token_accuracy": 0.24060789197683335,
      "num_tokens": 113323380.0,
      "step": 50765
    },
    {
      "entropy": 5.299519109725952,
      "epoch": 5.701971134946931,
      "grad_norm": 1.109375,
      "learning_rate": 0.00022912088223494138,
      "loss": 4.7493,
      "mean_token_accuracy": 0.24485370069742202,
      "num_tokens": 113334406.0,
      "step": 50770
    },
    {
      "entropy": 5.274255180358887,
      "epoch": 5.7025327118548885,
      "grad_norm": 1.125,
      "learning_rate": 0.00022908157776214556,
      "loss": 4.7913,
      "mean_token_accuracy": 0.22991242557764052,
      "num_tokens": 113344902.0,
      "step": 50775
    },
    {
      "entropy": 5.350745010375976,
      "epoch": 5.703094288762846,
      "grad_norm": 1.0,
      "learning_rate": 0.00022904227475141015,
      "loss": 4.8557,
      "mean_token_accuracy": 0.2279915079474449,
      "num_tokens": 113355408.0,
      "step": 50780
    },
    {
      "entropy": 5.321910381317139,
      "epoch": 5.703655865670804,
      "grad_norm": 1.15625,
      "learning_rate": 0.00022900297320398654,
      "loss": 4.7164,
      "mean_token_accuracy": 0.25021869093179705,
      "num_tokens": 113365704.0,
      "step": 50785
    },
    {
      "entropy": 5.397921180725097,
      "epoch": 5.704217442578761,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022896367312112608,
      "loss": 4.8253,
      "mean_token_accuracy": 0.23364744633436202,
      "num_tokens": 113376918.0,
      "step": 50790
    },
    {
      "entropy": 5.198722457885742,
      "epoch": 5.704779019486718,
      "grad_norm": 0.9375,
      "learning_rate": 0.00022892437450408015,
      "loss": 4.696,
      "mean_token_accuracy": 0.24758501648902892,
      "num_tokens": 113388567.0,
      "step": 50795
    },
    {
      "entropy": 5.232199096679688,
      "epoch": 5.705340596394676,
      "grad_norm": 1.109375,
      "learning_rate": 0.00022888507735409997,
      "loss": 4.6875,
      "mean_token_accuracy": 0.24776776880025864,
      "num_tokens": 113399408.0,
      "step": 50800
    },
    {
      "entropy": 5.50755672454834,
      "epoch": 5.705902173302634,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022884578167243675,
      "loss": 5.0289,
      "mean_token_accuracy": 0.22923431694507598,
      "num_tokens": 113410605.0,
      "step": 50805
    },
    {
      "entropy": 5.288316011428833,
      "epoch": 5.706463750210592,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022880648746034183,
      "loss": 4.8132,
      "mean_token_accuracy": 0.23807477056980134,
      "num_tokens": 113422552.0,
      "step": 50810
    },
    {
      "entropy": 5.384940719604492,
      "epoch": 5.707025327118549,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022876719471906626,
      "loss": 4.8437,
      "mean_token_accuracy": 0.24064173698425292,
      "num_tokens": 113434423.0,
      "step": 50815
    },
    {
      "entropy": 5.277596664428711,
      "epoch": 5.707586904026506,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022872790344986117,
      "loss": 4.8484,
      "mean_token_accuracy": 0.23502280712127685,
      "num_tokens": 113446743.0,
      "step": 50820
    },
    {
      "entropy": 5.397785377502442,
      "epoch": 5.708148480934464,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022868861365397752,
      "loss": 4.8936,
      "mean_token_accuracy": 0.2255231872200966,
      "num_tokens": 113457630.0,
      "step": 50825
    },
    {
      "entropy": 5.348792648315429,
      "epoch": 5.7087100578424215,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022864932533266642,
      "loss": 4.7772,
      "mean_token_accuracy": 0.23757467567920684,
      "num_tokens": 113469637.0,
      "step": 50830
    },
    {
      "entropy": 5.3823925971984865,
      "epoch": 5.709271634750379,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022861003848717877,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23722977191209793,
      "num_tokens": 113480686.0,
      "step": 50835
    },
    {
      "entropy": 5.251714897155762,
      "epoch": 5.709833211658337,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022857075311876542,
      "loss": 4.7037,
      "mean_token_accuracy": 0.2446419969201088,
      "num_tokens": 113491952.0,
      "step": 50840
    },
    {
      "entropy": 5.18622407913208,
      "epoch": 5.710394788566294,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022853146922867734,
      "loss": 4.6865,
      "mean_token_accuracy": 0.24995354562997818,
      "num_tokens": 113502393.0,
      "step": 50845
    },
    {
      "entropy": 5.326667356491089,
      "epoch": 5.710956365474251,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002284921868181653,
      "loss": 4.8814,
      "mean_token_accuracy": 0.23814730644226073,
      "num_tokens": 113514012.0,
      "step": 50850
    },
    {
      "entropy": 5.294448280334473,
      "epoch": 5.711517942382209,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022845290588848012,
      "loss": 4.7909,
      "mean_token_accuracy": 0.2405279889702797,
      "num_tokens": 113525387.0,
      "step": 50855
    },
    {
      "entropy": 5.32027645111084,
      "epoch": 5.7120795192901666,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022841362644087244,
      "loss": 4.7888,
      "mean_token_accuracy": 0.23910442888736724,
      "num_tokens": 113536478.0,
      "step": 50860
    },
    {
      "entropy": 5.340518283843994,
      "epoch": 5.712641096198125,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022837434847659295,
      "loss": 4.7435,
      "mean_token_accuracy": 0.24288913160562514,
      "num_tokens": 113547737.0,
      "step": 50865
    },
    {
      "entropy": 5.446155214309693,
      "epoch": 5.713202673106082,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002283350719968923,
      "loss": 4.9657,
      "mean_token_accuracy": 0.23434963822364807,
      "num_tokens": 113559326.0,
      "step": 50870
    },
    {
      "entropy": 5.301633548736572,
      "epoch": 5.713764250014039,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000228295797003021,
      "loss": 4.7344,
      "mean_token_accuracy": 0.2411084070801735,
      "num_tokens": 113569737.0,
      "step": 50875
    },
    {
      "entropy": 5.309129428863526,
      "epoch": 5.714325826921997,
      "grad_norm": 1.0,
      "learning_rate": 0.00022825652349622966,
      "loss": 4.7675,
      "mean_token_accuracy": 0.24523212164640426,
      "num_tokens": 113581154.0,
      "step": 50880
    },
    {
      "entropy": 5.325250911712646,
      "epoch": 5.714887403829954,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022821725147776877,
      "loss": 4.7756,
      "mean_token_accuracy": 0.2419283866882324,
      "num_tokens": 113592125.0,
      "step": 50885
    },
    {
      "entropy": 5.243553256988525,
      "epoch": 5.7154489807379125,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022817798094888875,
      "loss": 4.6568,
      "mean_token_accuracy": 0.24634029120206832,
      "num_tokens": 113602384.0,
      "step": 50890
    },
    {
      "entropy": 5.228353834152221,
      "epoch": 5.71601055764587,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022813871191083995,
      "loss": 4.7014,
      "mean_token_accuracy": 0.23961537778377534,
      "num_tokens": 113613191.0,
      "step": 50895
    },
    {
      "entropy": 5.247710943222046,
      "epoch": 5.716572134553827,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022809944436487278,
      "loss": 4.7715,
      "mean_token_accuracy": 0.2362819045782089,
      "num_tokens": 113624615.0,
      "step": 50900
    },
    {
      "entropy": 5.315955972671508,
      "epoch": 5.717133711461785,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022806017831223746,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2377579018473625,
      "num_tokens": 113636461.0,
      "step": 50905
    },
    {
      "entropy": 5.372476291656494,
      "epoch": 5.717695288369742,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022802091375418422,
      "loss": 4.8525,
      "mean_token_accuracy": 0.2356565311551094,
      "num_tokens": 113647452.0,
      "step": 50910
    },
    {
      "entropy": 5.425834131240845,
      "epoch": 5.7182568652776995,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022798165069196335,
      "loss": 4.8675,
      "mean_token_accuracy": 0.23659952580928803,
      "num_tokens": 113658358.0,
      "step": 50915
    },
    {
      "entropy": 5.378121757507325,
      "epoch": 5.718818442185658,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022794238912682497,
      "loss": 4.8503,
      "mean_token_accuracy": 0.23739357441663742,
      "num_tokens": 113668649.0,
      "step": 50920
    },
    {
      "entropy": 5.315393877029419,
      "epoch": 5.719380019093615,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022790312906001919,
      "loss": 4.7648,
      "mean_token_accuracy": 0.2309473931789398,
      "num_tokens": 113679802.0,
      "step": 50925
    },
    {
      "entropy": 5.319543075561524,
      "epoch": 5.719941596001572,
      "grad_norm": 0.9375,
      "learning_rate": 0.00022786387049279606,
      "loss": 4.8402,
      "mean_token_accuracy": 0.23391697257757188,
      "num_tokens": 113691879.0,
      "step": 50930
    },
    {
      "entropy": 5.493599081039429,
      "epoch": 5.72050317290953,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022782461342640554,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23335328102111816,
      "num_tokens": 113701854.0,
      "step": 50935
    },
    {
      "entropy": 5.33527626991272,
      "epoch": 5.721064749817487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022778535786209758,
      "loss": 4.7823,
      "mean_token_accuracy": 0.24058614820241928,
      "num_tokens": 113712771.0,
      "step": 50940
    },
    {
      "entropy": 5.27732286453247,
      "epoch": 5.7216263267254455,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002277461038011222,
      "loss": 4.7097,
      "mean_token_accuracy": 0.24012369513511658,
      "num_tokens": 113723590.0,
      "step": 50945
    },
    {
      "entropy": 5.398259115219116,
      "epoch": 5.722187903633403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022770685124472907,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2377983510494232,
      "num_tokens": 113735451.0,
      "step": 50950
    },
    {
      "entropy": 5.247649717330932,
      "epoch": 5.72274948054136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022766760019416826,
      "loss": 4.7362,
      "mean_token_accuracy": 0.23561370223760605,
      "num_tokens": 113746455.0,
      "step": 50955
    },
    {
      "entropy": 5.270652151107788,
      "epoch": 5.723311057449318,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022762835065068943,
      "loss": 4.7167,
      "mean_token_accuracy": 0.25138472616672514,
      "num_tokens": 113757616.0,
      "step": 50960
    },
    {
      "entropy": 5.269488382339477,
      "epoch": 5.723872634357275,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022758910261554228,
      "loss": 4.8001,
      "mean_token_accuracy": 0.23454590886831284,
      "num_tokens": 113769187.0,
      "step": 50965
    },
    {
      "entropy": 5.270131540298462,
      "epoch": 5.724434211265232,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022754985608997648,
      "loss": 4.7602,
      "mean_token_accuracy": 0.24329565167427064,
      "num_tokens": 113779690.0,
      "step": 50970
    },
    {
      "entropy": 5.3525238037109375,
      "epoch": 5.7249957881731905,
      "grad_norm": 1.109375,
      "learning_rate": 0.00022751061107524167,
      "loss": 4.8581,
      "mean_token_accuracy": 0.23489189893007278,
      "num_tokens": 113789888.0,
      "step": 50975
    },
    {
      "entropy": 5.362554597854614,
      "epoch": 5.725557365081148,
      "grad_norm": 1.0,
      "learning_rate": 0.00022747136757258742,
      "loss": 4.8469,
      "mean_token_accuracy": 0.23648446649312974,
      "num_tokens": 113800577.0,
      "step": 50980
    },
    {
      "entropy": 5.217343330383301,
      "epoch": 5.726118941989105,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002274321255832632,
      "loss": 4.6712,
      "mean_token_accuracy": 0.25313189178705214,
      "num_tokens": 113812262.0,
      "step": 50985
    },
    {
      "entropy": 5.295279884338379,
      "epoch": 5.726680518897063,
      "grad_norm": 0.90625,
      "learning_rate": 0.00022739288510851863,
      "loss": 4.7536,
      "mean_token_accuracy": 0.2437872990965843,
      "num_tokens": 113824468.0,
      "step": 50990
    },
    {
      "entropy": 5.322895479202271,
      "epoch": 5.72724209580502,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022735364614960302,
      "loss": 4.8082,
      "mean_token_accuracy": 0.2425532028079033,
      "num_tokens": 113835238.0,
      "step": 50995
    },
    {
      "entropy": 5.2680665969848635,
      "epoch": 5.727803672712978,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022731440870776582,
      "loss": 4.7891,
      "mean_token_accuracy": 0.23824004530906678,
      "num_tokens": 113846275.0,
      "step": 51000
    },
    {
      "epoch": 5.727803672712978,
      "eval_entropy": 5.115499059348983,
      "eval_loss": 4.9925947189331055,
      "eval_mean_token_accuracy": 0.23540482873773033,
      "eval_num_tokens": 113846275.0,
      "eval_runtime": 23.8902,
      "eval_samples_per_second": 1298.146,
      "eval_steps_per_second": 162.284,
      "step": 51000
    },
    {
      "entropy": 5.323248672485351,
      "epoch": 5.728365249620936,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022727517278425637,
      "loss": 4.7195,
      "mean_token_accuracy": 0.2379387781023979,
      "num_tokens": 113857666.0,
      "step": 51005
    },
    {
      "entropy": 5.355527067184449,
      "epoch": 5.728926826528893,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00022723593838032391,
      "loss": 4.8232,
      "mean_token_accuracy": 0.2316133886575699,
      "num_tokens": 113868279.0,
      "step": 51010
    },
    {
      "entropy": 5.35429105758667,
      "epoch": 5.729488403436851,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022719670549721765,
      "loss": 4.8636,
      "mean_token_accuracy": 0.23542719036340715,
      "num_tokens": 113879401.0,
      "step": 51015
    },
    {
      "entropy": 5.291626262664795,
      "epoch": 5.730049980344808,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00022715747413618677,
      "loss": 4.7996,
      "mean_token_accuracy": 0.23818872570991517,
      "num_tokens": 113890435.0,
      "step": 51020
    },
    {
      "entropy": 5.240498065948486,
      "epoch": 5.730611557252765,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022711824429848054,
      "loss": 4.6834,
      "mean_token_accuracy": 0.2471496045589447,
      "num_tokens": 113902169.0,
      "step": 51025
    },
    {
      "entropy": 5.399691581726074,
      "epoch": 5.7311731341607235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000227079015985348,
      "loss": 4.9192,
      "mean_token_accuracy": 0.2308892324566841,
      "num_tokens": 113914077.0,
      "step": 51030
    },
    {
      "entropy": 5.259225368499756,
      "epoch": 5.731734711068681,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002270397891980382,
      "loss": 4.6661,
      "mean_token_accuracy": 0.25196210443973543,
      "num_tokens": 113925889.0,
      "step": 51035
    },
    {
      "entropy": 5.423564195632935,
      "epoch": 5.732296287976638,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022700056393780012,
      "loss": 4.8761,
      "mean_token_accuracy": 0.23512988537549973,
      "num_tokens": 113936636.0,
      "step": 51040
    },
    {
      "entropy": 5.28269476890564,
      "epoch": 5.732857864884596,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022696134020588267,
      "loss": 4.7462,
      "mean_token_accuracy": 0.2454352483153343,
      "num_tokens": 113948078.0,
      "step": 51045
    },
    {
      "entropy": 5.317790889739991,
      "epoch": 5.733419441792553,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002269221180035348,
      "loss": 4.7818,
      "mean_token_accuracy": 0.24009400308132173,
      "num_tokens": 113959532.0,
      "step": 51050
    },
    {
      "entropy": 5.335780668258667,
      "epoch": 5.733981018700511,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022688289733200522,
      "loss": 4.8068,
      "mean_token_accuracy": 0.23624992221593857,
      "num_tokens": 113971443.0,
      "step": 51055
    },
    {
      "entropy": 5.333909225463867,
      "epoch": 5.734542595608469,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000226843678192543,
      "loss": 4.7668,
      "mean_token_accuracy": 0.24689988791942596,
      "num_tokens": 113981435.0,
      "step": 51060
    },
    {
      "entropy": 5.24518723487854,
      "epoch": 5.735104172516426,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002268044605863967,
      "loss": 4.7422,
      "mean_token_accuracy": 0.24779165983200074,
      "num_tokens": 113991755.0,
      "step": 51065
    },
    {
      "entropy": 5.338075637817383,
      "epoch": 5.735665749424384,
      "grad_norm": 1.125,
      "learning_rate": 0.00022676524451481505,
      "loss": 4.8466,
      "mean_token_accuracy": 0.24133355617523194,
      "num_tokens": 114002506.0,
      "step": 51070
    },
    {
      "entropy": 5.252970933914185,
      "epoch": 5.736227326332341,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022672602997904674,
      "loss": 4.7299,
      "mean_token_accuracy": 0.24329013973474503,
      "num_tokens": 114014743.0,
      "step": 51075
    },
    {
      "entropy": 5.203706073760986,
      "epoch": 5.736788903240299,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022668681698034032,
      "loss": 4.7325,
      "mean_token_accuracy": 0.24161656945943832,
      "num_tokens": 114025355.0,
      "step": 51080
    },
    {
      "entropy": 5.333952474594116,
      "epoch": 5.737350480148256,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022664760551994445,
      "loss": 4.8472,
      "mean_token_accuracy": 0.23403352200984956,
      "num_tokens": 114036600.0,
      "step": 51085
    },
    {
      "entropy": 5.35011920928955,
      "epoch": 5.737912057056214,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00022660839559910756,
      "loss": 4.792,
      "mean_token_accuracy": 0.24633708000183105,
      "num_tokens": 114047828.0,
      "step": 51090
    },
    {
      "entropy": 5.3461692333221436,
      "epoch": 5.738473633964172,
      "grad_norm": 1.03125,
      "learning_rate": 0.000226569187219078,
      "loss": 4.7527,
      "mean_token_accuracy": 0.2517297640442848,
      "num_tokens": 114058101.0,
      "step": 51095
    },
    {
      "entropy": 5.364392709732056,
      "epoch": 5.739035210872129,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002265299803811044,
      "loss": 4.8233,
      "mean_token_accuracy": 0.23378075361251832,
      "num_tokens": 114069255.0,
      "step": 51100
    },
    {
      "entropy": 5.310218715667725,
      "epoch": 5.739596787780086,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022649077508643507,
      "loss": 4.8658,
      "mean_token_accuracy": 0.2405985802412033,
      "num_tokens": 114081363.0,
      "step": 51105
    },
    {
      "entropy": 5.40250186920166,
      "epoch": 5.740158364688044,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022645157133631823,
      "loss": 4.8996,
      "mean_token_accuracy": 0.23601340800523757,
      "num_tokens": 114093276.0,
      "step": 51110
    },
    {
      "entropy": 5.394738483428955,
      "epoch": 5.7407199415960015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022641236913200215,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2316654533147812,
      "num_tokens": 114104178.0,
      "step": 51115
    },
    {
      "entropy": 5.305172252655029,
      "epoch": 5.741281518503959,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022637316847473515,
      "loss": 4.8295,
      "mean_token_accuracy": 0.23546912968158723,
      "num_tokens": 114115366.0,
      "step": 51120
    },
    {
      "entropy": 5.252777433395385,
      "epoch": 5.741843095411917,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022633396936576534,
      "loss": 4.719,
      "mean_token_accuracy": 0.2453773632645607,
      "num_tokens": 114126417.0,
      "step": 51125
    },
    {
      "entropy": 5.236069583892823,
      "epoch": 5.742404672319874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022629477180634072,
      "loss": 4.7421,
      "mean_token_accuracy": 0.23877759277820587,
      "num_tokens": 114136943.0,
      "step": 51130
    },
    {
      "entropy": 5.33218412399292,
      "epoch": 5.742966249227832,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002262555757977095,
      "loss": 4.7727,
      "mean_token_accuracy": 0.23420901596546173,
      "num_tokens": 114147347.0,
      "step": 51135
    },
    {
      "entropy": 5.263463115692138,
      "epoch": 5.743527826135789,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022621638134111966,
      "loss": 4.7426,
      "mean_token_accuracy": 0.2413904294371605,
      "num_tokens": 114158419.0,
      "step": 51140
    },
    {
      "entropy": 5.268434524536133,
      "epoch": 5.744089403043747,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002261771884378192,
      "loss": 4.7621,
      "mean_token_accuracy": 0.24676626771688462,
      "num_tokens": 114168731.0,
      "step": 51145
    },
    {
      "entropy": 5.274119138717651,
      "epoch": 5.744650979951705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002261379970890559,
      "loss": 4.7334,
      "mean_token_accuracy": 0.24730513840913773,
      "num_tokens": 114180062.0,
      "step": 51150
    },
    {
      "entropy": 5.398240852355957,
      "epoch": 5.745212556859662,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022609880729607785,
      "loss": 4.8684,
      "mean_token_accuracy": 0.23451490253210067,
      "num_tokens": 114191340.0,
      "step": 51155
    },
    {
      "entropy": 5.293097352981567,
      "epoch": 5.745774133767619,
      "grad_norm": 1.125,
      "learning_rate": 0.00022605961906013266,
      "loss": 4.7608,
      "mean_token_accuracy": 0.24814313948154448,
      "num_tokens": 114201934.0,
      "step": 51160
    },
    {
      "entropy": 5.38550877571106,
      "epoch": 5.746335710675577,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002260204323824681,
      "loss": 4.7961,
      "mean_token_accuracy": 0.23988536596298218,
      "num_tokens": 114212562.0,
      "step": 51165
    },
    {
      "entropy": 5.232830667495728,
      "epoch": 5.7468972875835345,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022598124726433205,
      "loss": 4.6778,
      "mean_token_accuracy": 0.24768627285957337,
      "num_tokens": 114223668.0,
      "step": 51170
    },
    {
      "entropy": 5.154384660720825,
      "epoch": 5.747458864491492,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022594206370697212,
      "loss": 4.6664,
      "mean_token_accuracy": 0.2570901349186897,
      "num_tokens": 114235391.0,
      "step": 51175
    },
    {
      "entropy": 5.253238248825073,
      "epoch": 5.74802044139945,
      "grad_norm": 1.21875,
      "learning_rate": 0.00022590288171163593,
      "loss": 4.7538,
      "mean_token_accuracy": 0.2393921732902527,
      "num_tokens": 114246267.0,
      "step": 51180
    },
    {
      "entropy": 5.311155557632446,
      "epoch": 5.748582018307407,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000225863701279571,
      "loss": 4.7166,
      "mean_token_accuracy": 0.2527556478977203,
      "num_tokens": 114257394.0,
      "step": 51185
    },
    {
      "entropy": 5.322428512573242,
      "epoch": 5.749143595215365,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002258245224120249,
      "loss": 4.8205,
      "mean_token_accuracy": 0.23979092240333558,
      "num_tokens": 114268614.0,
      "step": 51190
    },
    {
      "entropy": 5.359666585922241,
      "epoch": 5.749705172123322,
      "grad_norm": 1.109375,
      "learning_rate": 0.00022578534511024506,
      "loss": 4.8303,
      "mean_token_accuracy": 0.23358473032712937,
      "num_tokens": 114279654.0,
      "step": 51195
    },
    {
      "entropy": 5.3894116401672365,
      "epoch": 5.7502667490312795,
      "grad_norm": 1.15625,
      "learning_rate": 0.00022574616937547887,
      "loss": 4.9331,
      "mean_token_accuracy": 0.235675410926342,
      "num_tokens": 114291181.0,
      "step": 51200
    },
    {
      "entropy": 5.397780084609986,
      "epoch": 5.750828325939238,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022570699520897382,
      "loss": 4.9816,
      "mean_token_accuracy": 0.2292216420173645,
      "num_tokens": 114302729.0,
      "step": 51205
    },
    {
      "entropy": 5.286508178710937,
      "epoch": 5.751389902847195,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002256678226119771,
      "loss": 4.7516,
      "mean_token_accuracy": 0.24058356285095214,
      "num_tokens": 114314220.0,
      "step": 51210
    },
    {
      "entropy": 5.379350996017456,
      "epoch": 5.751951479755152,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022562865158573615,
      "loss": 4.7666,
      "mean_token_accuracy": 0.2405677855014801,
      "num_tokens": 114324950.0,
      "step": 51215
    },
    {
      "entropy": 5.318822145462036,
      "epoch": 5.75251305666311,
      "grad_norm": 0.921875,
      "learning_rate": 0.000225589482131498,
      "loss": 4.7855,
      "mean_token_accuracy": 0.24441753476858138,
      "num_tokens": 114338092.0,
      "step": 51220
    },
    {
      "entropy": 5.283265161514282,
      "epoch": 5.753074633571067,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022555031425050993,
      "loss": 4.7134,
      "mean_token_accuracy": 0.2425044521689415,
      "num_tokens": 114348184.0,
      "step": 51225
    },
    {
      "entropy": 5.2763965129852295,
      "epoch": 5.753636210479025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000225511147944019,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23975655883550645,
      "num_tokens": 114358156.0,
      "step": 51230
    },
    {
      "entropy": 5.269646167755127,
      "epoch": 5.754197787386983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022547198321327223,
      "loss": 4.7634,
      "mean_token_accuracy": 0.2446499839425087,
      "num_tokens": 114369330.0,
      "step": 51235
    },
    {
      "entropy": 5.405162239074707,
      "epoch": 5.75475936429494,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002254328200595168,
      "loss": 4.8888,
      "mean_token_accuracy": 0.23165274411439896,
      "num_tokens": 114380923.0,
      "step": 51240
    },
    {
      "entropy": 5.35341215133667,
      "epoch": 5.755320941202898,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022539365848399967,
      "loss": 4.8783,
      "mean_token_accuracy": 0.23412137478590012,
      "num_tokens": 114392281.0,
      "step": 51245
    },
    {
      "entropy": 5.406776285171508,
      "epoch": 5.755882518110855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022535449848796763,
      "loss": 4.8038,
      "mean_token_accuracy": 0.24122519046068192,
      "num_tokens": 114403295.0,
      "step": 51250
    },
    {
      "entropy": 5.39395923614502,
      "epoch": 5.7564440950188125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022531534007266764,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2369815558195114,
      "num_tokens": 114413507.0,
      "step": 51255
    },
    {
      "entropy": 5.296884489059448,
      "epoch": 5.757005671926771,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022527618323934652,
      "loss": 4.8014,
      "mean_token_accuracy": 0.24328305423259736,
      "num_tokens": 114426394.0,
      "step": 51260
    },
    {
      "entropy": 5.380506992340088,
      "epoch": 5.757567248834728,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022523702798925093,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23547255247831345,
      "num_tokens": 114438704.0,
      "step": 51265
    },
    {
      "entropy": 5.219447660446167,
      "epoch": 5.758128825742686,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022519787432362777,
      "loss": 4.7289,
      "mean_token_accuracy": 0.24249451607465744,
      "num_tokens": 114449881.0,
      "step": 51270
    },
    {
      "entropy": 5.31497802734375,
      "epoch": 5.758690402650643,
      "grad_norm": 1.0,
      "learning_rate": 0.00022515872224372353,
      "loss": 4.807,
      "mean_token_accuracy": 0.2391219586133957,
      "num_tokens": 114461504.0,
      "step": 51275
    },
    {
      "entropy": 5.345953798294067,
      "epoch": 5.7592519795586,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002251195717507849,
      "loss": 4.859,
      "mean_token_accuracy": 0.22882412523031234,
      "num_tokens": 114471901.0,
      "step": 51280
    },
    {
      "entropy": 5.275282716751098,
      "epoch": 5.7598135564665585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022508042284605852,
      "loss": 4.7709,
      "mean_token_accuracy": 0.24114848375320436,
      "num_tokens": 114483346.0,
      "step": 51285
    },
    {
      "entropy": 5.357812452316284,
      "epoch": 5.760375133374516,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002250412755307908,
      "loss": 4.8566,
      "mean_token_accuracy": 0.24404732286930084,
      "num_tokens": 114494802.0,
      "step": 51290
    },
    {
      "entropy": 5.32595386505127,
      "epoch": 5.760936710282473,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022500212980622818,
      "loss": 4.7208,
      "mean_token_accuracy": 0.24138077795505525,
      "num_tokens": 114505856.0,
      "step": 51295
    },
    {
      "entropy": 5.405333566665649,
      "epoch": 5.761498287190431,
      "grad_norm": 1.15625,
      "learning_rate": 0.00022496298567361724,
      "loss": 4.876,
      "mean_token_accuracy": 0.24198714792728424,
      "num_tokens": 114515944.0,
      "step": 51300
    },
    {
      "entropy": 5.377615594863892,
      "epoch": 5.762059864098388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022492384313420416,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22941087037324906,
      "num_tokens": 114527401.0,
      "step": 51305
    },
    {
      "entropy": 5.28571138381958,
      "epoch": 5.762621441006345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022488470218923534,
      "loss": 4.716,
      "mean_token_accuracy": 0.2456533804535866,
      "num_tokens": 114538335.0,
      "step": 51310
    },
    {
      "entropy": 5.303635120391846,
      "epoch": 5.7631830179143035,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022484556283995706,
      "loss": 4.76,
      "mean_token_accuracy": 0.24764000922441481,
      "num_tokens": 114550060.0,
      "step": 51315
    },
    {
      "entropy": 5.3178668975830075,
      "epoch": 5.763744594822261,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002248064250876155,
      "loss": 4.8541,
      "mean_token_accuracy": 0.23919204920530318,
      "num_tokens": 114561912.0,
      "step": 51320
    },
    {
      "entropy": 5.294459009170533,
      "epoch": 5.764306171730219,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022476728893345687,
      "loss": 4.8089,
      "mean_token_accuracy": 0.24784767180681228,
      "num_tokens": 114572910.0,
      "step": 51325
    },
    {
      "entropy": 5.316967725753784,
      "epoch": 5.764867748638176,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022472815437872724,
      "loss": 4.7931,
      "mean_token_accuracy": 0.23867156952619553,
      "num_tokens": 114584875.0,
      "step": 51330
    },
    {
      "entropy": 5.3189873695373535,
      "epoch": 5.765429325546133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022468902142467267,
      "loss": 4.7757,
      "mean_token_accuracy": 0.23896090388298036,
      "num_tokens": 114595123.0,
      "step": 51335
    },
    {
      "entropy": 5.367301273345947,
      "epoch": 5.765990902454091,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022464989007253922,
      "loss": 4.9026,
      "mean_token_accuracy": 0.22482929974794388,
      "num_tokens": 114606848.0,
      "step": 51340
    },
    {
      "entropy": 5.350270938873291,
      "epoch": 5.766552479362049,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002246107603235727,
      "loss": 4.805,
      "mean_token_accuracy": 0.23897681385278702,
      "num_tokens": 114618278.0,
      "step": 51345
    },
    {
      "entropy": 5.312136030197143,
      "epoch": 5.767114056270006,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022457163217901916,
      "loss": 4.7463,
      "mean_token_accuracy": 0.24646656066179276,
      "num_tokens": 114628158.0,
      "step": 51350
    },
    {
      "entropy": 5.372099304199219,
      "epoch": 5.767675633177964,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002245325056401244,
      "loss": 4.8071,
      "mean_token_accuracy": 0.2432974711060524,
      "num_tokens": 114639446.0,
      "step": 51355
    },
    {
      "entropy": 5.388066387176513,
      "epoch": 5.768237210085921,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00022449338070813424,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22843966633081436,
      "num_tokens": 114650995.0,
      "step": 51360
    },
    {
      "entropy": 5.228050470352173,
      "epoch": 5.768798786993878,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022445425738429442,
      "loss": 4.6706,
      "mean_token_accuracy": 0.2513738974928856,
      "num_tokens": 114663484.0,
      "step": 51365
    },
    {
      "entropy": 5.330328178405762,
      "epoch": 5.7693603639018365,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022441513566985068,
      "loss": 4.837,
      "mean_token_accuracy": 0.23621643632650374,
      "num_tokens": 114675186.0,
      "step": 51370
    },
    {
      "entropy": 5.342471027374268,
      "epoch": 5.769921940809794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022437601556604865,
      "loss": 4.7719,
      "mean_token_accuracy": 0.24322451651096344,
      "num_tokens": 114687539.0,
      "step": 51375
    },
    {
      "entropy": 5.363106822967529,
      "epoch": 5.770483517717752,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022433689707413373,
      "loss": 4.8552,
      "mean_token_accuracy": 0.24017221331596375,
      "num_tokens": 114698804.0,
      "step": 51380
    },
    {
      "entropy": 5.376868867874146,
      "epoch": 5.771045094625709,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022429778019535186,
      "loss": 4.8243,
      "mean_token_accuracy": 0.23726645559072496,
      "num_tokens": 114708927.0,
      "step": 51385
    },
    {
      "entropy": 5.323652362823486,
      "epoch": 5.771606671533666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022425866493094827,
      "loss": 4.8613,
      "mean_token_accuracy": 0.24192046970129014,
      "num_tokens": 114719835.0,
      "step": 51390
    },
    {
      "entropy": 5.323780965805054,
      "epoch": 5.772168248441624,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022421955128216854,
      "loss": 4.7868,
      "mean_token_accuracy": 0.2414919063448906,
      "num_tokens": 114730420.0,
      "step": 51395
    },
    {
      "entropy": 5.3395918846130375,
      "epoch": 5.772729825349582,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002241804392502579,
      "loss": 4.7397,
      "mean_token_accuracy": 0.2436084732413292,
      "num_tokens": 114740467.0,
      "step": 51400
    },
    {
      "entropy": 5.246948719024658,
      "epoch": 5.773291402257539,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022414132883646183,
      "loss": 4.7462,
      "mean_token_accuracy": 0.24216579049825668,
      "num_tokens": 114751981.0,
      "step": 51405
    },
    {
      "entropy": 5.321401882171631,
      "epoch": 5.773852979165497,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00022410222004202557,
      "loss": 4.8496,
      "mean_token_accuracy": 0.24375711530447006,
      "num_tokens": 114763054.0,
      "step": 51410
    },
    {
      "entropy": 5.4112447738647464,
      "epoch": 5.774414556073454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002240631128681943,
      "loss": 4.9416,
      "mean_token_accuracy": 0.22880327552556992,
      "num_tokens": 114774540.0,
      "step": 51415
    },
    {
      "entropy": 5.367655515670776,
      "epoch": 5.774976132981411,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022402400731621336,
      "loss": 4.7839,
      "mean_token_accuracy": 0.23801037222146987,
      "num_tokens": 114785964.0,
      "step": 51420
    },
    {
      "entropy": 5.389012145996094,
      "epoch": 5.775537709889369,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002239849033873278,
      "loss": 4.8143,
      "mean_token_accuracy": 0.23612545877695085,
      "num_tokens": 114797985.0,
      "step": 51425
    },
    {
      "entropy": 5.343607711791992,
      "epoch": 5.776099286797327,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00022394580108278263,
      "loss": 4.804,
      "mean_token_accuracy": 0.2486430287361145,
      "num_tokens": 114809082.0,
      "step": 51430
    },
    {
      "entropy": 5.335655975341797,
      "epoch": 5.776660863705285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000223906700403823,
      "loss": 4.8515,
      "mean_token_accuracy": 0.23501545190811157,
      "num_tokens": 114820234.0,
      "step": 51435
    },
    {
      "entropy": 5.230275583267212,
      "epoch": 5.777222440613242,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022386760135169388,
      "loss": 4.6969,
      "mean_token_accuracy": 0.24931626468896867,
      "num_tokens": 114829901.0,
      "step": 51440
    },
    {
      "entropy": 5.284165477752685,
      "epoch": 5.777784017521199,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022382850392764015,
      "loss": 4.7852,
      "mean_token_accuracy": 0.2374560832977295,
      "num_tokens": 114839959.0,
      "step": 51445
    },
    {
      "entropy": 5.3351202487945555,
      "epoch": 5.778345594429157,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022378940813290666,
      "loss": 4.8606,
      "mean_token_accuracy": 0.23566230982542039,
      "num_tokens": 114851385.0,
      "step": 51450
    },
    {
      "entropy": 5.3154796123504635,
      "epoch": 5.7789071713371145,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022375031396873818,
      "loss": 4.7641,
      "mean_token_accuracy": 0.24268782287836074,
      "num_tokens": 114861971.0,
      "step": 51455
    },
    {
      "entropy": 5.265054559707641,
      "epoch": 5.779468748245073,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022371122143637973,
      "loss": 4.6892,
      "mean_token_accuracy": 0.24518761932849883,
      "num_tokens": 114873511.0,
      "step": 51460
    },
    {
      "entropy": 5.3151203155517575,
      "epoch": 5.78003032515303,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022367213053707587,
      "loss": 4.8242,
      "mean_token_accuracy": 0.2332860827445984,
      "num_tokens": 114885166.0,
      "step": 51465
    },
    {
      "entropy": 5.344426774978638,
      "epoch": 5.780591902060987,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022363304127207123,
      "loss": 4.8078,
      "mean_token_accuracy": 0.23814645558595657,
      "num_tokens": 114896015.0,
      "step": 51470
    },
    {
      "entropy": 5.320563697814942,
      "epoch": 5.781153478968945,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022359395364261048,
      "loss": 4.7564,
      "mean_token_accuracy": 0.244650399684906,
      "num_tokens": 114906369.0,
      "step": 51475
    },
    {
      "entropy": 5.2923667430877686,
      "epoch": 5.781715055876902,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022355486764993822,
      "loss": 4.7404,
      "mean_token_accuracy": 0.2464059665799141,
      "num_tokens": 114917991.0,
      "step": 51480
    },
    {
      "entropy": 5.270185279846191,
      "epoch": 5.78227663278486,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022351578329529893,
      "loss": 4.7696,
      "mean_token_accuracy": 0.24113014787435533,
      "num_tokens": 114928506.0,
      "step": 51485
    },
    {
      "entropy": 5.269151830673218,
      "epoch": 5.782838209692818,
      "grad_norm": 1.0625,
      "learning_rate": 0.000223476700579937,
      "loss": 4.731,
      "mean_token_accuracy": 0.24178647994995117,
      "num_tokens": 114939702.0,
      "step": 51490
    },
    {
      "entropy": 5.3656494140625,
      "epoch": 5.783399786600775,
      "grad_norm": 1.03125,
      "learning_rate": 0.000223437619505097,
      "loss": 4.8433,
      "mean_token_accuracy": 0.2451056182384491,
      "num_tokens": 114952062.0,
      "step": 51495
    },
    {
      "entropy": 5.305613088607788,
      "epoch": 5.783961363508732,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022339854007202316,
      "loss": 4.7226,
      "mean_token_accuracy": 0.24643427431583403,
      "num_tokens": 114962166.0,
      "step": 51500
    },
    {
      "entropy": 5.281415271759033,
      "epoch": 5.78452294041669,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002233594622819598,
      "loss": 4.799,
      "mean_token_accuracy": 0.23726112842559816,
      "num_tokens": 114972795.0,
      "step": 51505
    },
    {
      "entropy": 5.4138189315795895,
      "epoch": 5.7850845173246475,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00022332038613615118,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2319990172982216,
      "num_tokens": 114983761.0,
      "step": 51510
    },
    {
      "entropy": 5.3415721416473385,
      "epoch": 5.785646094232606,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022328131163584154,
      "loss": 4.8347,
      "mean_token_accuracy": 0.23886471688747407,
      "num_tokens": 114995071.0,
      "step": 51515
    },
    {
      "entropy": 5.372035837173462,
      "epoch": 5.786207671140563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000223242238782275,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2369135022163391,
      "num_tokens": 115007063.0,
      "step": 51520
    },
    {
      "entropy": 5.324918222427368,
      "epoch": 5.78676924804852,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022320316757669557,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23629413843154906,
      "num_tokens": 115017839.0,
      "step": 51525
    },
    {
      "entropy": 5.287717485427857,
      "epoch": 5.787330824956478,
      "grad_norm": 1.0,
      "learning_rate": 0.00022316409802034743,
      "loss": 4.771,
      "mean_token_accuracy": 0.24148973673582078,
      "num_tokens": 115029154.0,
      "step": 51530
    },
    {
      "entropy": 5.2532322883605955,
      "epoch": 5.787892401864435,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002231250301144745,
      "loss": 4.6991,
      "mean_token_accuracy": 0.24504282772541047,
      "num_tokens": 115039818.0,
      "step": 51535
    },
    {
      "entropy": 5.306731748580932,
      "epoch": 5.7884539787723925,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002230859638603208,
      "loss": 4.7823,
      "mean_token_accuracy": 0.2314349427819252,
      "num_tokens": 115050531.0,
      "step": 51540
    },
    {
      "entropy": 5.331207895278931,
      "epoch": 5.789015555680351,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022304689925913007,
      "loss": 4.81,
      "mean_token_accuracy": 0.24350003600120546,
      "num_tokens": 115060906.0,
      "step": 51545
    },
    {
      "entropy": 5.423318338394165,
      "epoch": 5.789577132588308,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022300783631214627,
      "loss": 4.9401,
      "mean_token_accuracy": 0.22641288489103317,
      "num_tokens": 115072231.0,
      "step": 51550
    },
    {
      "entropy": 5.295700311660767,
      "epoch": 5.790138709496265,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022296877502061307,
      "loss": 4.8388,
      "mean_token_accuracy": 0.23854895681142807,
      "num_tokens": 115083931.0,
      "step": 51555
    },
    {
      "entropy": 5.315231084823608,
      "epoch": 5.790700286404223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022292971538577427,
      "loss": 4.8233,
      "mean_token_accuracy": 0.24155119955539703,
      "num_tokens": 115095471.0,
      "step": 51560
    },
    {
      "entropy": 5.293683624267578,
      "epoch": 5.79126186331218,
      "grad_norm": 0.984375,
      "learning_rate": 0.00022289065740887354,
      "loss": 4.7644,
      "mean_token_accuracy": 0.24360760152339936,
      "num_tokens": 115107174.0,
      "step": 51565
    },
    {
      "entropy": 5.361393308639526,
      "epoch": 5.7918234402201385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022285160109115448,
      "loss": 4.8449,
      "mean_token_accuracy": 0.24522320926189423,
      "num_tokens": 115117328.0,
      "step": 51570
    },
    {
      "entropy": 5.377897119522094,
      "epoch": 5.792385017128096,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022281254643386072,
      "loss": 4.911,
      "mean_token_accuracy": 0.23208729773759842,
      "num_tokens": 115128299.0,
      "step": 51575
    },
    {
      "entropy": 5.353327655792237,
      "epoch": 5.792946594036053,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022277349343823567,
      "loss": 4.8446,
      "mean_token_accuracy": 0.23603860288858414,
      "num_tokens": 115139521.0,
      "step": 51580
    },
    {
      "entropy": 5.380072593688965,
      "epoch": 5.793508170944011,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022273444210552286,
      "loss": 4.7917,
      "mean_token_accuracy": 0.23643328845500947,
      "num_tokens": 115149869.0,
      "step": 51585
    },
    {
      "entropy": 5.271111726760864,
      "epoch": 5.794069747851968,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022269539243696568,
      "loss": 4.6942,
      "mean_token_accuracy": 0.24440861493349075,
      "num_tokens": 115160390.0,
      "step": 51590
    },
    {
      "entropy": 5.220657014846802,
      "epoch": 5.7946313247599255,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00022265634443380743,
      "loss": 4.7364,
      "mean_token_accuracy": 0.24712844491004943,
      "num_tokens": 115171686.0,
      "step": 51595
    },
    {
      "entropy": 5.369271564483642,
      "epoch": 5.795192901667884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022261729809729152,
      "loss": 4.9346,
      "mean_token_accuracy": 0.23092354089021683,
      "num_tokens": 115184285.0,
      "step": 51600
    },
    {
      "entropy": 5.3955573558807375,
      "epoch": 5.795754478575841,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00022257825342866116,
      "loss": 4.8218,
      "mean_token_accuracy": 0.2459690362215042,
      "num_tokens": 115195258.0,
      "step": 51605
    },
    {
      "entropy": 5.388172435760498,
      "epoch": 5.796316055483798,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002225392104291596,
      "loss": 4.8543,
      "mean_token_accuracy": 0.2352003797888756,
      "num_tokens": 115206155.0,
      "step": 51610
    },
    {
      "entropy": 5.248282527923584,
      "epoch": 5.796877632391756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002225001691000299,
      "loss": 4.6949,
      "mean_token_accuracy": 0.25024876147508623,
      "num_tokens": 115215545.0,
      "step": 51615
    },
    {
      "entropy": 5.306143712997437,
      "epoch": 5.797439209299713,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002224611294425151,
      "loss": 4.8467,
      "mean_token_accuracy": 0.2379753917455673,
      "num_tokens": 115226586.0,
      "step": 51620
    },
    {
      "entropy": 5.423135709762573,
      "epoch": 5.7980007862076715,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002224220914578584,
      "loss": 4.8695,
      "mean_token_accuracy": 0.23780225664377214,
      "num_tokens": 115237338.0,
      "step": 51625
    },
    {
      "entropy": 5.358355283737183,
      "epoch": 5.798562363115629,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022238305514730267,
      "loss": 4.8153,
      "mean_token_accuracy": 0.24581610709428786,
      "num_tokens": 115247727.0,
      "step": 51630
    },
    {
      "entropy": 5.332873296737671,
      "epoch": 5.799123940023586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022234402051209086,
      "loss": 4.7845,
      "mean_token_accuracy": 0.2404308408498764,
      "num_tokens": 115258410.0,
      "step": 51635
    },
    {
      "entropy": 5.384312629699707,
      "epoch": 5.799685516931544,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00022230498755346583,
      "loss": 4.9032,
      "mean_token_accuracy": 0.23063233494758606,
      "num_tokens": 115270474.0,
      "step": 51640
    },
    {
      "entropy": 5.383729124069214,
      "epoch": 5.800247093839501,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022226595627267048,
      "loss": 4.8913,
      "mean_token_accuracy": 0.23396565169095992,
      "num_tokens": 115280603.0,
      "step": 51645
    },
    {
      "entropy": 5.282470846176148,
      "epoch": 5.800808670747459,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022222692667094756,
      "loss": 4.7175,
      "mean_token_accuracy": 0.24562754482030869,
      "num_tokens": 115291413.0,
      "step": 51650
    },
    {
      "entropy": 5.272460842132569,
      "epoch": 5.8013702476554165,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022218789874953976,
      "loss": 4.7366,
      "mean_token_accuracy": 0.24315459579229354,
      "num_tokens": 115302480.0,
      "step": 51655
    },
    {
      "entropy": 5.367280578613281,
      "epoch": 5.801931824563374,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022214887250968974,
      "loss": 4.8197,
      "mean_token_accuracy": 0.24274391531944275,
      "num_tokens": 115313159.0,
      "step": 51660
    },
    {
      "entropy": 5.248348999023437,
      "epoch": 5.802493401471332,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002221098479526401,
      "loss": 4.7483,
      "mean_token_accuracy": 0.2378095030784607,
      "num_tokens": 115323188.0,
      "step": 51665
    },
    {
      "entropy": 5.310281276702881,
      "epoch": 5.803054978379289,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022207082507963327,
      "loss": 4.7632,
      "mean_token_accuracy": 0.23651974499225617,
      "num_tokens": 115335865.0,
      "step": 51670
    },
    {
      "entropy": 5.278801345825196,
      "epoch": 5.803616555287246,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022203180389191208,
      "loss": 4.7185,
      "mean_token_accuracy": 0.24759314358234405,
      "num_tokens": 115347334.0,
      "step": 51675
    },
    {
      "entropy": 5.375073099136353,
      "epoch": 5.804178132195204,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00022199278439071873,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23553008139133452,
      "num_tokens": 115358658.0,
      "step": 51680
    },
    {
      "entropy": 5.223998165130615,
      "epoch": 5.804739709103162,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022195376657729578,
      "loss": 4.6414,
      "mean_token_accuracy": 0.25045770555734637,
      "num_tokens": 115368299.0,
      "step": 51685
    },
    {
      "entropy": 5.238648414611816,
      "epoch": 5.805301286011119,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00022191475045288533,
      "loss": 4.7184,
      "mean_token_accuracy": 0.24589795619249344,
      "num_tokens": 115379258.0,
      "step": 51690
    },
    {
      "entropy": 5.3334434032440186,
      "epoch": 5.805862862919077,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00022187573601872994,
      "loss": 4.8873,
      "mean_token_accuracy": 0.23369628190994263,
      "num_tokens": 115391103.0,
      "step": 51695
    },
    {
      "entropy": 5.341215324401856,
      "epoch": 5.806424439827034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00022183672327607168,
      "loss": 4.7833,
      "mean_token_accuracy": 0.24056948572397233,
      "num_tokens": 115401502.0,
      "step": 51700
    },
    {
      "entropy": 5.316876697540283,
      "epoch": 5.806986016734992,
      "grad_norm": 0.953125,
      "learning_rate": 0.00022179771222615268,
      "loss": 4.8492,
      "mean_token_accuracy": 0.23198105692863463,
      "num_tokens": 115412918.0,
      "step": 51705
    },
    {
      "entropy": 5.367710447311401,
      "epoch": 5.8075475936429495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022175870287021528,
      "loss": 4.8416,
      "mean_token_accuracy": 0.24036489576101303,
      "num_tokens": 115425497.0,
      "step": 51710
    },
    {
      "entropy": 5.254408740997315,
      "epoch": 5.808109170550907,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00022171969520950143,
      "loss": 4.7579,
      "mean_token_accuracy": 0.24395632445812226,
      "num_tokens": 115436426.0,
      "step": 51715
    },
    {
      "entropy": 5.3099676132202145,
      "epoch": 5.808670747458865,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022168068924525308,
      "loss": 4.7174,
      "mean_token_accuracy": 0.2398879274725914,
      "num_tokens": 115446876.0,
      "step": 51720
    },
    {
      "entropy": 5.311697006225586,
      "epoch": 5.809232324366822,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00022164168497871227,
      "loss": 4.7647,
      "mean_token_accuracy": 0.24503686726093293,
      "num_tokens": 115457319.0,
      "step": 51725
    },
    {
      "entropy": 5.26786241531372,
      "epoch": 5.809793901274779,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022160268241112096,
      "loss": 4.7269,
      "mean_token_accuracy": 0.24327296763658524,
      "num_tokens": 115469605.0,
      "step": 51730
    },
    {
      "entropy": 5.236423921585083,
      "epoch": 5.810355478182737,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002215636815437209,
      "loss": 4.6793,
      "mean_token_accuracy": 0.24247221350669862,
      "num_tokens": 115480182.0,
      "step": 51735
    },
    {
      "entropy": 5.272430849075318,
      "epoch": 5.810917055090695,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022152468237775387,
      "loss": 4.7413,
      "mean_token_accuracy": 0.24355512261390685,
      "num_tokens": 115490746.0,
      "step": 51740
    },
    {
      "entropy": 5.300336837768555,
      "epoch": 5.811478631998652,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00022148568491446181,
      "loss": 4.7732,
      "mean_token_accuracy": 0.2334764838218689,
      "num_tokens": 115501798.0,
      "step": 51745
    },
    {
      "entropy": 5.31250376701355,
      "epoch": 5.81204020890661,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00022144668915508625,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2364277720451355,
      "num_tokens": 115512855.0,
      "step": 51750
    },
    {
      "entropy": 5.377389907836914,
      "epoch": 5.812601785814567,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022140769510086894,
      "loss": 4.8303,
      "mean_token_accuracy": 0.2327611967921257,
      "num_tokens": 115523983.0,
      "step": 51755
    },
    {
      "entropy": 5.329485845565796,
      "epoch": 5.813163362722525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002213687027530514,
      "loss": 4.8144,
      "mean_token_accuracy": 0.2399829849600792,
      "num_tokens": 115534947.0,
      "step": 51760
    },
    {
      "entropy": 5.236224269866943,
      "epoch": 5.813724939630482,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002213297121128751,
      "loss": 4.726,
      "mean_token_accuracy": 0.23789919018745423,
      "num_tokens": 115546687.0,
      "step": 51765
    },
    {
      "entropy": 5.3435369491577145,
      "epoch": 5.81428651653844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002212907231815816,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2403624624013901,
      "num_tokens": 115558462.0,
      "step": 51770
    },
    {
      "entropy": 5.355800294876099,
      "epoch": 5.814848093446398,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00022125173596041232,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2339407354593277,
      "num_tokens": 115568445.0,
      "step": 51775
    },
    {
      "entropy": 5.393836832046508,
      "epoch": 5.815409670354355,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00022121275045060857,
      "loss": 4.789,
      "mean_token_accuracy": 0.2412220522761345,
      "num_tokens": 115578435.0,
      "step": 51780
    },
    {
      "entropy": 5.3118504047393795,
      "epoch": 5.815971247262313,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022117376665341178,
      "loss": 4.7741,
      "mean_token_accuracy": 0.2413351207971573,
      "num_tokens": 115589393.0,
      "step": 51785
    },
    {
      "entropy": 5.38281021118164,
      "epoch": 5.81653282417027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022113478457006308,
      "loss": 4.9036,
      "mean_token_accuracy": 0.23469117432832717,
      "num_tokens": 115601458.0,
      "step": 51790
    },
    {
      "entropy": 5.217300224304199,
      "epoch": 5.8170944010782275,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022109580420180374,
      "loss": 4.6144,
      "mean_token_accuracy": 0.2551250696182251,
      "num_tokens": 115611784.0,
      "step": 51795
    },
    {
      "entropy": 5.225303792953492,
      "epoch": 5.817655977986185,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022105682554987493,
      "loss": 4.7148,
      "mean_token_accuracy": 0.2429789885878563,
      "num_tokens": 115623717.0,
      "step": 51800
    },
    {
      "entropy": 5.330233764648438,
      "epoch": 5.818217554894143,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002210178486155177,
      "loss": 4.8045,
      "mean_token_accuracy": 0.23886442184448242,
      "num_tokens": 115634834.0,
      "step": 51805
    },
    {
      "entropy": 5.330279016494751,
      "epoch": 5.8187791318021,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022097887339997312,
      "loss": 4.7724,
      "mean_token_accuracy": 0.24004240781068803,
      "num_tokens": 115646175.0,
      "step": 51810
    },
    {
      "entropy": 5.3096692085266115,
      "epoch": 5.819340708710058,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022093989990448203,
      "loss": 4.7201,
      "mean_token_accuracy": 0.2427333563566208,
      "num_tokens": 115657398.0,
      "step": 51815
    },
    {
      "entropy": 5.334406566619873,
      "epoch": 5.819902285618015,
      "grad_norm": 0.921875,
      "learning_rate": 0.00022090092813028563,
      "loss": 4.7862,
      "mean_token_accuracy": 0.23620370626449586,
      "num_tokens": 115670273.0,
      "step": 51820
    },
    {
      "entropy": 5.300399112701416,
      "epoch": 5.820463862525973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022086195807862464,
      "loss": 4.7827,
      "mean_token_accuracy": 0.24166414141654968,
      "num_tokens": 115681771.0,
      "step": 51825
    },
    {
      "entropy": 5.261226463317871,
      "epoch": 5.821025439433931,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022082298975073988,
      "loss": 4.7234,
      "mean_token_accuracy": 0.243558931350708,
      "num_tokens": 115693405.0,
      "step": 51830
    },
    {
      "entropy": 5.33147087097168,
      "epoch": 5.821587016341888,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002207840231478722,
      "loss": 4.8365,
      "mean_token_accuracy": 0.23550816327333451,
      "num_tokens": 115703683.0,
      "step": 51835
    },
    {
      "entropy": 5.290785312652588,
      "epoch": 5.822148593249846,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022074505827126222,
      "loss": 4.6972,
      "mean_token_accuracy": 0.24416192024946212,
      "num_tokens": 115714815.0,
      "step": 51840
    },
    {
      "entropy": 5.304741716384887,
      "epoch": 5.822710170157803,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00022070609512215062,
      "loss": 4.8412,
      "mean_token_accuracy": 0.23058947026729584,
      "num_tokens": 115725068.0,
      "step": 51845
    },
    {
      "entropy": 5.303268241882324,
      "epoch": 5.8232717470657605,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000220667133701778,
      "loss": 4.8021,
      "mean_token_accuracy": 0.2359074205160141,
      "num_tokens": 115736090.0,
      "step": 51850
    },
    {
      "entropy": 5.335567951202393,
      "epoch": 5.823833323973719,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022062817401138492,
      "loss": 4.7866,
      "mean_token_accuracy": 0.24162097424268722,
      "num_tokens": 115747727.0,
      "step": 51855
    },
    {
      "entropy": 5.22287769317627,
      "epoch": 5.824394900881676,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00022058921605221189,
      "loss": 4.7005,
      "mean_token_accuracy": 0.2438298225402832,
      "num_tokens": 115758970.0,
      "step": 51860
    },
    {
      "entropy": 5.290819072723389,
      "epoch": 5.824956477789633,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022055025982549924,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23729415088891984,
      "num_tokens": 115771503.0,
      "step": 51865
    },
    {
      "entropy": 5.28175950050354,
      "epoch": 5.825518054697591,
      "grad_norm": 1.109375,
      "learning_rate": 0.00022051130533248754,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23354617655277252,
      "num_tokens": 115782161.0,
      "step": 51870
    },
    {
      "entropy": 5.300750398635865,
      "epoch": 5.826079631605548,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000220472352574417,
      "loss": 4.7362,
      "mean_token_accuracy": 0.2411913365125656,
      "num_tokens": 115793368.0,
      "step": 51875
    },
    {
      "entropy": 5.329979801177979,
      "epoch": 5.8266412085135055,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00022043340155252788,
      "loss": 4.7873,
      "mean_token_accuracy": 0.24106606245040893,
      "num_tokens": 115804784.0,
      "step": 51880
    },
    {
      "entropy": 5.281415939331055,
      "epoch": 5.827202785421464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022039445226806033,
      "loss": 4.7314,
      "mean_token_accuracy": 0.24624460488557814,
      "num_tokens": 115814993.0,
      "step": 51885
    },
    {
      "entropy": 5.345640182495117,
      "epoch": 5.827764362329421,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022035550472225467,
      "loss": 4.868,
      "mean_token_accuracy": 0.23753280192613602,
      "num_tokens": 115826819.0,
      "step": 51890
    },
    {
      "entropy": 5.387383985519409,
      "epoch": 5.828325939237379,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022031655891635094,
      "loss": 4.8842,
      "mean_token_accuracy": 0.23434273451566695,
      "num_tokens": 115837948.0,
      "step": 51895
    },
    {
      "entropy": 5.455547571182251,
      "epoch": 5.828887516145336,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002202776148515892,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2226386159658432,
      "num_tokens": 115849404.0,
      "step": 51900
    },
    {
      "entropy": 5.363751935958862,
      "epoch": 5.829449093053293,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022023867252920942,
      "loss": 4.7737,
      "mean_token_accuracy": 0.24247547835111619,
      "num_tokens": 115860885.0,
      "step": 51905
    },
    {
      "entropy": 5.318176174163819,
      "epoch": 5.8300106699612515,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022019973195045155,
      "loss": 4.7859,
      "mean_token_accuracy": 0.24244073778390884,
      "num_tokens": 115872858.0,
      "step": 51910
    },
    {
      "entropy": 5.354626798629761,
      "epoch": 5.830572246869209,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022016079311655546,
      "loss": 4.8442,
      "mean_token_accuracy": 0.24042485505342484,
      "num_tokens": 115884945.0,
      "step": 51915
    },
    {
      "entropy": 5.281793785095215,
      "epoch": 5.831133823777166,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022012185602876094,
      "loss": 4.7502,
      "mean_token_accuracy": 0.23937497437000274,
      "num_tokens": 115896271.0,
      "step": 51920
    },
    {
      "entropy": 5.272312307357788,
      "epoch": 5.831695400685124,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022008292068830787,
      "loss": 4.7912,
      "mean_token_accuracy": 0.23827672898769378,
      "num_tokens": 115907355.0,
      "step": 51925
    },
    {
      "entropy": 5.3403085231781,
      "epoch": 5.832256977593081,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022004398709643587,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23658114820718765,
      "num_tokens": 115918700.0,
      "step": 51930
    },
    {
      "entropy": 5.280960559844971,
      "epoch": 5.8328185545010385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022000505525438462,
      "loss": 4.7054,
      "mean_token_accuracy": 0.24480810016393661,
      "num_tokens": 115930216.0,
      "step": 51935
    },
    {
      "entropy": 5.263738822937012,
      "epoch": 5.833380131408997,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021996612516339382,
      "loss": 4.7527,
      "mean_token_accuracy": 0.23691657930612564,
      "num_tokens": 115941740.0,
      "step": 51940
    },
    {
      "entropy": 5.26071982383728,
      "epoch": 5.833941708316954,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002199271968247029,
      "loss": 4.8061,
      "mean_token_accuracy": 0.24480031430721283,
      "num_tokens": 115952462.0,
      "step": 51945
    },
    {
      "entropy": 5.304532289505005,
      "epoch": 5.834503285224912,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021988827023955145,
      "loss": 4.7821,
      "mean_token_accuracy": 0.24809825867414476,
      "num_tokens": 115964133.0,
      "step": 51950
    },
    {
      "entropy": 5.282430410385132,
      "epoch": 5.835064862132869,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021984934540917873,
      "loss": 4.7722,
      "mean_token_accuracy": 0.23781523257493972,
      "num_tokens": 115975614.0,
      "step": 51955
    },
    {
      "entropy": 5.324610710144043,
      "epoch": 5.835626439040826,
      "grad_norm": 0.90625,
      "learning_rate": 0.00021981042233482423,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23029084950685502,
      "num_tokens": 115987653.0,
      "step": 51960
    },
    {
      "entropy": 5.246792459487915,
      "epoch": 5.8361880159487844,
      "grad_norm": 1.234375,
      "learning_rate": 0.00021977150101772737,
      "loss": 4.6558,
      "mean_token_accuracy": 0.2535387337207794,
      "num_tokens": 115999401.0,
      "step": 51965
    },
    {
      "entropy": 5.313159990310669,
      "epoch": 5.836749592856742,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021973258145912728,
      "loss": 4.8207,
      "mean_token_accuracy": 0.24127539545297622,
      "num_tokens": 116011187.0,
      "step": 51970
    },
    {
      "entropy": 5.355815315246582,
      "epoch": 5.8373111697647,
      "grad_norm": 1.171875,
      "learning_rate": 0.00021969366366026327,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22961999326944352,
      "num_tokens": 116022387.0,
      "step": 51975
    },
    {
      "entropy": 5.3730579853057865,
      "epoch": 5.837872746672657,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021965474762237447,
      "loss": 4.8724,
      "mean_token_accuracy": 0.24265945106744766,
      "num_tokens": 116034002.0,
      "step": 51980
    },
    {
      "entropy": 5.350648546218872,
      "epoch": 5.838434323580614,
      "grad_norm": 1.0,
      "learning_rate": 0.0002196158333467,
      "loss": 4.8844,
      "mean_token_accuracy": 0.23468669950962068,
      "num_tokens": 116046380.0,
      "step": 51985
    },
    {
      "entropy": 5.213736057281494,
      "epoch": 5.838995900488571,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021957692083447884,
      "loss": 4.6987,
      "mean_token_accuracy": 0.24591419100761414,
      "num_tokens": 116057898.0,
      "step": 51990
    },
    {
      "entropy": 5.319465494155883,
      "epoch": 5.8395574773965295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021953801008694996,
      "loss": 4.7524,
      "mean_token_accuracy": 0.23952192813158035,
      "num_tokens": 116069553.0,
      "step": 51995
    },
    {
      "entropy": 5.2992555618286135,
      "epoch": 5.840119054304487,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00021949910110535242,
      "loss": 4.713,
      "mean_token_accuracy": 0.2466202825307846,
      "num_tokens": 116079782.0,
      "step": 52000
    },
    {
      "entropy": 5.27800931930542,
      "epoch": 5.840680631212445,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00021946019389092503,
      "loss": 4.7745,
      "mean_token_accuracy": 0.2390301540493965,
      "num_tokens": 116091608.0,
      "step": 52005
    },
    {
      "entropy": 5.316182613372803,
      "epoch": 5.841242208120402,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002194212884449066,
      "loss": 4.8066,
      "mean_token_accuracy": 0.23927389979362487,
      "num_tokens": 116102118.0,
      "step": 52010
    },
    {
      "entropy": 5.347042798995972,
      "epoch": 5.841803785028359,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002193823847685359,
      "loss": 4.8634,
      "mean_token_accuracy": 0.23595324009656907,
      "num_tokens": 116113490.0,
      "step": 52015
    },
    {
      "entropy": 5.282469320297241,
      "epoch": 5.842365361936317,
      "grad_norm": 1.0625,
      "learning_rate": 0.00021934348286305167,
      "loss": 4.7906,
      "mean_token_accuracy": 0.24581867307424546,
      "num_tokens": 116125082.0,
      "step": 52020
    },
    {
      "entropy": 5.278296756744385,
      "epoch": 5.842926938844275,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00021930458272969243,
      "loss": 4.7757,
      "mean_token_accuracy": 0.23865014165639878,
      "num_tokens": 116135217.0,
      "step": 52025
    },
    {
      "entropy": 5.293130588531494,
      "epoch": 5.843488515752233,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021926568436969685,
      "loss": 4.7384,
      "mean_token_accuracy": 0.23734684884548188,
      "num_tokens": 116146565.0,
      "step": 52030
    },
    {
      "entropy": 5.438349866867066,
      "epoch": 5.84405009266019,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002192267877843036,
      "loss": 4.9475,
      "mean_token_accuracy": 0.2344321444630623,
      "num_tokens": 116157805.0,
      "step": 52035
    },
    {
      "entropy": 5.413228321075439,
      "epoch": 5.844611669568147,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021918789297475106,
      "loss": 4.8828,
      "mean_token_accuracy": 0.23845165967941284,
      "num_tokens": 116170073.0,
      "step": 52040
    },
    {
      "entropy": 5.31778359413147,
      "epoch": 5.845173246476105,
      "grad_norm": 1.15625,
      "learning_rate": 0.00021914899994227767,
      "loss": 4.723,
      "mean_token_accuracy": 0.2449008882045746,
      "num_tokens": 116180102.0,
      "step": 52045
    },
    {
      "entropy": 5.324584054946899,
      "epoch": 5.8457348233840625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00021911010868812174,
      "loss": 4.804,
      "mean_token_accuracy": 0.23759617358446122,
      "num_tokens": 116190160.0,
      "step": 52050
    },
    {
      "entropy": 5.205485677719116,
      "epoch": 5.84629640029202,
      "grad_norm": 1.0,
      "learning_rate": 0.00021907121921352162,
      "loss": 4.667,
      "mean_token_accuracy": 0.24842192083597184,
      "num_tokens": 116201496.0,
      "step": 52055
    },
    {
      "entropy": 5.363065195083618,
      "epoch": 5.846857977199978,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021903233151971562,
      "loss": 4.8339,
      "mean_token_accuracy": 0.23541479110717772,
      "num_tokens": 116213265.0,
      "step": 52060
    },
    {
      "entropy": 5.279172420501709,
      "epoch": 5.847419554107935,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00021899344560794183,
      "loss": 4.7612,
      "mean_token_accuracy": 0.24774031341075897,
      "num_tokens": 116225034.0,
      "step": 52065
    },
    {
      "entropy": 5.314350938796997,
      "epoch": 5.847981131015892,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00021895456147943848,
      "loss": 4.7636,
      "mean_token_accuracy": 0.24711343348026277,
      "num_tokens": 116236358.0,
      "step": 52070
    },
    {
      "entropy": 5.294891357421875,
      "epoch": 5.84854270792385,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021891567913544363,
      "loss": 4.8259,
      "mean_token_accuracy": 0.23810407370328904,
      "num_tokens": 116248433.0,
      "step": 52075
    },
    {
      "entropy": 5.350520086288452,
      "epoch": 5.8491042848318076,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00021887679857719535,
      "loss": 4.7881,
      "mean_token_accuracy": 0.2427320033311844,
      "num_tokens": 116260324.0,
      "step": 52080
    },
    {
      "entropy": 5.350294780731201,
      "epoch": 5.849665861739766,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00021883791980593153,
      "loss": 4.8086,
      "mean_token_accuracy": 0.2520646572113037,
      "num_tokens": 116271712.0,
      "step": 52085
    },
    {
      "entropy": 5.354401159286499,
      "epoch": 5.850227438647723,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021879904282289008,
      "loss": 4.8432,
      "mean_token_accuracy": 0.23567491471767427,
      "num_tokens": 116282923.0,
      "step": 52090
    },
    {
      "entropy": 5.213904142379761,
      "epoch": 5.85078901555568,
      "grad_norm": 1.03125,
      "learning_rate": 0.000218760167629309,
      "loss": 4.7121,
      "mean_token_accuracy": 0.24323923885822296,
      "num_tokens": 116294064.0,
      "step": 52095
    },
    {
      "entropy": 5.254127311706543,
      "epoch": 5.851350592463638,
      "grad_norm": 1.125,
      "learning_rate": 0.0002187212942264259,
      "loss": 4.7065,
      "mean_token_accuracy": 0.24513959735631943,
      "num_tokens": 116305861.0,
      "step": 52100
    },
    {
      "entropy": 5.268340396881103,
      "epoch": 5.851912169371595,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002186824226154786,
      "loss": 4.8285,
      "mean_token_accuracy": 0.23865508735179902,
      "num_tokens": 116317255.0,
      "step": 52105
    },
    {
      "entropy": 5.3726695537567135,
      "epoch": 5.852473746279553,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00021864355279770494,
      "loss": 4.8984,
      "mean_token_accuracy": 0.23277236223220826,
      "num_tokens": 116328506.0,
      "step": 52110
    },
    {
      "entropy": 5.278265047073364,
      "epoch": 5.853035323187511,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002186046847743423,
      "loss": 4.7242,
      "mean_token_accuracy": 0.24662187844514846,
      "num_tokens": 116339456.0,
      "step": 52115
    },
    {
      "entropy": 5.238146543502808,
      "epoch": 5.853596900095468,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021856581854662845,
      "loss": 4.7687,
      "mean_token_accuracy": 0.2484685882925987,
      "num_tokens": 116349915.0,
      "step": 52120
    },
    {
      "entropy": 5.305943298339844,
      "epoch": 5.854158477003425,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00021852695411580083,
      "loss": 4.8327,
      "mean_token_accuracy": 0.23384024053812028,
      "num_tokens": 116362362.0,
      "step": 52125
    },
    {
      "entropy": 5.255475950241089,
      "epoch": 5.854720053911383,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021848809148309684,
      "loss": 4.6974,
      "mean_token_accuracy": 0.24392309188842773,
      "num_tokens": 116373955.0,
      "step": 52130
    },
    {
      "entropy": 5.339780998229981,
      "epoch": 5.8552816308193405,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00021844923064975397,
      "loss": 4.7505,
      "mean_token_accuracy": 0.24256467670202256,
      "num_tokens": 116384528.0,
      "step": 52135
    },
    {
      "entropy": 5.313487434387207,
      "epoch": 5.855843207727299,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00021841037161700943,
      "loss": 4.7688,
      "mean_token_accuracy": 0.24961293041706084,
      "num_tokens": 116395600.0,
      "step": 52140
    },
    {
      "entropy": 5.319756507873535,
      "epoch": 5.856404784635256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021837151438610074,
      "loss": 4.7059,
      "mean_token_accuracy": 0.24923174530267717,
      "num_tokens": 116406427.0,
      "step": 52145
    },
    {
      "entropy": 5.315813732147217,
      "epoch": 5.856966361543213,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00021833265895826492,
      "loss": 4.7855,
      "mean_token_accuracy": 0.24354251325130463,
      "num_tokens": 116417707.0,
      "step": 52150
    },
    {
      "entropy": 5.327625799179077,
      "epoch": 5.857527938451171,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021829380533473924,
      "loss": 4.756,
      "mean_token_accuracy": 0.2469869777560234,
      "num_tokens": 116428311.0,
      "step": 52155
    },
    {
      "entropy": 5.288520002365113,
      "epoch": 5.858089515359128,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021825495351676083,
      "loss": 4.782,
      "mean_token_accuracy": 0.24147259145975114,
      "num_tokens": 116438579.0,
      "step": 52160
    },
    {
      "entropy": 5.25305871963501,
      "epoch": 5.8586510922670865,
      "grad_norm": 1.0625,
      "learning_rate": 0.00021821610350556664,
      "loss": 4.7322,
      "mean_token_accuracy": 0.2450373038649559,
      "num_tokens": 116449360.0,
      "step": 52165
    },
    {
      "entropy": 5.351304626464843,
      "epoch": 5.859212669175044,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002181772553023938,
      "loss": 4.9107,
      "mean_token_accuracy": 0.22663556337356566,
      "num_tokens": 116459583.0,
      "step": 52170
    },
    {
      "entropy": 5.308622932434082,
      "epoch": 5.859774246083001,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021813840890847904,
      "loss": 4.7782,
      "mean_token_accuracy": 0.2513938143849373,
      "num_tokens": 116470771.0,
      "step": 52175
    },
    {
      "entropy": 5.263450574874878,
      "epoch": 5.860335822990958,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002180995643250595,
      "loss": 4.7352,
      "mean_token_accuracy": 0.24555165320634842,
      "num_tokens": 116482900.0,
      "step": 52180
    },
    {
      "entropy": 5.340573453903199,
      "epoch": 5.860897399898916,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002180607215533719,
      "loss": 4.8098,
      "mean_token_accuracy": 0.24643135219812393,
      "num_tokens": 116493499.0,
      "step": 52185
    },
    {
      "entropy": 5.319741725921631,
      "epoch": 5.8614589768068734,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021802188059465304,
      "loss": 4.7964,
      "mean_token_accuracy": 0.24311316907405853,
      "num_tokens": 116504437.0,
      "step": 52190
    },
    {
      "entropy": 5.264760208129883,
      "epoch": 5.8620205537148315,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002179830414501396,
      "loss": 4.6867,
      "mean_token_accuracy": 0.24633391350507736,
      "num_tokens": 116514142.0,
      "step": 52195
    },
    {
      "entropy": 5.3389074325561525,
      "epoch": 5.862582130622789,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021794420412106825,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2331741690635681,
      "num_tokens": 116525408.0,
      "step": 52200
    },
    {
      "entropy": 5.371243286132812,
      "epoch": 5.863143707530746,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00021790536860867555,
      "loss": 4.8399,
      "mean_token_accuracy": 0.24093413203954697,
      "num_tokens": 116536920.0,
      "step": 52205
    },
    {
      "entropy": 5.302493715286255,
      "epoch": 5.863705284438704,
      "grad_norm": 1.078125,
      "learning_rate": 0.000217866534914198,
      "loss": 4.7923,
      "mean_token_accuracy": 0.2368660166859627,
      "num_tokens": 116547437.0,
      "step": 52210
    },
    {
      "entropy": 5.2621105194091795,
      "epoch": 5.864266861346661,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00021782770303887218,
      "loss": 4.726,
      "mean_token_accuracy": 0.24794248938560487,
      "num_tokens": 116557718.0,
      "step": 52215
    },
    {
      "entropy": 5.260115718841552,
      "epoch": 5.864828438254619,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00021778887298393447,
      "loss": 4.7781,
      "mean_token_accuracy": 0.23761823773384094,
      "num_tokens": 116569486.0,
      "step": 52220
    },
    {
      "entropy": 5.36255407333374,
      "epoch": 5.865390015162577,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002177500447506212,
      "loss": 4.9232,
      "mean_token_accuracy": 0.23078247755765915,
      "num_tokens": 116580617.0,
      "step": 52225
    },
    {
      "entropy": 5.391385078430176,
      "epoch": 5.865951592070534,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021771121834016877,
      "loss": 4.9655,
      "mean_token_accuracy": 0.23164264261722564,
      "num_tokens": 116590815.0,
      "step": 52230
    },
    {
      "entropy": 5.36133131980896,
      "epoch": 5.866513168978492,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00021767239375381338,
      "loss": 4.8356,
      "mean_token_accuracy": 0.2349330872297287,
      "num_tokens": 116601755.0,
      "step": 52235
    },
    {
      "entropy": 5.3380897521972654,
      "epoch": 5.867074745886449,
      "grad_norm": 1.140625,
      "learning_rate": 0.00021763357099279114,
      "loss": 4.7835,
      "mean_token_accuracy": 0.2450447142124176,
      "num_tokens": 116611895.0,
      "step": 52240
    },
    {
      "entropy": 5.312600564956665,
      "epoch": 5.867636322794406,
      "grad_norm": 1.109375,
      "learning_rate": 0.00021759475005833818,
      "loss": 4.8027,
      "mean_token_accuracy": 0.23749465495347977,
      "num_tokens": 116623054.0,
      "step": 52245
    },
    {
      "entropy": 5.41942024230957,
      "epoch": 5.8681978997023645,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021755593095169074,
      "loss": 4.8852,
      "mean_token_accuracy": 0.23144979923963546,
      "num_tokens": 116634700.0,
      "step": 52250
    },
    {
      "entropy": 5.393616104125977,
      "epoch": 5.868759476610322,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002175171136740847,
      "loss": 4.8609,
      "mean_token_accuracy": 0.236311237514019,
      "num_tokens": 116645907.0,
      "step": 52255
    },
    {
      "entropy": 5.353059196472168,
      "epoch": 5.869321053518279,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021747829822675612,
      "loss": 4.7994,
      "mean_token_accuracy": 0.24315326362848283,
      "num_tokens": 116656584.0,
      "step": 52260
    },
    {
      "entropy": 5.364986085891724,
      "epoch": 5.869882630426237,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021743948461094083,
      "loss": 4.8435,
      "mean_token_accuracy": 0.23231004625558854,
      "num_tokens": 116667934.0,
      "step": 52265
    },
    {
      "entropy": 5.334745597839356,
      "epoch": 5.870444207334194,
      "grad_norm": 1.125,
      "learning_rate": 0.00021740067282787463,
      "loss": 4.8038,
      "mean_token_accuracy": 0.242742058634758,
      "num_tokens": 116678773.0,
      "step": 52270
    },
    {
      "entropy": 5.361050701141357,
      "epoch": 5.871005784242152,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00021736186287879338,
      "loss": 4.8652,
      "mean_token_accuracy": 0.23464329093694686,
      "num_tokens": 116690376.0,
      "step": 52275
    },
    {
      "entropy": 5.381990718841553,
      "epoch": 5.87156736115011,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002173230547649327,
      "loss": 4.8324,
      "mean_token_accuracy": 0.24207522720098495,
      "num_tokens": 116700831.0,
      "step": 52280
    },
    {
      "entropy": 5.384974098205566,
      "epoch": 5.872128938058067,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00021728424848752842,
      "loss": 4.871,
      "mean_token_accuracy": 0.24117059260606766,
      "num_tokens": 116712675.0,
      "step": 52285
    },
    {
      "entropy": 5.304960775375366,
      "epoch": 5.872690514966025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021724544404781598,
      "loss": 4.8344,
      "mean_token_accuracy": 0.23482557535171508,
      "num_tokens": 116724821.0,
      "step": 52290
    },
    {
      "entropy": 5.293739318847656,
      "epoch": 5.873252091873982,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021720664144703105,
      "loss": 4.7962,
      "mean_token_accuracy": 0.23864217847585678,
      "num_tokens": 116736216.0,
      "step": 52295
    },
    {
      "entropy": 5.3224530696868895,
      "epoch": 5.873813668781939,
      "grad_norm": 1.140625,
      "learning_rate": 0.00021716784068640904,
      "loss": 4.8128,
      "mean_token_accuracy": 0.24054273962974548,
      "num_tokens": 116747827.0,
      "step": 52300
    },
    {
      "entropy": 5.323281192779541,
      "epoch": 5.874375245689897,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021712904176718545,
      "loss": 4.7971,
      "mean_token_accuracy": 0.2366036906838417,
      "num_tokens": 116757849.0,
      "step": 52305
    },
    {
      "entropy": 5.3227475643157955,
      "epoch": 5.874936822597855,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021709024469059564,
      "loss": 4.7709,
      "mean_token_accuracy": 0.23898344188928605,
      "num_tokens": 116769341.0,
      "step": 52310
    },
    {
      "entropy": 5.3447352886199955,
      "epoch": 5.875498399505812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021705144945787487,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23225641697645188,
      "num_tokens": 116780655.0,
      "step": 52315
    },
    {
      "entropy": 5.288556528091431,
      "epoch": 5.87605997641377,
      "grad_norm": 1.0,
      "learning_rate": 0.00021701265607025828,
      "loss": 4.76,
      "mean_token_accuracy": 0.24420306384563445,
      "num_tokens": 116792130.0,
      "step": 52320
    },
    {
      "entropy": 5.278102397918701,
      "epoch": 5.876621553321727,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00021697386452898133,
      "loss": 4.8275,
      "mean_token_accuracy": 0.2405288577079773,
      "num_tokens": 116803902.0,
      "step": 52325
    },
    {
      "entropy": 5.2946936130523685,
      "epoch": 5.877183130229685,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00021693507483527908,
      "loss": 4.7879,
      "mean_token_accuracy": 0.23899690806865692,
      "num_tokens": 116815657.0,
      "step": 52330
    },
    {
      "entropy": 5.336142873764038,
      "epoch": 5.8777447071376425,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00021689628699038648,
      "loss": 4.8804,
      "mean_token_accuracy": 0.2349557563662529,
      "num_tokens": 116826969.0,
      "step": 52335
    },
    {
      "entropy": 5.314441108703614,
      "epoch": 5.8783062840456,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021685750099553874,
      "loss": 4.7059,
      "mean_token_accuracy": 0.24036201238632202,
      "num_tokens": 116837532.0,
      "step": 52340
    },
    {
      "entropy": 5.251613807678223,
      "epoch": 5.878867860953558,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021681871685197068,
      "loss": 4.6691,
      "mean_token_accuracy": 0.2471660017967224,
      "num_tokens": 116848248.0,
      "step": 52345
    },
    {
      "entropy": 5.319881296157837,
      "epoch": 5.879429437861515,
      "grad_norm": 1.125,
      "learning_rate": 0.00021677993456091727,
      "loss": 4.801,
      "mean_token_accuracy": 0.23996011465787886,
      "num_tokens": 116858352.0,
      "step": 52350
    },
    {
      "entropy": 5.290220880508423,
      "epoch": 5.879991014769473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00021674115412361324,
      "loss": 4.7978,
      "mean_token_accuracy": 0.24188912957906722,
      "num_tokens": 116869323.0,
      "step": 52355
    },
    {
      "entropy": 5.3869682312011715,
      "epoch": 5.88055259167743,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021670237554129347,
      "loss": 4.9371,
      "mean_token_accuracy": 0.236181902885437,
      "num_tokens": 116881254.0,
      "step": 52360
    },
    {
      "entropy": 5.46140341758728,
      "epoch": 5.881114168585388,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002166635988151927,
      "loss": 4.9588,
      "mean_token_accuracy": 0.22845247983932496,
      "num_tokens": 116893614.0,
      "step": 52365
    },
    {
      "entropy": 5.376405715942383,
      "epoch": 5.881675745493345,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002166248239465456,
      "loss": 4.8556,
      "mean_token_accuracy": 0.23973244726657866,
      "num_tokens": 116904404.0,
      "step": 52370
    },
    {
      "entropy": 5.414957189559937,
      "epoch": 5.882237322401303,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021658605093658674,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2324964538216591,
      "num_tokens": 116916349.0,
      "step": 52375
    },
    {
      "entropy": 5.327965879440308,
      "epoch": 5.88279889930926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021654727978655066,
      "loss": 4.7467,
      "mean_token_accuracy": 0.24598519653081893,
      "num_tokens": 116926414.0,
      "step": 52380
    },
    {
      "entropy": 5.307403135299682,
      "epoch": 5.883360476217218,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021650851049767185,
      "loss": 4.8121,
      "mean_token_accuracy": 0.2532970324158669,
      "num_tokens": 116937821.0,
      "step": 52385
    },
    {
      "entropy": 5.273918962478637,
      "epoch": 5.8839220531251755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021646974307118467,
      "loss": 4.812,
      "mean_token_accuracy": 0.23287705183029175,
      "num_tokens": 116949125.0,
      "step": 52390
    },
    {
      "entropy": 5.320056772232055,
      "epoch": 5.884483630033133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021643097750832365,
      "loss": 4.708,
      "mean_token_accuracy": 0.24672023206949234,
      "num_tokens": 116959994.0,
      "step": 52395
    },
    {
      "entropy": 5.37435245513916,
      "epoch": 5.885045206941091,
      "grad_norm": 0.984375,
      "learning_rate": 0.00021639221381032303,
      "loss": 4.7798,
      "mean_token_accuracy": 0.24144816100597383,
      "num_tokens": 116971248.0,
      "step": 52400
    },
    {
      "entropy": 5.300677299499512,
      "epoch": 5.885606783849048,
      "grad_norm": 0.953125,
      "learning_rate": 0.00021635345197841712,
      "loss": 4.7509,
      "mean_token_accuracy": 0.2427266389131546,
      "num_tokens": 116982013.0,
      "step": 52405
    },
    {
      "entropy": 5.3028887748718265,
      "epoch": 5.886168360757006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002163146920138399,
      "loss": 4.8486,
      "mean_token_accuracy": 0.23087591975927352,
      "num_tokens": 116993469.0,
      "step": 52410
    },
    {
      "entropy": 5.3081221103668215,
      "epoch": 5.886729937664963,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021627593391782574,
      "loss": 4.7649,
      "mean_token_accuracy": 0.23843368142843246,
      "num_tokens": 117004171.0,
      "step": 52415
    },
    {
      "entropy": 5.293523454666138,
      "epoch": 5.8872915145729205,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021623717769160862,
      "loss": 4.7426,
      "mean_token_accuracy": 0.2504502832889557,
      "num_tokens": 117014894.0,
      "step": 52420
    },
    {
      "entropy": 5.3123941898345945,
      "epoch": 5.887853091480879,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021619842333642252,
      "loss": 4.79,
      "mean_token_accuracy": 0.2436143770813942,
      "num_tokens": 117026450.0,
      "step": 52425
    },
    {
      "entropy": 5.296234464645385,
      "epoch": 5.888414668388836,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00021615967085350147,
      "loss": 4.8238,
      "mean_token_accuracy": 0.23865356892347336,
      "num_tokens": 117039158.0,
      "step": 52430
    },
    {
      "entropy": 5.332911109924316,
      "epoch": 5.888976245296793,
      "grad_norm": 1.109375,
      "learning_rate": 0.00021612092024407925,
      "loss": 4.8293,
      "mean_token_accuracy": 0.24340758025646209,
      "num_tokens": 117048996.0,
      "step": 52435
    },
    {
      "entropy": 5.362594938278198,
      "epoch": 5.889537822204751,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00021608217150938987,
      "loss": 4.7997,
      "mean_token_accuracy": 0.2314944490790367,
      "num_tokens": 117060139.0,
      "step": 52440
    },
    {
      "entropy": 5.34909987449646,
      "epoch": 5.890099399112708,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002160434246506669,
      "loss": 4.7941,
      "mean_token_accuracy": 0.23605460524559022,
      "num_tokens": 117071766.0,
      "step": 52445
    },
    {
      "entropy": 5.346598100662232,
      "epoch": 5.890660976020666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002160046796691442,
      "loss": 4.8723,
      "mean_token_accuracy": 0.22950120866298676,
      "num_tokens": 117083459.0,
      "step": 52450
    },
    {
      "entropy": 5.2549303531646725,
      "epoch": 5.891222552928624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002159659365660554,
      "loss": 4.7509,
      "mean_token_accuracy": 0.2412627026438713,
      "num_tokens": 117096538.0,
      "step": 52455
    },
    {
      "entropy": 5.284158992767334,
      "epoch": 5.891784129836581,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000215927195342634,
      "loss": 4.6897,
      "mean_token_accuracy": 0.24352446049451829,
      "num_tokens": 117107787.0,
      "step": 52460
    },
    {
      "entropy": 5.403974819183349,
      "epoch": 5.892345706744539,
      "grad_norm": 1.140625,
      "learning_rate": 0.00021588845600011364,
      "loss": 4.8266,
      "mean_token_accuracy": 0.23382043838500977,
      "num_tokens": 117117734.0,
      "step": 52465
    },
    {
      "entropy": 5.313464975357055,
      "epoch": 5.892907283652496,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002158497185397278,
      "loss": 4.7779,
      "mean_token_accuracy": 0.2442946583032608,
      "num_tokens": 117129571.0,
      "step": 52470
    },
    {
      "entropy": 5.3234447002410885,
      "epoch": 5.8934688605604535,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021581098296270985,
      "loss": 4.7341,
      "mean_token_accuracy": 0.2465335786342621,
      "num_tokens": 117140696.0,
      "step": 52475
    },
    {
      "entropy": 5.318042325973511,
      "epoch": 5.894030437468412,
      "grad_norm": 1.109375,
      "learning_rate": 0.00021577224927029322,
      "loss": 4.7893,
      "mean_token_accuracy": 0.23674688190221788,
      "num_tokens": 117151491.0,
      "step": 52480
    },
    {
      "entropy": 5.355563735961914,
      "epoch": 5.894592014376369,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00021573351746371106,
      "loss": 4.8632,
      "mean_token_accuracy": 0.23320467621088029,
      "num_tokens": 117163515.0,
      "step": 52485
    },
    {
      "entropy": 5.402954483032227,
      "epoch": 5.895153591284326,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021569478754419674,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22959672957658767,
      "num_tokens": 117174936.0,
      "step": 52490
    },
    {
      "entropy": 5.39686861038208,
      "epoch": 5.895715168192284,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021565605951298335,
      "loss": 4.8799,
      "mean_token_accuracy": 0.23701776564121246,
      "num_tokens": 117185888.0,
      "step": 52495
    },
    {
      "entropy": 5.297519111633301,
      "epoch": 5.896276745100241,
      "grad_norm": 1.09375,
      "learning_rate": 0.000215617333371304,
      "loss": 4.6682,
      "mean_token_accuracy": 0.24988609701395034,
      "num_tokens": 117195762.0,
      "step": 52500
    },
    {
      "entropy": 5.375298023223877,
      "epoch": 5.896838322008199,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00021557860912039184,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23327751010656356,
      "num_tokens": 117206431.0,
      "step": 52505
    },
    {
      "entropy": 5.3422607421875,
      "epoch": 5.897399898916157,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00021553988676147984,
      "loss": 4.8126,
      "mean_token_accuracy": 0.2339986115694046,
      "num_tokens": 117217712.0,
      "step": 52510
    },
    {
      "entropy": 5.241910648345947,
      "epoch": 5.897961475824114,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002155011662958009,
      "loss": 4.6784,
      "mean_token_accuracy": 0.24025077521800994,
      "num_tokens": 117228103.0,
      "step": 52515
    },
    {
      "entropy": 5.222621440887451,
      "epoch": 5.898523052732072,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00021546244772458785,
      "loss": 4.66,
      "mean_token_accuracy": 0.2508206024765968,
      "num_tokens": 117238851.0,
      "step": 52520
    },
    {
      "entropy": 5.376227617263794,
      "epoch": 5.899084629640029,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021542373104907364,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23815090954303741,
      "num_tokens": 117250138.0,
      "step": 52525
    },
    {
      "entropy": 5.425632762908935,
      "epoch": 5.899646206547986,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00021538501627049085,
      "loss": 4.9811,
      "mean_token_accuracy": 0.22889527082443237,
      "num_tokens": 117263572.0,
      "step": 52530
    },
    {
      "entropy": 5.336065006256104,
      "epoch": 5.9002077834559445,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021534630339007223,
      "loss": 4.8559,
      "mean_token_accuracy": 0.23768599778413774,
      "num_tokens": 117273824.0,
      "step": 52535
    },
    {
      "entropy": 5.398649072647094,
      "epoch": 5.900769360363902,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002153075924090505,
      "loss": 4.8896,
      "mean_token_accuracy": 0.23595982044935226,
      "num_tokens": 117285099.0,
      "step": 52540
    },
    {
      "entropy": 5.38257622718811,
      "epoch": 5.90133093727186,
      "grad_norm": 0.953125,
      "learning_rate": 0.00021526888332865817,
      "loss": 4.8162,
      "mean_token_accuracy": 0.23543430119752884,
      "num_tokens": 117297007.0,
      "step": 52545
    },
    {
      "entropy": 5.355287551879883,
      "epoch": 5.901892514179817,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021523017615012786,
      "loss": 4.8612,
      "mean_token_accuracy": 0.2337565615773201,
      "num_tokens": 117308429.0,
      "step": 52550
    },
    {
      "entropy": 5.247079992294312,
      "epoch": 5.902454091087774,
      "grad_norm": 1.125,
      "learning_rate": 0.00021519147087469182,
      "loss": 4.6534,
      "mean_token_accuracy": 0.25455627888441085,
      "num_tokens": 117318643.0,
      "step": 52555
    },
    {
      "entropy": 5.322048091888428,
      "epoch": 5.9030156679957315,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021515276750358258,
      "loss": 4.881,
      "mean_token_accuracy": 0.23063185662031174,
      "num_tokens": 117331810.0,
      "step": 52560
    },
    {
      "entropy": 5.383693647384644,
      "epoch": 5.90357724490369,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021511406603803247,
      "loss": 4.808,
      "mean_token_accuracy": 0.23875497430562972,
      "num_tokens": 117342963.0,
      "step": 52565
    },
    {
      "entropy": 5.318366146087646,
      "epoch": 5.904138821811647,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00021507536647927361,
      "loss": 4.7631,
      "mean_token_accuracy": 0.24394513219594954,
      "num_tokens": 117352421.0,
      "step": 52570
    },
    {
      "entropy": 5.333449363708496,
      "epoch": 5.904700398719605,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002150366688285384,
      "loss": 4.8259,
      "mean_token_accuracy": 0.23856592625379563,
      "num_tokens": 117362947.0,
      "step": 52575
    },
    {
      "entropy": 5.362803411483765,
      "epoch": 5.905261975627562,
      "grad_norm": 1.0,
      "learning_rate": 0.00021499797308705893,
      "loss": 4.8913,
      "mean_token_accuracy": 0.23900776505470275,
      "num_tokens": 117375102.0,
      "step": 52580
    },
    {
      "entropy": 5.297739887237549,
      "epoch": 5.905823552535519,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00021495927925606722,
      "loss": 4.6975,
      "mean_token_accuracy": 0.24490772187709808,
      "num_tokens": 117385356.0,
      "step": 52585
    },
    {
      "entropy": 5.4017750263214115,
      "epoch": 5.9063851294434775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021492058733679542,
      "loss": 4.8577,
      "mean_token_accuracy": 0.23617078810930253,
      "num_tokens": 117396106.0,
      "step": 52590
    },
    {
      "entropy": 5.373436164855957,
      "epoch": 5.906946706351435,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00021488189733047535,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22724589109420776,
      "num_tokens": 117407176.0,
      "step": 52595
    },
    {
      "entropy": 5.363849067687989,
      "epoch": 5.907508283259393,
      "grad_norm": 1.0625,
      "learning_rate": 0.000214843209238339,
      "loss": 4.8167,
      "mean_token_accuracy": 0.23439643830060958,
      "num_tokens": 117418407.0,
      "step": 52600
    },
    {
      "entropy": 5.279997968673706,
      "epoch": 5.90806986016735,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00021480452306161813,
      "loss": 4.7818,
      "mean_token_accuracy": 0.24056798219680786,
      "num_tokens": 117431178.0,
      "step": 52605
    },
    {
      "entropy": 5.298519945144653,
      "epoch": 5.908631437075307,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021476583880154464,
      "loss": 4.7752,
      "mean_token_accuracy": 0.24063548892736436,
      "num_tokens": 117442806.0,
      "step": 52610
    },
    {
      "entropy": 5.382386302947998,
      "epoch": 5.909193013983265,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021472715645935026,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2243045151233673,
      "num_tokens": 117454949.0,
      "step": 52615
    },
    {
      "entropy": 5.351309967041016,
      "epoch": 5.909754590891223,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00021468847603626657,
      "loss": 4.787,
      "mean_token_accuracy": 0.24099792093038558,
      "num_tokens": 117467330.0,
      "step": 52620
    },
    {
      "entropy": 5.293655967712402,
      "epoch": 5.91031616779918,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002146497975335252,
      "loss": 4.7312,
      "mean_token_accuracy": 0.23782020956277847,
      "num_tokens": 117477555.0,
      "step": 52625
    },
    {
      "entropy": 5.350171041488648,
      "epoch": 5.910877744707138,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021461112095235764,
      "loss": 4.8889,
      "mean_token_accuracy": 0.23674318343400955,
      "num_tokens": 117487847.0,
      "step": 52630
    },
    {
      "entropy": 5.244684648513794,
      "epoch": 5.911439321615095,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021457244629399547,
      "loss": 4.8195,
      "mean_token_accuracy": 0.23943972438573838,
      "num_tokens": 117499347.0,
      "step": 52635
    },
    {
      "entropy": 5.331181478500366,
      "epoch": 5.912000898523052,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00021453377355967003,
      "loss": 4.7839,
      "mean_token_accuracy": 0.24056043177843095,
      "num_tokens": 117511201.0,
      "step": 52640
    },
    {
      "entropy": 5.369244050979614,
      "epoch": 5.91256247543101,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021449510275061264,
      "loss": 4.8104,
      "mean_token_accuracy": 0.24250543415546416,
      "num_tokens": 117522742.0,
      "step": 52645
    },
    {
      "entropy": 5.320102119445801,
      "epoch": 5.913124052338968,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021445643386805468,
      "loss": 4.7973,
      "mean_token_accuracy": 0.23455713540315629,
      "num_tokens": 117533609.0,
      "step": 52650
    },
    {
      "entropy": 5.3027314186096195,
      "epoch": 5.913685629246926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002144177669132274,
      "loss": 4.7897,
      "mean_token_accuracy": 0.24138564467430115,
      "num_tokens": 117544265.0,
      "step": 52655
    },
    {
      "entropy": 5.269420480728149,
      "epoch": 5.914247206154883,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021437910188736188,
      "loss": 4.8251,
      "mean_token_accuracy": 0.2385624796152115,
      "num_tokens": 117556088.0,
      "step": 52660
    },
    {
      "entropy": 5.380385637283325,
      "epoch": 5.91480878306284,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00021434043879168925,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23730861395597458,
      "num_tokens": 117566891.0,
      "step": 52665
    },
    {
      "entropy": 5.272003602981568,
      "epoch": 5.915370359970798,
      "grad_norm": 1.0,
      "learning_rate": 0.00021430177762744057,
      "loss": 4.722,
      "mean_token_accuracy": 0.24348686784505844,
      "num_tokens": 117577552.0,
      "step": 52670
    },
    {
      "entropy": 5.358322477340698,
      "epoch": 5.9159319368787555,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00021426311839584688,
      "loss": 4.8494,
      "mean_token_accuracy": 0.23722697347402572,
      "num_tokens": 117589436.0,
      "step": 52675
    },
    {
      "entropy": 5.326896333694458,
      "epoch": 5.916493513786713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021422446109813892,
      "loss": 4.8228,
      "mean_token_accuracy": 0.24012438356876373,
      "num_tokens": 117599728.0,
      "step": 52680
    },
    {
      "entropy": 5.276667499542237,
      "epoch": 5.917055090694671,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002141858057355478,
      "loss": 4.7109,
      "mean_token_accuracy": 0.24763632118701934,
      "num_tokens": 117610961.0,
      "step": 52685
    },
    {
      "entropy": 5.344430780410766,
      "epoch": 5.917616667602628,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002141471523093042,
      "loss": 4.837,
      "mean_token_accuracy": 0.2464147001504898,
      "num_tokens": 117622857.0,
      "step": 52690
    },
    {
      "entropy": 5.358357858657837,
      "epoch": 5.918178244510585,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021410850082063888,
      "loss": 4.8505,
      "mean_token_accuracy": 0.23661022186279296,
      "num_tokens": 117634533.0,
      "step": 52695
    },
    {
      "entropy": 5.319080257415772,
      "epoch": 5.918739821418543,
      "grad_norm": 1.125,
      "learning_rate": 0.0002140698512707825,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23772307187318803,
      "num_tokens": 117646136.0,
      "step": 52700
    },
    {
      "entropy": 5.238615465164185,
      "epoch": 5.919301398326501,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021403120366096572,
      "loss": 4.722,
      "mean_token_accuracy": 0.24417992681264877,
      "num_tokens": 117656900.0,
      "step": 52705
    },
    {
      "entropy": 5.2939342021942135,
      "epoch": 5.919862975234459,
      "grad_norm": 1.0,
      "learning_rate": 0.00021399255799241901,
      "loss": 4.8477,
      "mean_token_accuracy": 0.24088621884584427,
      "num_tokens": 117668689.0,
      "step": 52710
    },
    {
      "entropy": 5.3165863990783695,
      "epoch": 5.920424552142416,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021395391426637288,
      "loss": 4.7855,
      "mean_token_accuracy": 0.24513758420944215,
      "num_tokens": 117679926.0,
      "step": 52715
    },
    {
      "entropy": 5.282705640792846,
      "epoch": 5.920986129050373,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002139152724840579,
      "loss": 4.7558,
      "mean_token_accuracy": 0.24554292261600494,
      "num_tokens": 117691027.0,
      "step": 52720
    },
    {
      "entropy": 5.3088014125823975,
      "epoch": 5.921547705958331,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00021387663264670425,
      "loss": 4.7382,
      "mean_token_accuracy": 0.2531947508454323,
      "num_tokens": 117702098.0,
      "step": 52725
    },
    {
      "entropy": 5.30010404586792,
      "epoch": 5.9221092828662885,
      "grad_norm": 0.984375,
      "learning_rate": 0.00021383799475554238,
      "loss": 4.7951,
      "mean_token_accuracy": 0.23933669924736023,
      "num_tokens": 117712865.0,
      "step": 52730
    },
    {
      "entropy": 5.3108593940734865,
      "epoch": 5.922670859774247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002137993588118025,
      "loss": 4.8597,
      "mean_token_accuracy": 0.2429556280374527,
      "num_tokens": 117723750.0,
      "step": 52735
    },
    {
      "entropy": 5.2819146633148195,
      "epoch": 5.923232436682204,
      "grad_norm": 1.0625,
      "learning_rate": 0.00021376072481671477,
      "loss": 4.7054,
      "mean_token_accuracy": 0.24451127499341965,
      "num_tokens": 117734587.0,
      "step": 52740
    },
    {
      "entropy": 5.289522266387939,
      "epoch": 5.923794013590161,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0002137220927715093,
      "loss": 4.8047,
      "mean_token_accuracy": 0.23737290501594543,
      "num_tokens": 117747557.0,
      "step": 52745
    },
    {
      "entropy": 5.3407814502716064,
      "epoch": 5.924355590498118,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000213683462677416,
      "loss": 4.8207,
      "mean_token_accuracy": 0.23531724363565446,
      "num_tokens": 117758256.0,
      "step": 52750
    },
    {
      "entropy": 5.299745225906372,
      "epoch": 5.924917167406076,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021364483453566519,
      "loss": 4.7034,
      "mean_token_accuracy": 0.24976298063993455,
      "num_tokens": 117769976.0,
      "step": 52755
    },
    {
      "entropy": 5.349279880523682,
      "epoch": 5.9254787443140335,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021360620834748662,
      "loss": 4.8787,
      "mean_token_accuracy": 0.23477881401777267,
      "num_tokens": 117782284.0,
      "step": 52760
    },
    {
      "entropy": 5.311539220809936,
      "epoch": 5.926040321221992,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002135675841141102,
      "loss": 4.8379,
      "mean_token_accuracy": 0.23427960723638536,
      "num_tokens": 117794131.0,
      "step": 52765
    },
    {
      "entropy": 5.3468310832977295,
      "epoch": 5.926601898129949,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021352896183676574,
      "loss": 4.8929,
      "mean_token_accuracy": 0.23108877688646318,
      "num_tokens": 117804785.0,
      "step": 52770
    },
    {
      "entropy": 5.239675712585449,
      "epoch": 5.927163475037906,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021349034151668295,
      "loss": 4.6393,
      "mean_token_accuracy": 0.24674164354801179,
      "num_tokens": 117815691.0,
      "step": 52775
    },
    {
      "entropy": 5.432074499130249,
      "epoch": 5.927725051945864,
      "grad_norm": 1.109375,
      "learning_rate": 0.00021345172315509153,
      "loss": 4.9151,
      "mean_token_accuracy": 0.23812775015830995,
      "num_tokens": 117827044.0,
      "step": 52780
    },
    {
      "entropy": 5.353216552734375,
      "epoch": 5.928286628853821,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00021341310675322112,
      "loss": 4.7864,
      "mean_token_accuracy": 0.23355242013931274,
      "num_tokens": 117837702.0,
      "step": 52785
    },
    {
      "entropy": 5.334795665740967,
      "epoch": 5.9288482057617795,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00021337449231230128,
      "loss": 4.8628,
      "mean_token_accuracy": 0.23625075668096543,
      "num_tokens": 117848204.0,
      "step": 52790
    },
    {
      "entropy": 5.266196584701538,
      "epoch": 5.929409782669737,
      "grad_norm": 1.0,
      "learning_rate": 0.00021333587983356157,
      "loss": 4.7294,
      "mean_token_accuracy": 0.24152374118566514,
      "num_tokens": 117858517.0,
      "step": 52795
    },
    {
      "entropy": 5.34162015914917,
      "epoch": 5.929971359577694,
      "grad_norm": 1.109375,
      "learning_rate": 0.00021329726931823128,
      "loss": 4.8333,
      "mean_token_accuracy": 0.2348298668861389,
      "num_tokens": 117868874.0,
      "step": 52800
    },
    {
      "entropy": 5.334166860580444,
      "epoch": 5.930532936485652,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002132586607675399,
      "loss": 4.8209,
      "mean_token_accuracy": 0.24213824570178985,
      "num_tokens": 117879450.0,
      "step": 52805
    },
    {
      "entropy": 5.297774696350098,
      "epoch": 5.931094513393609,
      "grad_norm": 1.0625,
      "learning_rate": 0.00021322005418271674,
      "loss": 4.7599,
      "mean_token_accuracy": 0.24300872385501862,
      "num_tokens": 117890111.0,
      "step": 52810
    },
    {
      "entropy": 5.348769235610962,
      "epoch": 5.9316560903015665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021318144956499098,
      "loss": 4.8322,
      "mean_token_accuracy": 0.23776338696479798,
      "num_tokens": 117901836.0,
      "step": 52815
    },
    {
      "entropy": 5.2751970291137695,
      "epoch": 5.932217667209525,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021314284691559182,
      "loss": 4.676,
      "mean_token_accuracy": 0.24950366616249084,
      "num_tokens": 117911854.0,
      "step": 52820
    },
    {
      "entropy": 5.265320301055908,
      "epoch": 5.932779244117482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002131042462357483,
      "loss": 4.744,
      "mean_token_accuracy": 0.24879120886325837,
      "num_tokens": 117922967.0,
      "step": 52825
    },
    {
      "entropy": 5.31804141998291,
      "epoch": 5.933340821025439,
      "grad_norm": 1.109375,
      "learning_rate": 0.00021306564752668979,
      "loss": 4.8161,
      "mean_token_accuracy": 0.2378914475440979,
      "num_tokens": 117933991.0,
      "step": 52830
    },
    {
      "entropy": 5.353575944900513,
      "epoch": 5.933902397933397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00021302705078964497,
      "loss": 4.8476,
      "mean_token_accuracy": 0.23956012427806855,
      "num_tokens": 117945991.0,
      "step": 52835
    },
    {
      "entropy": 5.385207462310791,
      "epoch": 5.934463974841354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021298845602584293,
      "loss": 4.7949,
      "mean_token_accuracy": 0.24706200361251832,
      "num_tokens": 117956588.0,
      "step": 52840
    },
    {
      "entropy": 5.367385482788086,
      "epoch": 5.9350255517493125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002129498632365125,
      "loss": 4.8256,
      "mean_token_accuracy": 0.2423715576529503,
      "num_tokens": 117966425.0,
      "step": 52845
    },
    {
      "entropy": 5.402268600463867,
      "epoch": 5.93558712865727,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002129112724228825,
      "loss": 4.8023,
      "mean_token_accuracy": 0.2361139491200447,
      "num_tokens": 117976794.0,
      "step": 52850
    },
    {
      "entropy": 5.333890628814697,
      "epoch": 5.936148705565227,
      "grad_norm": 1.140625,
      "learning_rate": 0.00021287268358618168,
      "loss": 4.7871,
      "mean_token_accuracy": 0.24049043655395508,
      "num_tokens": 117988236.0,
      "step": 52855
    },
    {
      "entropy": 5.228446960449219,
      "epoch": 5.936710282473185,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021283409672763864,
      "loss": 4.6823,
      "mean_token_accuracy": 0.2531230911612511,
      "num_tokens": 117999699.0,
      "step": 52860
    },
    {
      "entropy": 5.287493324279785,
      "epoch": 5.937271859381142,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002127955118484821,
      "loss": 4.7659,
      "mean_token_accuracy": 0.24034328758716583,
      "num_tokens": 118008788.0,
      "step": 52865
    },
    {
      "entropy": 5.305104970932007,
      "epoch": 5.937833436289099,
      "grad_norm": 0.96875,
      "learning_rate": 0.00021275692894994058,
      "loss": 4.8218,
      "mean_token_accuracy": 0.23868513852357864,
      "num_tokens": 118019925.0,
      "step": 52870
    },
    {
      "entropy": 5.302529907226562,
      "epoch": 5.9383950131970575,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00021271834803324264,
      "loss": 4.6648,
      "mean_token_accuracy": 0.2500045567750931,
      "num_tokens": 118031403.0,
      "step": 52875
    },
    {
      "entropy": 5.2565765380859375,
      "epoch": 5.938956590105015,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002126797690996166,
      "loss": 4.7024,
      "mean_token_accuracy": 0.246938119828701,
      "num_tokens": 118044565.0,
      "step": 52880
    },
    {
      "entropy": 5.364192008972168,
      "epoch": 5.939518167012972,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00021264119215029092,
      "loss": 4.8273,
      "mean_token_accuracy": 0.23944331854581832,
      "num_tokens": 118055684.0,
      "step": 52885
    },
    {
      "entropy": 5.23955078125,
      "epoch": 5.94007974392093,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021260261718649388,
      "loss": 4.7102,
      "mean_token_accuracy": 0.24631700962781905,
      "num_tokens": 118066667.0,
      "step": 52890
    },
    {
      "entropy": 5.288236284255982,
      "epoch": 5.940641320828887,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021256404420945358,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23086397498846054,
      "num_tokens": 118077699.0,
      "step": 52895
    },
    {
      "entropy": 5.2241761684417725,
      "epoch": 5.941202897736845,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002125254732203984,
      "loss": 4.6662,
      "mean_token_accuracy": 0.2462477818131447,
      "num_tokens": 118088732.0,
      "step": 52900
    },
    {
      "entropy": 5.403057956695557,
      "epoch": 5.941764474644803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021248690422055643,
      "loss": 4.8975,
      "mean_token_accuracy": 0.23873988837003707,
      "num_tokens": 118100838.0,
      "step": 52905
    },
    {
      "entropy": 5.357292795181275,
      "epoch": 5.94232605155276,
      "grad_norm": 0.9375,
      "learning_rate": 0.00021244833721115563,
      "loss": 4.8199,
      "mean_token_accuracy": 0.24359432756900787,
      "num_tokens": 118112448.0,
      "step": 52910
    },
    {
      "entropy": 5.4224019050598145,
      "epoch": 5.942887628460718,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002124097721934241,
      "loss": 4.8978,
      "mean_token_accuracy": 0.23739907443523406,
      "num_tokens": 118124236.0,
      "step": 52915
    },
    {
      "entropy": 5.367053604125976,
      "epoch": 5.943449205368675,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00021237120916858967,
      "loss": 4.812,
      "mean_token_accuracy": 0.23633493930101396,
      "num_tokens": 118133871.0,
      "step": 52920
    },
    {
      "entropy": 5.274646186828614,
      "epoch": 5.944010782276633,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00021233264813788022,
      "loss": 4.7589,
      "mean_token_accuracy": 0.24510617107152938,
      "num_tokens": 118144543.0,
      "step": 52925
    },
    {
      "entropy": 5.313476037979126,
      "epoch": 5.9445723591845905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002122940891025235,
      "loss": 4.8217,
      "mean_token_accuracy": 0.24542275220155715,
      "num_tokens": 118155933.0,
      "step": 52930
    },
    {
      "entropy": 5.397573041915893,
      "epoch": 5.945133936092548,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00021225553206374737,
      "loss": 4.8673,
      "mean_token_accuracy": 0.24101074486970903,
      "num_tokens": 118166726.0,
      "step": 52935
    },
    {
      "entropy": 5.296338605880737,
      "epoch": 5.945695513000506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00021221697702277943,
      "loss": 4.7701,
      "mean_token_accuracy": 0.24418648481369018,
      "num_tokens": 118177412.0,
      "step": 52940
    },
    {
      "entropy": 5.306379127502441,
      "epoch": 5.946257089908463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002121784239808473,
      "loss": 4.7765,
      "mean_token_accuracy": 0.24094390869140625,
      "num_tokens": 118188445.0,
      "step": 52945
    },
    {
      "entropy": 5.319080448150634,
      "epoch": 5.94681866681642,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00021213987293917846,
      "loss": 4.8404,
      "mean_token_accuracy": 0.23668431341648102,
      "num_tokens": 118198897.0,
      "step": 52950
    },
    {
      "entropy": 5.322534608840942,
      "epoch": 5.947380243724378,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00021210132389900055,
      "loss": 4.8405,
      "mean_token_accuracy": 0.24147406667470933,
      "num_tokens": 118209815.0,
      "step": 52955
    },
    {
      "entropy": 5.380500173568725,
      "epoch": 5.947941820632336,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00021206277686154078,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23685044497251512,
      "num_tokens": 118221336.0,
      "step": 52960
    },
    {
      "entropy": 5.2973381042480465,
      "epoch": 5.948503397540293,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021202423182802656,
      "loss": 4.7638,
      "mean_token_accuracy": 0.23716307431459427,
      "num_tokens": 118231585.0,
      "step": 52965
    },
    {
      "entropy": 5.373033332824707,
      "epoch": 5.949064974448251,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021198568879968527,
      "loss": 4.935,
      "mean_token_accuracy": 0.2354410097002983,
      "num_tokens": 118241643.0,
      "step": 52970
    },
    {
      "entropy": 5.366337680816651,
      "epoch": 5.949626551356208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00021194714777774415,
      "loss": 4.787,
      "mean_token_accuracy": 0.2425659030675888,
      "num_tokens": 118252344.0,
      "step": 52975
    },
    {
      "entropy": 5.324969148635864,
      "epoch": 5.950188128264166,
      "grad_norm": 1.125,
      "learning_rate": 0.0002119086087634303,
      "loss": 4.8118,
      "mean_token_accuracy": 0.24386746436357498,
      "num_tokens": 118263187.0,
      "step": 52980
    },
    {
      "entropy": 5.302988195419312,
      "epoch": 5.950749705172123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021187007175797078,
      "loss": 4.7874,
      "mean_token_accuracy": 0.24221671968698502,
      "num_tokens": 118274602.0,
      "step": 52985
    },
    {
      "entropy": 5.352237033843994,
      "epoch": 5.951311282080081,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002118315367625927,
      "loss": 4.8259,
      "mean_token_accuracy": 0.2456492230296135,
      "num_tokens": 118286173.0,
      "step": 52990
    },
    {
      "entropy": 5.368943166732788,
      "epoch": 5.951872858988039,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00021179300377852296,
      "loss": 4.8679,
      "mean_token_accuracy": 0.23175786286592484,
      "num_tokens": 118295785.0,
      "step": 52995
    },
    {
      "entropy": 5.446654272079468,
      "epoch": 5.952434435895996,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002117544728069885,
      "loss": 5.006,
      "mean_token_accuracy": 0.2269887074828148,
      "num_tokens": 118308026.0,
      "step": 53000
    },
    {
      "entropy": 5.445450639724731,
      "epoch": 5.952996012803953,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002117159438492161,
      "loss": 4.9271,
      "mean_token_accuracy": 0.236155666410923,
      "num_tokens": 118319283.0,
      "step": 53005
    },
    {
      "entropy": 5.297165822982788,
      "epoch": 5.953557589711911,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00021167741690643268,
      "loss": 4.6876,
      "mean_token_accuracy": 0.2489714056253433,
      "num_tokens": 118330094.0,
      "step": 53010
    },
    {
      "entropy": 5.244270133972168,
      "epoch": 5.9541191666198685,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002116388919798648,
      "loss": 4.6736,
      "mean_token_accuracy": 0.25826686769723894,
      "num_tokens": 118341271.0,
      "step": 53015
    },
    {
      "entropy": 5.3319660186767575,
      "epoch": 5.954680743527826,
      "grad_norm": 1.109375,
      "learning_rate": 0.00021160036907073921,
      "loss": 4.7682,
      "mean_token_accuracy": 0.2347572922706604,
      "num_tokens": 118351451.0,
      "step": 53020
    },
    {
      "entropy": 5.285858821868897,
      "epoch": 5.955242320435784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021156184818028247,
      "loss": 4.7487,
      "mean_token_accuracy": 0.24658646136522294,
      "num_tokens": 118362375.0,
      "step": 53025
    },
    {
      "entropy": 5.215389680862427,
      "epoch": 5.955803897343741,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021152332930972107,
      "loss": 4.6837,
      "mean_token_accuracy": 0.24302284121513368,
      "num_tokens": 118372020.0,
      "step": 53030
    },
    {
      "entropy": 5.27696213722229,
      "epoch": 5.956365474251699,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021148481246028146,
      "loss": 4.752,
      "mean_token_accuracy": 0.24155394434928895,
      "num_tokens": 118381750.0,
      "step": 53035
    },
    {
      "entropy": 5.265909719467163,
      "epoch": 5.956927051159656,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00021144629763318997,
      "loss": 4.6896,
      "mean_token_accuracy": 0.2454355239868164,
      "num_tokens": 118392648.0,
      "step": 53040
    },
    {
      "entropy": 5.3780759334564205,
      "epoch": 5.957488628067614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002114077848296731,
      "loss": 4.8953,
      "mean_token_accuracy": 0.23652449399232864,
      "num_tokens": 118403441.0,
      "step": 53045
    },
    {
      "entropy": 5.398523998260498,
      "epoch": 5.958050204975572,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021136927405095707,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2348824828863144,
      "num_tokens": 118414520.0,
      "step": 53050
    },
    {
      "entropy": 5.354370164871216,
      "epoch": 5.958611781883529,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000211330765298268,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2407940909266472,
      "num_tokens": 118425048.0,
      "step": 53055
    },
    {
      "entropy": 5.317968893051147,
      "epoch": 5.959173358791486,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002112922585728321,
      "loss": 4.7129,
      "mean_token_accuracy": 0.24368281811475753,
      "num_tokens": 118435780.0,
      "step": 53060
    },
    {
      "entropy": 5.264994192123413,
      "epoch": 5.959734935699444,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002112537538758753,
      "loss": 4.7451,
      "mean_token_accuracy": 0.2422763079404831,
      "num_tokens": 118446733.0,
      "step": 53065
    },
    {
      "entropy": 5.30240912437439,
      "epoch": 5.9602965126074015,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021121525120862384,
      "loss": 4.8271,
      "mean_token_accuracy": 0.23815187960863113,
      "num_tokens": 118456943.0,
      "step": 53070
    },
    {
      "entropy": 5.2836628437042235,
      "epoch": 5.960858089515359,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021117675057230335,
      "loss": 4.791,
      "mean_token_accuracy": 0.25062100738286974,
      "num_tokens": 118467085.0,
      "step": 53075
    },
    {
      "entropy": 5.3532023429870605,
      "epoch": 5.961419666423317,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00021113825196813997,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23332354575395584,
      "num_tokens": 118478050.0,
      "step": 53080
    },
    {
      "entropy": 5.306705617904663,
      "epoch": 5.961981243331274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021109975539735943,
      "loss": 4.7708,
      "mean_token_accuracy": 0.24110810160636903,
      "num_tokens": 118488982.0,
      "step": 53085
    },
    {
      "entropy": 5.389169216156006,
      "epoch": 5.962542820239232,
      "grad_norm": 0.953125,
      "learning_rate": 0.00021106126086118743,
      "loss": 4.9238,
      "mean_token_accuracy": 0.23730454444885254,
      "num_tokens": 118501760.0,
      "step": 53090
    },
    {
      "entropy": 5.382950973510742,
      "epoch": 5.963104397147189,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021102276836084972,
      "loss": 4.8065,
      "mean_token_accuracy": 0.2388884648680687,
      "num_tokens": 118513960.0,
      "step": 53095
    },
    {
      "entropy": 5.366236972808838,
      "epoch": 5.9636659740551465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021098427789757184,
      "loss": 4.7728,
      "mean_token_accuracy": 0.23083711564540862,
      "num_tokens": 118524810.0,
      "step": 53100
    },
    {
      "entropy": 5.331679391860962,
      "epoch": 5.964227550963105,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002109457894725794,
      "loss": 4.7613,
      "mean_token_accuracy": 0.2375050291419029,
      "num_tokens": 118537327.0,
      "step": 53105
    },
    {
      "entropy": 5.303065204620362,
      "epoch": 5.964789127871062,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00021090730308709777,
      "loss": 4.7467,
      "mean_token_accuracy": 0.24354559779167176,
      "num_tokens": 118549578.0,
      "step": 53110
    },
    {
      "entropy": 5.299870300292969,
      "epoch": 5.96535070477902,
      "grad_norm": 1.140625,
      "learning_rate": 0.00021086881874235252,
      "loss": 4.807,
      "mean_token_accuracy": 0.24000062495470048,
      "num_tokens": 118559872.0,
      "step": 53115
    },
    {
      "entropy": 5.354321575164795,
      "epoch": 5.965912281686977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021083033643956906,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2383415549993515,
      "num_tokens": 118570721.0,
      "step": 53120
    },
    {
      "entropy": 5.300224542617798,
      "epoch": 5.966473858594934,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021079185617997248,
      "loss": 4.7229,
      "mean_token_accuracy": 0.24282481372356415,
      "num_tokens": 118581694.0,
      "step": 53125
    },
    {
      "entropy": 5.313095426559448,
      "epoch": 5.9670354355028925,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021075337796478815,
      "loss": 4.7662,
      "mean_token_accuracy": 0.2511210322380066,
      "num_tokens": 118594026.0,
      "step": 53130
    },
    {
      "entropy": 5.376242160797119,
      "epoch": 5.96759701241085,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00021071490179524116,
      "loss": 4.8446,
      "mean_token_accuracy": 0.24164272844791412,
      "num_tokens": 118607795.0,
      "step": 53135
    },
    {
      "entropy": 5.3043395519256595,
      "epoch": 5.968158589318807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021067642767255663,
      "loss": 4.8281,
      "mean_token_accuracy": 0.24110422432422637,
      "num_tokens": 118618639.0,
      "step": 53140
    },
    {
      "entropy": 5.3816815376281735,
      "epoch": 5.968720166226765,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002106379555979595,
      "loss": 4.81,
      "mean_token_accuracy": 0.23706227093935012,
      "num_tokens": 118628324.0,
      "step": 53145
    },
    {
      "entropy": 5.218087148666382,
      "epoch": 5.969281743134722,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021059948557267493,
      "loss": 4.687,
      "mean_token_accuracy": 0.2440139725804329,
      "num_tokens": 118640005.0,
      "step": 53150
    },
    {
      "entropy": 5.362035751342773,
      "epoch": 5.9698433200426795,
      "grad_norm": 1.171875,
      "learning_rate": 0.00021056101759792773,
      "loss": 4.8956,
      "mean_token_accuracy": 0.23357243835926056,
      "num_tokens": 118651179.0,
      "step": 53155
    },
    {
      "entropy": 5.261061573028565,
      "epoch": 5.970404896950638,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021052255167494272,
      "loss": 4.8129,
      "mean_token_accuracy": 0.23382124900817872,
      "num_tokens": 118661506.0,
      "step": 53160
    },
    {
      "entropy": 5.343864393234253,
      "epoch": 5.970966473858595,
      "grad_norm": 1.109375,
      "learning_rate": 0.00021048408780494467,
      "loss": 4.8332,
      "mean_token_accuracy": 0.23804474025964736,
      "num_tokens": 118672252.0,
      "step": 53165
    },
    {
      "entropy": 5.387560415267944,
      "epoch": 5.971528050766553,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00021044562598915824,
      "loss": 4.8684,
      "mean_token_accuracy": 0.23996058106422424,
      "num_tokens": 118684407.0,
      "step": 53170
    },
    {
      "entropy": 5.396267986297607,
      "epoch": 5.97208962767451,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002104071662288081,
      "loss": 4.8862,
      "mean_token_accuracy": 0.2367575541138649,
      "num_tokens": 118695808.0,
      "step": 53175
    },
    {
      "entropy": 5.3975057125091555,
      "epoch": 5.972651204582467,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00021036870852511892,
      "loss": 4.8129,
      "mean_token_accuracy": 0.23958427906036378,
      "num_tokens": 118708701.0,
      "step": 53180
    },
    {
      "entropy": 5.388815546035767,
      "epoch": 5.9732127814904254,
      "grad_norm": 1.0,
      "learning_rate": 0.000210330252879315,
      "loss": 4.8477,
      "mean_token_accuracy": 0.2427206814289093,
      "num_tokens": 118720618.0,
      "step": 53185
    },
    {
      "entropy": 5.26605257987976,
      "epoch": 5.973774358398383,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021029179929262098,
      "loss": 4.7498,
      "mean_token_accuracy": 0.24906917959451674,
      "num_tokens": 118732536.0,
      "step": 53190
    },
    {
      "entropy": 5.258680295944214,
      "epoch": 5.97433593530634,
      "grad_norm": 1.15625,
      "learning_rate": 0.00021025334776626122,
      "loss": 4.7547,
      "mean_token_accuracy": 0.24200102537870408,
      "num_tokens": 118742052.0,
      "step": 53195
    },
    {
      "entropy": 5.227660465240478,
      "epoch": 5.974897512214298,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021021489830145995,
      "loss": 4.7056,
      "mean_token_accuracy": 0.2503634437918663,
      "num_tokens": 118752760.0,
      "step": 53200
    },
    {
      "entropy": 5.306587028503418,
      "epoch": 5.975459089122255,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021017645089944142,
      "loss": 4.7608,
      "mean_token_accuracy": 0.24137572050094605,
      "num_tokens": 118763512.0,
      "step": 53205
    },
    {
      "entropy": 5.311144495010376,
      "epoch": 5.976020666030212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021013800556142986,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2340591236948967,
      "num_tokens": 118774838.0,
      "step": 53210
    },
    {
      "entropy": 5.32604603767395,
      "epoch": 5.9765822429381705,
      "grad_norm": 1.0625,
      "learning_rate": 0.00021009956228864934,
      "loss": 4.7354,
      "mean_token_accuracy": 0.2497787967324257,
      "num_tokens": 118785548.0,
      "step": 53215
    },
    {
      "entropy": 5.286301183700561,
      "epoch": 5.977143819846128,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002100611210823239,
      "loss": 4.7189,
      "mean_token_accuracy": 0.24076679348945618,
      "num_tokens": 118796804.0,
      "step": 53220
    },
    {
      "entropy": 5.267124986648559,
      "epoch": 5.977705396754086,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00021002268194367764,
      "loss": 4.7775,
      "mean_token_accuracy": 0.2398952916264534,
      "num_tokens": 118808232.0,
      "step": 53225
    },
    {
      "entropy": 5.334344100952149,
      "epoch": 5.978266973662043,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002099842448739343,
      "loss": 4.7911,
      "mean_token_accuracy": 0.24159238636493682,
      "num_tokens": 118820331.0,
      "step": 53230
    },
    {
      "entropy": 5.350548028945923,
      "epoch": 5.97882855057,
      "grad_norm": 1.109375,
      "learning_rate": 0.00020994580987431787,
      "loss": 4.8438,
      "mean_token_accuracy": 0.2356039047241211,
      "num_tokens": 118830484.0,
      "step": 53235
    },
    {
      "entropy": 5.322480058670044,
      "epoch": 5.979390127477958,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00020990737694605204,
      "loss": 4.833,
      "mean_token_accuracy": 0.2343960091471672,
      "num_tokens": 118842130.0,
      "step": 53240
    },
    {
      "entropy": 5.3551851272583,
      "epoch": 5.979951704385916,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002098689460903606,
      "loss": 4.8013,
      "mean_token_accuracy": 0.24249498397111893,
      "num_tokens": 118853449.0,
      "step": 53245
    },
    {
      "entropy": 5.32408766746521,
      "epoch": 5.980513281293873,
      "grad_norm": 0.984375,
      "learning_rate": 0.00020983051730846724,
      "loss": 4.8087,
      "mean_token_accuracy": 0.2341926321387291,
      "num_tokens": 118864694.0,
      "step": 53250
    },
    {
      "entropy": 5.34204134941101,
      "epoch": 5.981074858201831,
      "grad_norm": 1.09375,
      "learning_rate": 0.00020979209060159533,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23728945553302766,
      "num_tokens": 118875628.0,
      "step": 53255
    },
    {
      "entropy": 5.35408124923706,
      "epoch": 5.981636435109788,
      "grad_norm": 1.015625,
      "learning_rate": 0.00020975366597096863,
      "loss": 4.8018,
      "mean_token_accuracy": 0.2440395623445511,
      "num_tokens": 118887241.0,
      "step": 53260
    },
    {
      "entropy": 5.325093936920166,
      "epoch": 5.982198012017745,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002097152434178105,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23288943171501159,
      "num_tokens": 118898642.0,
      "step": 53265
    },
    {
      "entropy": 5.280015325546264,
      "epoch": 5.9827595889257035,
      "grad_norm": 1.140625,
      "learning_rate": 0.00020967682294334439,
      "loss": 4.7994,
      "mean_token_accuracy": 0.23759506940841674,
      "num_tokens": 118908313.0,
      "step": 53270
    },
    {
      "entropy": 5.231676769256592,
      "epoch": 5.983321165833661,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00020963840454879358,
      "loss": 4.7091,
      "mean_token_accuracy": 0.24996860921382905,
      "num_tokens": 118919021.0,
      "step": 53275
    },
    {
      "entropy": 5.302547645568848,
      "epoch": 5.983882742741619,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002095999882353813,
      "loss": 4.7608,
      "mean_token_accuracy": 0.24193157702684404,
      "num_tokens": 118931178.0,
      "step": 53280
    },
    {
      "entropy": 5.262881803512573,
      "epoch": 5.984444319649576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00020956157400433073,
      "loss": 4.7452,
      "mean_token_accuracy": 0.24571855068206788,
      "num_tokens": 118942652.0,
      "step": 53285
    },
    {
      "entropy": 5.311073446273804,
      "epoch": 5.985005896557533,
      "grad_norm": 1.078125,
      "learning_rate": 0.000209523161856865,
      "loss": 4.7816,
      "mean_token_accuracy": 0.23954159170389175,
      "num_tokens": 118952539.0,
      "step": 53290
    },
    {
      "entropy": 5.360652685165405,
      "epoch": 5.985567473465491,
      "grad_norm": 0.984375,
      "learning_rate": 0.00020948475179420723,
      "loss": 4.8605,
      "mean_token_accuracy": 0.2444551780819893,
      "num_tokens": 118965389.0,
      "step": 53295
    },
    {
      "entropy": 5.39256477355957,
      "epoch": 5.986129050373449,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00020944634381758038,
      "loss": 4.8633,
      "mean_token_accuracy": 0.24350703060626983,
      "num_tokens": 118978009.0,
      "step": 53300
    },
    {
      "entropy": 5.340012693405152,
      "epoch": 5.986690627281407,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00020940793792820735,
      "loss": 4.8567,
      "mean_token_accuracy": 0.23352669179439545,
      "num_tokens": 118988729.0,
      "step": 53305
    },
    {
      "entropy": 5.416486978530884,
      "epoch": 5.987252204189364,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00020936953412731103,
      "loss": 4.8593,
      "mean_token_accuracy": 0.23563915193080903,
      "num_tokens": 119001533.0,
      "step": 53310
    },
    {
      "entropy": 5.344382286071777,
      "epoch": 5.987813781097321,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00020933113241611423,
      "loss": 4.787,
      "mean_token_accuracy": 0.240438774228096,
      "num_tokens": 119012104.0,
      "step": 53315
    },
    {
      "entropy": 5.357608366012573,
      "epoch": 5.988375358005279,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002092927327958396,
      "loss": 4.8747,
      "mean_token_accuracy": 0.23360259234905242,
      "num_tokens": 119022684.0,
      "step": 53320
    },
    {
      "entropy": 5.304161500930786,
      "epoch": 5.988936934913236,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00020925433526770986,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23920630663633347,
      "num_tokens": 119033430.0,
      "step": 53325
    },
    {
      "entropy": 5.346327590942383,
      "epoch": 5.989498511821194,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00020921593983294745,
      "loss": 4.7947,
      "mean_token_accuracy": 0.2369563415646553,
      "num_tokens": 119043783.0,
      "step": 53330
    },
    {
      "entropy": 5.353311538696289,
      "epoch": 5.990060088729152,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002091775464927752,
      "loss": 4.8132,
      "mean_token_accuracy": 0.2379778727889061,
      "num_tokens": 119053555.0,
      "step": 53335
    },
    {
      "entropy": 5.341504144668579,
      "epoch": 5.990621665637109,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00020913915524841536,
      "loss": 4.7652,
      "mean_token_accuracy": 0.2414470434188843,
      "num_tokens": 119064780.0,
      "step": 53340
    },
    {
      "entropy": 5.294338893890381,
      "epoch": 5.991183242545066,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00020910076610109035,
      "loss": 4.7751,
      "mean_token_accuracy": 0.2391240805387497,
      "num_tokens": 119075991.0,
      "step": 53345
    },
    {
      "entropy": 5.332740449905396,
      "epoch": 5.991744819453024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00020906237905202253,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2357625275850296,
      "num_tokens": 119088076.0,
      "step": 53350
    },
    {
      "entropy": 5.293495845794678,
      "epoch": 5.9923063963609815,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00020902399410243405,
      "loss": 4.8134,
      "mean_token_accuracy": 0.23866510689258574,
      "num_tokens": 119098775.0,
      "step": 53355
    },
    {
      "entropy": 5.429359769821167,
      "epoch": 5.99286797326894,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002089856112535473,
      "loss": 4.9259,
      "mean_token_accuracy": 0.23437703102827073,
      "num_tokens": 119110129.0,
      "step": 53360
    },
    {
      "entropy": 5.344730186462402,
      "epoch": 5.993429550176897,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00020894723050658421,
      "loss": 4.7686,
      "mean_token_accuracy": 0.24627690464258195,
      "num_tokens": 119122117.0,
      "step": 53365
    },
    {
      "entropy": 5.378541707992554,
      "epoch": 5.993991127084854,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00020890885186276693,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2370368078351021,
      "num_tokens": 119135426.0,
      "step": 53370
    },
    {
      "entropy": 5.308378028869629,
      "epoch": 5.994552703992812,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002088704753233175,
      "loss": 4.7972,
      "mean_token_accuracy": 0.23776538521051407,
      "num_tokens": 119146365.0,
      "step": 53375
    },
    {
      "entropy": 5.391438007354736,
      "epoch": 5.995114280900769,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002088321008894577,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23801590204238893,
      "num_tokens": 119157580.0,
      "step": 53380
    },
    {
      "entropy": 5.278424882888794,
      "epoch": 5.995675857808727,
      "grad_norm": 1.015625,
      "learning_rate": 0.00020879372856240952,
      "loss": 4.7287,
      "mean_token_accuracy": 0.24455349147319794,
      "num_tokens": 119168866.0,
      "step": 53385
    },
    {
      "entropy": 5.356719446182251,
      "epoch": 5.996237434716685,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002087553583433947,
      "loss": 4.8196,
      "mean_token_accuracy": 0.2372434049844742,
      "num_tokens": 119180548.0,
      "step": 53390
    },
    {
      "entropy": 5.3705016613006595,
      "epoch": 5.996799011624642,
      "grad_norm": 1.078125,
      "learning_rate": 0.00020871699023363493,
      "loss": 4.8662,
      "mean_token_accuracy": 0.23941477090120317,
      "num_tokens": 119192089.0,
      "step": 53395
    },
    {
      "entropy": 5.297486782073975,
      "epoch": 5.997360588532599,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002086786242343519,
      "loss": 4.7333,
      "mean_token_accuracy": 0.2439983755350113,
      "num_tokens": 119202869.0,
      "step": 53400
    },
    {
      "entropy": 5.311602640151977,
      "epoch": 5.997922165440557,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00020864026034676718,
      "loss": 4.7879,
      "mean_token_accuracy": 0.2419580340385437,
      "num_tokens": 119214107.0,
      "step": 53405
    },
    {
      "entropy": 5.3338478088378904,
      "epoch": 5.9984837423485144,
      "grad_norm": 1.078125,
      "learning_rate": 0.00020860189857210238,
      "loss": 4.8499,
      "mean_token_accuracy": 0.23969611674547195,
      "num_tokens": 119224557.0,
      "step": 53410
    },
    {
      "entropy": 5.25014100074768,
      "epoch": 5.9990453192564726,
      "grad_norm": 1.078125,
      "learning_rate": 0.00020856353891157882,
      "loss": 4.7614,
      "mean_token_accuracy": 0.2412363976240158,
      "num_tokens": 119234792.0,
      "step": 53415
    },
    {
      "entropy": 5.3125158786773685,
      "epoch": 5.99960689616443,
      "grad_norm": 1.0625,
      "learning_rate": 0.00020852518136641802,
      "loss": 4.758,
      "mean_token_accuracy": 0.23795027583837508,
      "num_tokens": 119246228.0,
      "step": 53420
    },
    {
      "entropy": 5.393482526143392,
      "epoch": 6.000112315381592,
      "grad_norm": 1.09375,
      "learning_rate": 0.00020848682593784117,
      "loss": 4.8777,
      "mean_token_accuracy": 0.2334193785985311,
      "num_tokens": 119255127.0,
      "step": 53425
    },
    {
      "entropy": 5.3250633716583256,
      "epoch": 6.000673892289549,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002084484726270696,
      "loss": 4.7653,
      "mean_token_accuracy": 0.23627821654081343,
      "num_tokens": 119265817.0,
      "step": 53430
    },
    {
      "entropy": 5.291588640213012,
      "epoch": 6.001235469197507,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00020841012143532446,
      "loss": 4.6871,
      "mean_token_accuracy": 0.25043177604675293,
      "num_tokens": 119278105.0,
      "step": 53435
    },
    {
      "entropy": 5.33927435874939,
      "epoch": 6.001797046105464,
      "grad_norm": 1.046875,
      "learning_rate": 0.00020837177236382687,
      "loss": 4.8381,
      "mean_token_accuracy": 0.23494777232408523,
      "num_tokens": 119289704.0,
      "step": 53440
    },
    {
      "entropy": 5.286820936203003,
      "epoch": 6.002358623013421,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002083334254137979,
      "loss": 4.6601,
      "mean_token_accuracy": 0.24776484370231627,
      "num_tokens": 119300992.0,
      "step": 53445
    },
    {
      "entropy": 5.334121751785278,
      "epoch": 6.0029201999213795,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002082950805864585,
      "loss": 4.7602,
      "mean_token_accuracy": 0.24448883831501006,
      "num_tokens": 119312428.0,
      "step": 53450
    },
    {
      "entropy": 5.313629293441773,
      "epoch": 6.003481776829337,
      "grad_norm": 1.09375,
      "learning_rate": 0.00020825673788302967,
      "loss": 4.7503,
      "mean_token_accuracy": 0.24739554077386855,
      "num_tokens": 119323696.0,
      "step": 53455
    },
    {
      "entropy": 5.327935266494751,
      "epoch": 6.004043353737294,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002082183973047322,
      "loss": 4.711,
      "mean_token_accuracy": 0.24840424358844757,
      "num_tokens": 119335419.0,
      "step": 53460
    },
    {
      "entropy": 5.295620727539062,
      "epoch": 6.004604930645252,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002081800588527868,
      "loss": 4.7,
      "mean_token_accuracy": 0.2451683133840561,
      "num_tokens": 119346922.0,
      "step": 53465
    },
    {
      "entropy": 5.362142276763916,
      "epoch": 6.005166507553209,
      "grad_norm": 1.28125,
      "learning_rate": 0.00020814172252841417,
      "loss": 4.8049,
      "mean_token_accuracy": 0.2466015785932541,
      "num_tokens": 119359137.0,
      "step": 53470
    },
    {
      "entropy": 5.232109117507934,
      "epoch": 6.005728084461167,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002081033883328351,
      "loss": 4.717,
      "mean_token_accuracy": 0.24987836182117462,
      "num_tokens": 119369021.0,
      "step": 53475
    },
    {
      "entropy": 5.247609329223633,
      "epoch": 6.006289661369125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00020806505626727012,
      "loss": 4.7171,
      "mean_token_accuracy": 0.2446214810013771,
      "num_tokens": 119379210.0,
      "step": 53480
    },
    {
      "entropy": 5.3424656867980955,
      "epoch": 6.006851238277082,
      "grad_norm": 1.1875,
      "learning_rate": 0.00020802672633293967,
      "loss": 4.7633,
      "mean_token_accuracy": 0.24457223415374757,
      "num_tokens": 119389869.0,
      "step": 53485
    },
    {
      "entropy": 5.332297706604004,
      "epoch": 6.00741281518504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00020798839853106434,
      "loss": 4.754,
      "mean_token_accuracy": 0.23884520232677459,
      "num_tokens": 119401158.0,
      "step": 53490
    },
    {
      "entropy": 5.307502937316895,
      "epoch": 6.007974392092997,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00020795007286286427,
      "loss": 4.6743,
      "mean_token_accuracy": 0.2553784683346748,
      "num_tokens": 119412873.0,
      "step": 53495
    },
    {
      "entropy": 5.2159504890441895,
      "epoch": 6.008535969000954,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00020791174932955993,
      "loss": 4.6233,
      "mean_token_accuracy": 0.2597205027937889,
      "num_tokens": 119422796.0,
      "step": 53500
    },
    {
      "entropy": 5.274893712997437,
      "epoch": 6.009097545908912,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00020787342793237147,
      "loss": 4.7125,
      "mean_token_accuracy": 0.24182466715574263,
      "num_tokens": 119433131.0,
      "step": 53505
    },
    {
      "entropy": 5.276225662231445,
      "epoch": 6.00965912281687,
      "grad_norm": 1.0625,
      "learning_rate": 0.00020783510867251908,
      "loss": 4.7692,
      "mean_token_accuracy": 0.2395051121711731,
      "num_tokens": 119444257.0,
      "step": 53510
    },
    {
      "entropy": 5.334344148635864,
      "epoch": 6.010220699724828,
      "grad_norm": 1.078125,
      "learning_rate": 0.00020779679155122295,
      "loss": 4.759,
      "mean_token_accuracy": 0.24111316800117494,
      "num_tokens": 119454824.0,
      "step": 53515
    },
    {
      "entropy": 5.248088455200195,
      "epoch": 6.010782276632785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00020775847656970302,
      "loss": 4.7286,
      "mean_token_accuracy": 0.24759822636842727,
      "num_tokens": 119465550.0,
      "step": 53520
    },
    {
      "entropy": 5.280000162124634,
      "epoch": 6.011343853540742,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002077201637291793,
      "loss": 4.6869,
      "mean_token_accuracy": 0.25074815154075625,
      "num_tokens": 119476513.0,
      "step": 53525
    },
    {
      "entropy": 5.399546384811401,
      "epoch": 6.0119054304487,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002076818530308716,
      "loss": 4.7975,
      "mean_token_accuracy": 0.24101306796073912,
      "num_tokens": 119488431.0,
      "step": 53530
    },
    {
      "entropy": 5.361117219924926,
      "epoch": 6.0124670073566575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00020764354447599988,
      "loss": 4.7463,
      "mean_token_accuracy": 0.24519618451595307,
      "num_tokens": 119499953.0,
      "step": 53535
    },
    {
      "entropy": 5.271987533569336,
      "epoch": 6.013028584264615,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002076052380657838,
      "loss": 4.63,
      "mean_token_accuracy": 0.2565925642848015,
      "num_tokens": 119510179.0,
      "step": 53540
    },
    {
      "entropy": 5.357620191574097,
      "epoch": 6.013590161172573,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00020756693380144297,
      "loss": 4.8219,
      "mean_token_accuracy": 0.240038725733757,
      "num_tokens": 119521428.0,
      "step": 53545
    },
    {
      "entropy": 5.225446367263794,
      "epoch": 6.01415173808053,
      "grad_norm": 1.109375,
      "learning_rate": 0.00020752863168419723,
      "loss": 4.6073,
      "mean_token_accuracy": 0.2577981397509575,
      "num_tokens": 119531369.0,
      "step": 53550
    },
    {
      "entropy": 5.146821594238281,
      "epoch": 6.014713314988487,
      "grad_norm": 1.140625,
      "learning_rate": 0.000207490331715266,
      "loss": 4.5259,
      "mean_token_accuracy": 0.2683349967002869,
      "num_tokens": 119542324.0,
      "step": 53555
    },
    {
      "entropy": 5.1985608577728275,
      "epoch": 6.015274891896445,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00020745203389586886,
      "loss": 4.6902,
      "mean_token_accuracy": 0.24531855136156083,
      "num_tokens": 119552691.0,
      "step": 53560
    },
    {
      "entropy": 5.170269536972046,
      "epoch": 6.015836468804403,
      "grad_norm": 1.078125,
      "learning_rate": 0.00020741373822722508,
      "loss": 4.5884,
      "mean_token_accuracy": 0.2524944469332695,
      "num_tokens": 119563966.0,
      "step": 53565
    },
    {
      "entropy": 5.3174481868743895,
      "epoch": 6.016398045712361,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00020737544471055415,
      "loss": 4.7654,
      "mean_token_accuracy": 0.2415078103542328,
      "num_tokens": 119575862.0,
      "step": 53570
    },
    {
      "entropy": 5.292827939987182,
      "epoch": 6.016959622620318,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002073371533470752,
      "loss": 4.6996,
      "mean_token_accuracy": 0.2440560281276703,
      "num_tokens": 119586285.0,
      "step": 53575
    },
    {
      "entropy": 5.298908329010009,
      "epoch": 6.017521199528275,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002072988641380076,
      "loss": 4.7161,
      "mean_token_accuracy": 0.24395707249641418,
      "num_tokens": 119598201.0,
      "step": 53580
    },
    {
      "entropy": 5.345401334762573,
      "epoch": 6.018082776436233,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00020726057708457036,
      "loss": 4.8161,
      "mean_token_accuracy": 0.24366234689950944,
      "num_tokens": 119608841.0,
      "step": 53585
    },
    {
      "entropy": 5.363494205474853,
      "epoch": 6.0186443533441905,
      "grad_norm": 1.0,
      "learning_rate": 0.00020722229218798267,
      "loss": 4.8763,
      "mean_token_accuracy": 0.23985317945480347,
      "num_tokens": 119622388.0,
      "step": 53590
    },
    {
      "entropy": 5.34201021194458,
      "epoch": 6.019205930252148,
      "grad_norm": 1.03125,
      "learning_rate": 0.00020718400944946347,
      "loss": 4.7252,
      "mean_token_accuracy": 0.24286043792963027,
      "num_tokens": 119634749.0,
      "step": 53595
    },
    {
      "entropy": 5.293387413024902,
      "epoch": 6.019767507160106,
      "grad_norm": 1.125,
      "learning_rate": 0.00020714572887023174,
      "loss": 4.6722,
      "mean_token_accuracy": 0.2493817463517189,
      "num_tokens": 119645962.0,
      "step": 53600
    },
    {
      "entropy": 5.324564838409424,
      "epoch": 6.020329084068063,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002071074504515063,
      "loss": 4.7564,
      "mean_token_accuracy": 0.24140122532844543,
      "num_tokens": 119656444.0,
      "step": 53605
    },
    {
      "entropy": 5.355208826065064,
      "epoch": 6.020890660976021,
      "grad_norm": 1.125,
      "learning_rate": 0.000207069174194506,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23851526528596878,
      "num_tokens": 119668752.0,
      "step": 53610
    },
    {
      "entropy": 5.316359424591065,
      "epoch": 6.021452237883978,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002070309001004494,
      "loss": 4.7715,
      "mean_token_accuracy": 0.2448462039232254,
      "num_tokens": 119680305.0,
      "step": 53615
    },
    {
      "entropy": 5.2875635623931885,
      "epoch": 6.0220138147919355,
      "grad_norm": 1.03125,
      "learning_rate": 0.00020699262817055542,
      "loss": 4.641,
      "mean_token_accuracy": 0.25850672125816343,
      "num_tokens": 119690788.0,
      "step": 53620
    },
    {
      "entropy": 5.243199348449707,
      "epoch": 6.022575391699894,
      "grad_norm": 1.125,
      "learning_rate": 0.00020695435840604255,
      "loss": 4.6526,
      "mean_token_accuracy": 0.24040796160697936,
      "num_tokens": 119702514.0,
      "step": 53625
    },
    {
      "entropy": 5.406851720809937,
      "epoch": 6.023136968607851,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002069160908081293,
      "loss": 4.8272,
      "mean_token_accuracy": 0.2372339442372322,
      "num_tokens": 119713321.0,
      "step": 53630
    },
    {
      "entropy": 5.250484037399292,
      "epoch": 6.023698545515808,
      "grad_norm": 1.046875,
      "learning_rate": 0.00020687782537803407,
      "loss": 4.6712,
      "mean_token_accuracy": 0.24900168031454087,
      "num_tokens": 119723163.0,
      "step": 53635
    },
    {
      "entropy": 5.1846764087677,
      "epoch": 6.024260122423766,
      "grad_norm": 1.078125,
      "learning_rate": 0.00020683956211697537,
      "loss": 4.6594,
      "mean_token_accuracy": 0.25145803540945055,
      "num_tokens": 119733807.0,
      "step": 53640
    },
    {
      "entropy": 5.304272031784057,
      "epoch": 6.024821699331723,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00020680130102617138,
      "loss": 4.7556,
      "mean_token_accuracy": 0.24679191112518312,
      "num_tokens": 119746112.0,
      "step": 53645
    },
    {
      "entropy": 5.249809980392456,
      "epoch": 6.025383276239681,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00020676304210684045,
      "loss": 4.6638,
      "mean_token_accuracy": 0.24794524908065796,
      "num_tokens": 119758086.0,
      "step": 53650
    },
    {
      "entropy": 5.271464824676514,
      "epoch": 6.025944853147639,
      "grad_norm": 1.171875,
      "learning_rate": 0.00020672478536020068,
      "loss": 4.6768,
      "mean_token_accuracy": 0.2579238101840019,
      "num_tokens": 119767814.0,
      "step": 53655
    },
    {
      "entropy": 5.284898281097412,
      "epoch": 6.026506430055596,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002066865307874703,
      "loss": 4.724,
      "mean_token_accuracy": 0.24601031392812728,
      "num_tokens": 119778549.0,
      "step": 53660
    },
    {
      "entropy": 5.375846338272095,
      "epoch": 6.027068006963554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002066482783898672,
      "loss": 4.8412,
      "mean_token_accuracy": 0.24369590878486633,
      "num_tokens": 119791037.0,
      "step": 53665
    },
    {
      "entropy": 5.301203680038452,
      "epoch": 6.027629583871511,
      "grad_norm": 1.125,
      "learning_rate": 0.0002066100281686095,
      "loss": 4.6932,
      "mean_token_accuracy": 0.2427634298801422,
      "num_tokens": 119801455.0,
      "step": 53670
    },
    {
      "entropy": 5.39236626625061,
      "epoch": 6.0281911607794685,
      "grad_norm": 1.03125,
      "learning_rate": 0.000206571780124915,
      "loss": 4.8401,
      "mean_token_accuracy": 0.24015963673591614,
      "num_tokens": 119813312.0,
      "step": 53675
    },
    {
      "entropy": 5.388398456573486,
      "epoch": 6.028752737687427,
      "grad_norm": 1.125,
      "learning_rate": 0.00020653353426000154,
      "loss": 4.7919,
      "mean_token_accuracy": 0.24793309420347215,
      "num_tokens": 119824006.0,
      "step": 53680
    },
    {
      "entropy": 5.306180667877197,
      "epoch": 6.029314314595384,
      "grad_norm": 1.171875,
      "learning_rate": 0.00020649529057508698,
      "loss": 4.7228,
      "mean_token_accuracy": 0.24863775670528412,
      "num_tokens": 119834420.0,
      "step": 53685
    },
    {
      "entropy": 5.348553419113159,
      "epoch": 6.029875891503341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00020645704907138875,
      "loss": 4.7762,
      "mean_token_accuracy": 0.24985164552927017,
      "num_tokens": 119845733.0,
      "step": 53690
    },
    {
      "entropy": 5.332263660430908,
      "epoch": 6.030437468411299,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00020641880975012477,
      "loss": 4.7942,
      "mean_token_accuracy": 0.24499650597572326,
      "num_tokens": 119855990.0,
      "step": 53695
    },
    {
      "entropy": 5.359835910797119,
      "epoch": 6.030999045319256,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00020638057261251254,
      "loss": 4.7886,
      "mean_token_accuracy": 0.23799397349357604,
      "num_tokens": 119866211.0,
      "step": 53700
    },
    {
      "entropy": 5.395473051071167,
      "epoch": 6.0315606222272145,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00020634233765976946,
      "loss": 4.7826,
      "mean_token_accuracy": 0.24104208648204803,
      "num_tokens": 119878149.0,
      "step": 53705
    },
    {
      "entropy": 5.322150230407715,
      "epoch": 6.032122199135172,
      "grad_norm": 1.125,
      "learning_rate": 0.00020630410489311298,
      "loss": 4.7436,
      "mean_token_accuracy": 0.2502454072237015,
      "num_tokens": 119888745.0,
      "step": 53710
    },
    {
      "entropy": 5.335811805725098,
      "epoch": 6.032683776043129,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00020626587431376043,
      "loss": 4.7784,
      "mean_token_accuracy": 0.23833724558353425,
      "num_tokens": 119900904.0,
      "step": 53715
    },
    {
      "entropy": 5.317043924331665,
      "epoch": 6.033245352951087,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002062276459229291,
      "loss": 4.7623,
      "mean_token_accuracy": 0.23794427514076233,
      "num_tokens": 119911624.0,
      "step": 53720
    },
    {
      "entropy": 5.336026000976562,
      "epoch": 6.033806929859044,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00020618941972183613,
      "loss": 4.7934,
      "mean_token_accuracy": 0.23747270405292512,
      "num_tokens": 119922914.0,
      "step": 53725
    },
    {
      "entropy": 5.31144871711731,
      "epoch": 6.034368506767001,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002061511957116988,
      "loss": 4.7246,
      "mean_token_accuracy": 0.24084382653236389,
      "num_tokens": 119934337.0,
      "step": 53730
    },
    {
      "entropy": 5.321654272079468,
      "epoch": 6.0349300836749595,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00020611297389373413,
      "loss": 4.6864,
      "mean_token_accuracy": 0.2605693906545639,
      "num_tokens": 119944993.0,
      "step": 53735
    },
    {
      "entropy": 5.3523674488067625,
      "epoch": 6.035491660582917,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00020607475426915906,
      "loss": 4.7684,
      "mean_token_accuracy": 0.2466003566980362,
      "num_tokens": 119956044.0,
      "step": 53740
    },
    {
      "entropy": 5.248297691345215,
      "epoch": 6.036053237490874,
      "grad_norm": 1.140625,
      "learning_rate": 0.00020603653683919055,
      "loss": 4.6862,
      "mean_token_accuracy": 0.24862728267908096,
      "num_tokens": 119967375.0,
      "step": 53745
    },
    {
      "entropy": 5.178760671615601,
      "epoch": 6.036614814398832,
      "grad_norm": 1.125,
      "learning_rate": 0.00020599832160504545,
      "loss": 4.633,
      "mean_token_accuracy": 0.25857930034399035,
      "num_tokens": 119977901.0,
      "step": 53750
    },
    {
      "entropy": 5.341062021255493,
      "epoch": 6.037176391306789,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00020596010856794057,
      "loss": 4.7655,
      "mean_token_accuracy": 0.24295441508293153,
      "num_tokens": 119989345.0,
      "step": 53755
    },
    {
      "entropy": 5.310883331298828,
      "epoch": 6.037737968214747,
      "grad_norm": 0.96875,
      "learning_rate": 0.00020592189772909248,
      "loss": 4.7703,
      "mean_token_accuracy": 0.24296208918094636,
      "num_tokens": 120002148.0,
      "step": 53760
    },
    {
      "entropy": 5.355102491378784,
      "epoch": 6.038299545122705,
      "grad_norm": 1.03125,
      "learning_rate": 0.00020588368908971806,
      "loss": 4.7585,
      "mean_token_accuracy": 0.2512926608324051,
      "num_tokens": 120015928.0,
      "step": 53765
    },
    {
      "entropy": 5.381805086135865,
      "epoch": 6.038861122030662,
      "grad_norm": 1.078125,
      "learning_rate": 0.00020584548265103382,
      "loss": 4.8443,
      "mean_token_accuracy": 0.2414686769247055,
      "num_tokens": 120028008.0,
      "step": 53770
    },
    {
      "entropy": 5.234278917312622,
      "epoch": 6.03942269893862,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00020580727841425624,
      "loss": 4.6737,
      "mean_token_accuracy": 0.24798963218927383,
      "num_tokens": 120039214.0,
      "step": 53775
    },
    {
      "entropy": 5.194100999832154,
      "epoch": 6.039984275846577,
      "grad_norm": 1.109375,
      "learning_rate": 0.00020576907638060167,
      "loss": 4.6084,
      "mean_token_accuracy": 0.2558003902435303,
      "num_tokens": 120050189.0,
      "step": 53780
    },
    {
      "entropy": 5.223897886276245,
      "epoch": 6.040545852754534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00020573087655128658,
      "loss": 4.6103,
      "mean_token_accuracy": 0.25135239213705063,
      "num_tokens": 120061036.0,
      "step": 53785
    },
    {
      "entropy": 5.282159662246704,
      "epoch": 6.0411074296624925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002056926789275273,
      "loss": 4.7222,
      "mean_token_accuracy": 0.24673686176538467,
      "num_tokens": 120071208.0,
      "step": 53790
    },
    {
      "entropy": 5.30632905960083,
      "epoch": 6.04166900657045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00020565448351053995,
      "loss": 4.7952,
      "mean_token_accuracy": 0.2409922733902931,
      "num_tokens": 120083134.0,
      "step": 53795
    },
    {
      "entropy": 5.264612627029419,
      "epoch": 6.042230583478408,
      "grad_norm": 1.046875,
      "learning_rate": 0.00020561629030154077,
      "loss": 4.643,
      "mean_token_accuracy": 0.2554131865501404,
      "num_tokens": 120094564.0,
      "step": 53800
    },
    {
      "entropy": 5.33672022819519,
      "epoch": 6.042792160386365,
      "grad_norm": 1.078125,
      "learning_rate": 0.00020557809930174582,
      "loss": 4.8354,
      "mean_token_accuracy": 0.23911887258291245,
      "num_tokens": 120106027.0,
      "step": 53805
    },
    {
      "entropy": 5.457008361816406,
      "epoch": 6.043353737294322,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002055399105123711,
      "loss": 4.8213,
      "mean_token_accuracy": 0.22789683192968369,
      "num_tokens": 120117435.0,
      "step": 53810
    },
    {
      "entropy": 5.27997088432312,
      "epoch": 6.04391531420228,
      "grad_norm": 1.0625,
      "learning_rate": 0.00020550172393463256,
      "loss": 4.6986,
      "mean_token_accuracy": 0.24592396169900893,
      "num_tokens": 120127502.0,
      "step": 53815
    },
    {
      "entropy": 5.25882740020752,
      "epoch": 6.044476891110238,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002054635395697461,
      "loss": 4.7242,
      "mean_token_accuracy": 0.24966303259134293,
      "num_tokens": 120138080.0,
      "step": 53820
    },
    {
      "entropy": 5.342452144622802,
      "epoch": 6.045038468018195,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00020542535741892754,
      "loss": 4.779,
      "mean_token_accuracy": 0.24048155546188354,
      "num_tokens": 120148468.0,
      "step": 53825
    },
    {
      "entropy": 5.328269290924072,
      "epoch": 6.045600044926153,
      "grad_norm": 1.109375,
      "learning_rate": 0.00020538717748339248,
      "loss": 4.7626,
      "mean_token_accuracy": 0.24611260145902633,
      "num_tokens": 120159563.0,
      "step": 53830
    },
    {
      "entropy": 5.326879692077637,
      "epoch": 6.04616162183411,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00020534899976435672,
      "loss": 4.7688,
      "mean_token_accuracy": 0.2462201312184334,
      "num_tokens": 120169186.0,
      "step": 53835
    },
    {
      "entropy": 5.416034460067749,
      "epoch": 6.046723198742067,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002053108242630359,
      "loss": 4.9115,
      "mean_token_accuracy": 0.23397413343191148,
      "num_tokens": 120181046.0,
      "step": 53840
    },
    {
      "entropy": 5.249190187454223,
      "epoch": 6.047284775650025,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00020527265098064547,
      "loss": 4.701,
      "mean_token_accuracy": 0.243500055372715,
      "num_tokens": 120190808.0,
      "step": 53845
    },
    {
      "entropy": 5.278196096420288,
      "epoch": 6.047846352557983,
      "grad_norm": 0.984375,
      "learning_rate": 0.00020523447991840078,
      "loss": 4.7173,
      "mean_token_accuracy": 0.24394591748714448,
      "num_tokens": 120202946.0,
      "step": 53850
    },
    {
      "entropy": 5.334074211120606,
      "epoch": 6.048407929465941,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00020519631107751736,
      "loss": 4.7339,
      "mean_token_accuracy": 0.25441970080137255,
      "num_tokens": 120214571.0,
      "step": 53855
    },
    {
      "entropy": 5.324241590499878,
      "epoch": 6.048969506373898,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002051581444592105,
      "loss": 4.7508,
      "mean_token_accuracy": 0.24376674890518188,
      "num_tokens": 120224772.0,
      "step": 53860
    },
    {
      "entropy": 5.338378858566284,
      "epoch": 6.049531083281855,
      "grad_norm": 1.0625,
      "learning_rate": 0.00020511998006469533,
      "loss": 4.7367,
      "mean_token_accuracy": 0.24418913424015046,
      "num_tokens": 120236216.0,
      "step": 53865
    },
    {
      "entropy": 5.319238615036011,
      "epoch": 6.050092660189813,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002050818178951871,
      "loss": 4.7376,
      "mean_token_accuracy": 0.24092504978179932,
      "num_tokens": 120247230.0,
      "step": 53870
    },
    {
      "entropy": 5.228237628936768,
      "epoch": 6.0506542370977705,
      "grad_norm": 1.078125,
      "learning_rate": 0.000205043657951901,
      "loss": 4.675,
      "mean_token_accuracy": 0.24884769916534424,
      "num_tokens": 120258036.0,
      "step": 53875
    },
    {
      "entropy": 5.297597885131836,
      "epoch": 6.051215814005728,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002050055002360519,
      "loss": 4.735,
      "mean_token_accuracy": 0.2469738930463791,
      "num_tokens": 120269483.0,
      "step": 53880
    },
    {
      "entropy": 5.2449524879455565,
      "epoch": 6.051777390913686,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002049673447488548,
      "loss": 4.6667,
      "mean_token_accuracy": 0.24441668540239334,
      "num_tokens": 120280345.0,
      "step": 53885
    },
    {
      "entropy": 5.3247767925262455,
      "epoch": 6.052338967821643,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00020492919149152467,
      "loss": 4.7168,
      "mean_token_accuracy": 0.247561976313591,
      "num_tokens": 120291245.0,
      "step": 53890
    },
    {
      "entropy": 5.273043584823609,
      "epoch": 6.052900544729601,
      "grad_norm": 1.0625,
      "learning_rate": 0.00020489104046527624,
      "loss": 4.7002,
      "mean_token_accuracy": 0.24393037855625152,
      "num_tokens": 120302606.0,
      "step": 53895
    },
    {
      "entropy": 5.283026123046875,
      "epoch": 6.053462121637558,
      "grad_norm": 1.0625,
      "learning_rate": 0.00020485289167132426,
      "loss": 4.7185,
      "mean_token_accuracy": 0.24804951399564742,
      "num_tokens": 120314036.0,
      "step": 53900
    },
    {
      "entropy": 5.29703540802002,
      "epoch": 6.054023698545516,
      "grad_norm": 1.03125,
      "learning_rate": 0.00020481474511088333,
      "loss": 4.7665,
      "mean_token_accuracy": 0.23769553750753403,
      "num_tokens": 120326010.0,
      "step": 53905
    },
    {
      "entropy": 5.378128576278686,
      "epoch": 6.054585275453474,
      "grad_norm": 1.0,
      "learning_rate": 0.0002047766007851682,
      "loss": 4.8163,
      "mean_token_accuracy": 0.23940632492303848,
      "num_tokens": 120337530.0,
      "step": 53910
    },
    {
      "entropy": 5.366470098495483,
      "epoch": 6.055146852361431,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00020473845869539337,
      "loss": 4.7176,
      "mean_token_accuracy": 0.24492294192314149,
      "num_tokens": 120348141.0,
      "step": 53915
    },
    {
      "entropy": 5.301345539093018,
      "epoch": 6.055708429269388,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002047003188427733,
      "loss": 4.758,
      "mean_token_accuracy": 0.2506889134645462,
      "num_tokens": 120359675.0,
      "step": 53920
    },
    {
      "entropy": 5.2994472026824955,
      "epoch": 6.056270006177346,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00020466218122852225,
      "loss": 4.7447,
      "mean_token_accuracy": 0.24748701006174087,
      "num_tokens": 120370821.0,
      "step": 53925
    },
    {
      "entropy": 5.32097864151001,
      "epoch": 6.0568315830853034,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002046240458538547,
      "loss": 4.7664,
      "mean_token_accuracy": 0.251889505982399,
      "num_tokens": 120383198.0,
      "step": 53930
    },
    {
      "entropy": 5.349177122116089,
      "epoch": 6.057393159993261,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002045859127199848,
      "loss": 4.7982,
      "mean_token_accuracy": 0.24302783310413362,
      "num_tokens": 120395801.0,
      "step": 53935
    },
    {
      "entropy": 5.214062070846557,
      "epoch": 6.057954736901219,
      "grad_norm": 1.0625,
      "learning_rate": 0.00020454778182812673,
      "loss": 4.5584,
      "mean_token_accuracy": 0.2583622708916664,
      "num_tokens": 120406138.0,
      "step": 53940
    },
    {
      "entropy": 5.3003710269927975,
      "epoch": 6.058516313809176,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002045096531794946,
      "loss": 4.7899,
      "mean_token_accuracy": 0.2449087083339691,
      "num_tokens": 120417835.0,
      "step": 53945
    },
    {
      "entropy": 5.277628183364868,
      "epoch": 6.059077890717134,
      "grad_norm": 1.0,
      "learning_rate": 0.0002044715267753025,
      "loss": 4.7417,
      "mean_token_accuracy": 0.24306504279375077,
      "num_tokens": 120428831.0,
      "step": 53950
    },
    {
      "entropy": 5.294975519180298,
      "epoch": 6.059639467625091,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002044334026167643,
      "loss": 4.7121,
      "mean_token_accuracy": 0.24762635082006454,
      "num_tokens": 120438599.0,
      "step": 53955
    },
    {
      "entropy": 5.246413469314575,
      "epoch": 6.0602010445330485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000204395280705094,
      "loss": 4.6979,
      "mean_token_accuracy": 0.24047832638025285,
      "num_tokens": 120450113.0,
      "step": 53960
    },
    {
      "entropy": 5.2937822341918945,
      "epoch": 6.060762621441007,
      "grad_norm": 1.0625,
      "learning_rate": 0.00020435716104150525,
      "loss": 4.6836,
      "mean_token_accuracy": 0.25834870934486387,
      "num_tokens": 120461869.0,
      "step": 53965
    },
    {
      "entropy": 5.27721095085144,
      "epoch": 6.061324198348964,
      "grad_norm": 1.0625,
      "learning_rate": 0.00020431904362721198,
      "loss": 4.7271,
      "mean_token_accuracy": 0.24767681062221528,
      "num_tokens": 120472819.0,
      "step": 53970
    },
    {
      "entropy": 5.270786619186401,
      "epoch": 6.061885775256921,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00020428092846342766,
      "loss": 4.6856,
      "mean_token_accuracy": 0.2511787161231041,
      "num_tokens": 120483861.0,
      "step": 53975
    },
    {
      "entropy": 5.3022692680358885,
      "epoch": 6.062447352164879,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002042428155513661,
      "loss": 4.7024,
      "mean_token_accuracy": 0.25228515118360517,
      "num_tokens": 120495268.0,
      "step": 53980
    },
    {
      "entropy": 5.2378637313842775,
      "epoch": 6.063008929072836,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00020420470489224075,
      "loss": 4.6727,
      "mean_token_accuracy": 0.2400696113705635,
      "num_tokens": 120508032.0,
      "step": 53985
    },
    {
      "entropy": 5.282812690734863,
      "epoch": 6.0635705059807945,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000204166596487265,
      "loss": 4.7182,
      "mean_token_accuracy": 0.2561252564191818,
      "num_tokens": 120518791.0,
      "step": 53990
    },
    {
      "entropy": 5.314331531524658,
      "epoch": 6.064132082888752,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002041284903376523,
      "loss": 4.7806,
      "mean_token_accuracy": 0.24260967671871186,
      "num_tokens": 120530694.0,
      "step": 53995
    },
    {
      "entropy": 5.313423347473145,
      "epoch": 6.064693659796709,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00020409038644461598,
      "loss": 4.7897,
      "mean_token_accuracy": 0.24708643406629563,
      "num_tokens": 120542375.0,
      "step": 54000
    },
    {
      "epoch": 6.064693659796709,
      "eval_entropy": 5.093094628958595,
      "eval_loss": 4.980837821960449,
      "eval_mean_token_accuracy": 0.2373161247292683,
      "eval_num_tokens": 120542375.0,
      "eval_runtime": 23.9604,
      "eval_samples_per_second": 1294.342,
      "eval_steps_per_second": 161.808,
      "step": 54000
    }
  ],
  "logging_steps": 5,
  "max_steps": 89030,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.889107709184e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}