{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.322766570605188,
  "eval_steps": 3000,
  "global_step": 45000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 6.845805501937866,
      "epoch": 0.0004803073967339097,
      "grad_norm": 1.2734375,
      "learning_rate": 2e-06,
      "loss": 6.6326,
      "mean_token_accuracy": 0.12870624884963036,
      "num_tokens": 10821.0,
      "step": 5
    },
    {
      "entropy": 6.8324442386627195,
      "epoch": 0.0009606147934678194,
      "grad_norm": 1.5234375,
      "learning_rate": 4.5e-06,
      "loss": 6.5613,
      "mean_token_accuracy": 0.13194445222616197,
      "num_tokens": 21838.0,
      "step": 10
    },
    {
      "entropy": 6.819680023193359,
      "epoch": 0.001440922190201729,
      "grad_norm": 1.1171875,
      "learning_rate": 7e-06,
      "loss": 6.593,
      "mean_token_accuracy": 0.12538625448942184,
      "num_tokens": 33483.0,
      "step": 15
    },
    {
      "entropy": 6.857285118103027,
      "epoch": 0.0019212295869356388,
      "grad_norm": 1.1328125,
      "learning_rate": 9.5e-06,
      "loss": 6.6862,
      "mean_token_accuracy": 0.1256486490368843,
      "num_tokens": 45150.0,
      "step": 20
    },
    {
      "entropy": 6.892150020599365,
      "epoch": 0.0024015369836695487,
      "grad_norm": 1.2421875,
      "learning_rate": 1.2e-05,
      "loss": 6.5826,
      "mean_token_accuracy": 0.12348027229309082,
      "num_tokens": 55862.0,
      "step": 25
    },
    {
      "entropy": 6.904298925399781,
      "epoch": 0.002881844380403458,
      "grad_norm": 1.0703125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 6.6976,
      "mean_token_accuracy": 0.12837719544768333,
      "num_tokens": 66766.0,
      "step": 30
    },
    {
      "entropy": 6.884329032897949,
      "epoch": 0.0033621517771373678,
      "grad_norm": 1.2265625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.4329,
      "mean_token_accuracy": 0.13798720836639405,
      "num_tokens": 77012.0,
      "step": 35
    },
    {
      "entropy": 6.914054870605469,
      "epoch": 0.0038424591738712775,
      "grad_norm": 1.1171875,
      "learning_rate": 1.95e-05,
      "loss": 6.6479,
      "mean_token_accuracy": 0.12707344740629195,
      "num_tokens": 89034.0,
      "step": 40
    },
    {
      "entropy": 6.835504341125488,
      "epoch": 0.004322766570605188,
      "grad_norm": 1.1171875,
      "learning_rate": 2.2e-05,
      "loss": 6.5321,
      "mean_token_accuracy": 0.13442391008138657,
      "num_tokens": 100341.0,
      "step": 45
    },
    {
      "entropy": 6.8050981044769285,
      "epoch": 0.004803073967339097,
      "grad_norm": 1.078125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 6.4944,
      "mean_token_accuracy": 0.1337900787591934,
      "num_tokens": 111997.0,
      "step": 50
    },
    {
      "entropy": 6.8239950180053714,
      "epoch": 0.005283381364073006,
      "grad_norm": 1.078125,
      "learning_rate": 2.7e-05,
      "loss": 6.5452,
      "mean_token_accuracy": 0.12953593134880065,
      "num_tokens": 123705.0,
      "step": 55
    },
    {
      "entropy": 6.816039228439331,
      "epoch": 0.005763688760806916,
      "grad_norm": 1.2109375,
      "learning_rate": 2.95e-05,
      "loss": 6.4881,
      "mean_token_accuracy": 0.13681765496730805,
      "num_tokens": 134091.0,
      "step": 60
    },
    {
      "entropy": 6.790379428863526,
      "epoch": 0.006243996157540826,
      "grad_norm": 1.03125,
      "learning_rate": 3.2e-05,
      "loss": 6.5926,
      "mean_token_accuracy": 0.13416145518422126,
      "num_tokens": 145731.0,
      "step": 65
    },
    {
      "entropy": 6.8630434513092045,
      "epoch": 0.0067243035542747355,
      "grad_norm": 1.203125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 6.51,
      "mean_token_accuracy": 0.1349278561770916,
      "num_tokens": 156979.0,
      "step": 70
    },
    {
      "entropy": 6.868352890014648,
      "epoch": 0.007204610951008645,
      "grad_norm": 1.2421875,
      "learning_rate": 3.7e-05,
      "loss": 6.4964,
      "mean_token_accuracy": 0.1383296526968479,
      "num_tokens": 167021.0,
      "step": 75
    },
    {
      "entropy": 6.792302656173706,
      "epoch": 0.007684918347742555,
      "grad_norm": 1.2109375,
      "learning_rate": 3.95e-05,
      "loss": 6.5049,
      "mean_token_accuracy": 0.13316406682133675,
      "num_tokens": 179269.0,
      "step": 80
    },
    {
      "entropy": 6.745629453659058,
      "epoch": 0.008165225744476465,
      "grad_norm": 1.0859375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 6.4804,
      "mean_token_accuracy": 0.1358616307377815,
      "num_tokens": 190328.0,
      "step": 85
    },
    {
      "entropy": 6.760327434539795,
      "epoch": 0.008645533141210375,
      "grad_norm": 1.296875,
      "learning_rate": 4.45e-05,
      "loss": 6.3536,
      "mean_token_accuracy": 0.139989273250103,
      "num_tokens": 201522.0,
      "step": 90
    },
    {
      "entropy": 6.793478059768677,
      "epoch": 0.009125840537944284,
      "grad_norm": 1.0546875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 6.4743,
      "mean_token_accuracy": 0.14095518440008165,
      "num_tokens": 212963.0,
      "step": 95
    },
    {
      "entropy": 6.847437191009521,
      "epoch": 0.009606147934678195,
      "grad_norm": 1.0703125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.6791,
      "mean_token_accuracy": 0.12346139177680016,
      "num_tokens": 224446.0,
      "step": 100
    },
    {
      "entropy": 6.894752740859985,
      "epoch": 0.010086455331412104,
      "grad_norm": 1.171875,
      "learning_rate": 5.2e-05,
      "loss": 6.5087,
      "mean_token_accuracy": 0.13819111213088037,
      "num_tokens": 235334.0,
      "step": 105
    },
    {
      "entropy": 6.9679093837738035,
      "epoch": 0.010566762728146013,
      "grad_norm": 1.046875,
      "learning_rate": 5.45e-05,
      "loss": 6.5823,
      "mean_token_accuracy": 0.12812134698033334,
      "num_tokens": 247024.0,
      "step": 110
    },
    {
      "entropy": 6.816913843154907,
      "epoch": 0.011047070124879923,
      "grad_norm": 1.171875,
      "learning_rate": 5.7e-05,
      "loss": 6.4988,
      "mean_token_accuracy": 0.1326398141682148,
      "num_tokens": 258354.0,
      "step": 115
    },
    {
      "entropy": 6.757030248641968,
      "epoch": 0.011527377521613832,
      "grad_norm": 1.09375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 6.5505,
      "mean_token_accuracy": 0.13475370407104492,
      "num_tokens": 271202.0,
      "step": 120
    },
    {
      "entropy": 6.8415162563323975,
      "epoch": 0.012007684918347743,
      "grad_norm": 1.2578125,
      "learning_rate": 6.2e-05,
      "loss": 6.5545,
      "mean_token_accuracy": 0.12843683063983918,
      "num_tokens": 282405.0,
      "step": 125
    },
    {
      "entropy": 6.810266208648682,
      "epoch": 0.012487992315081652,
      "grad_norm": 1.0859375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 6.5196,
      "mean_token_accuracy": 0.1312414713203907,
      "num_tokens": 294211.0,
      "step": 130
    },
    {
      "entropy": 6.916629600524902,
      "epoch": 0.012968299711815562,
      "grad_norm": 1.234375,
      "learning_rate": 6.7e-05,
      "loss": 6.5531,
      "mean_token_accuracy": 0.12653037831187247,
      "num_tokens": 305420.0,
      "step": 135
    },
    {
      "entropy": 6.8390460968017575,
      "epoch": 0.013448607108549471,
      "grad_norm": 1.09375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 6.4628,
      "mean_token_accuracy": 0.1347823828458786,
      "num_tokens": 316430.0,
      "step": 140
    },
    {
      "entropy": 6.7258693218231205,
      "epoch": 0.013928914505283382,
      "grad_norm": 1.15625,
      "learning_rate": 7.2e-05,
      "loss": 6.4777,
      "mean_token_accuracy": 0.13652994483709335,
      "num_tokens": 328406.0,
      "step": 145
    },
    {
      "entropy": 6.8050377368927,
      "epoch": 0.01440922190201729,
      "grad_norm": 1.234375,
      "learning_rate": 7.45e-05,
      "loss": 6.4622,
      "mean_token_accuracy": 0.13485521599650382,
      "num_tokens": 339051.0,
      "step": 150
    },
    {
      "entropy": 6.775390768051148,
      "epoch": 0.014889529298751201,
      "grad_norm": 1.1328125,
      "learning_rate": 7.7e-05,
      "loss": 6.4295,
      "mean_token_accuracy": 0.13496242091059685,
      "num_tokens": 350308.0,
      "step": 155
    },
    {
      "entropy": 6.888868618011474,
      "epoch": 0.01536983669548511,
      "grad_norm": 1.1015625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 6.466,
      "mean_token_accuracy": 0.1328311488032341,
      "num_tokens": 362159.0,
      "step": 160
    },
    {
      "entropy": 6.7834686756134035,
      "epoch": 0.01585014409221902,
      "grad_norm": 1.21875,
      "learning_rate": 8.2e-05,
      "loss": 6.5197,
      "mean_token_accuracy": 0.1340256117284298,
      "num_tokens": 375157.0,
      "step": 165
    },
    {
      "entropy": 6.732735347747803,
      "epoch": 0.01633045148895293,
      "grad_norm": 1.125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 6.3494,
      "mean_token_accuracy": 0.1391011193394661,
      "num_tokens": 387318.0,
      "step": 170
    },
    {
      "entropy": 6.73529953956604,
      "epoch": 0.01681075888568684,
      "grad_norm": 1.25,
      "learning_rate": 8.7e-05,
      "loss": 6.4332,
      "mean_token_accuracy": 0.140291528403759,
      "num_tokens": 398933.0,
      "step": 175
    },
    {
      "entropy": 6.753161287307739,
      "epoch": 0.01729106628242075,
      "grad_norm": 1.15625,
      "learning_rate": 8.95e-05,
      "loss": 6.4212,
      "mean_token_accuracy": 0.14206939190626144,
      "num_tokens": 410460.0,
      "step": 180
    },
    {
      "entropy": 6.807398653030395,
      "epoch": 0.01777137367915466,
      "grad_norm": 1.1875,
      "learning_rate": 9.2e-05,
      "loss": 6.4542,
      "mean_token_accuracy": 0.13703907132148743,
      "num_tokens": 421894.0,
      "step": 185
    },
    {
      "entropy": 6.740799045562744,
      "epoch": 0.01825168107588857,
      "grad_norm": 1.203125,
      "learning_rate": 9.45e-05,
      "loss": 6.4923,
      "mean_token_accuracy": 0.13176504224538804,
      "num_tokens": 432976.0,
      "step": 190
    },
    {
      "entropy": 6.845755910873413,
      "epoch": 0.018731988472622477,
      "grad_norm": 1.2109375,
      "learning_rate": 9.7e-05,
      "loss": 6.5375,
      "mean_token_accuracy": 0.1358255296945572,
      "num_tokens": 444859.0,
      "step": 195
    },
    {
      "entropy": 6.780419158935547,
      "epoch": 0.01921229586935639,
      "grad_norm": 1.1640625,
      "learning_rate": 9.95e-05,
      "loss": 6.4609,
      "mean_token_accuracy": 0.1418136939406395,
      "num_tokens": 456508.0,
      "step": 200
    },
    {
      "entropy": 6.7831775665283205,
      "epoch": 0.0196926032660903,
      "grad_norm": 1.28125,
      "learning_rate": 0.000102,
      "loss": 6.4672,
      "mean_token_accuracy": 0.1363604933023453,
      "num_tokens": 469064.0,
      "step": 205
    },
    {
      "entropy": 6.7620916843414305,
      "epoch": 0.020172910662824207,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00010449999999999999,
      "loss": 6.4243,
      "mean_token_accuracy": 0.14104704037308693,
      "num_tokens": 480743.0,
      "step": 210
    },
    {
      "entropy": 6.745031690597534,
      "epoch": 0.020653218059558116,
      "grad_norm": 1.265625,
      "learning_rate": 0.000107,
      "loss": 6.4653,
      "mean_token_accuracy": 0.13698699995875357,
      "num_tokens": 493378.0,
      "step": 215
    },
    {
      "entropy": 6.814111566543579,
      "epoch": 0.021133525456292025,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0001095,
      "loss": 6.4723,
      "mean_token_accuracy": 0.1333048313856125,
      "num_tokens": 505641.0,
      "step": 220
    },
    {
      "entropy": 6.795113658905029,
      "epoch": 0.021613832853025938,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000112,
      "loss": 6.5338,
      "mean_token_accuracy": 0.13384313508868217,
      "num_tokens": 517783.0,
      "step": 225
    },
    {
      "entropy": 6.722929668426514,
      "epoch": 0.022094140249759846,
      "grad_norm": 1.359375,
      "learning_rate": 0.0001145,
      "loss": 6.3621,
      "mean_token_accuracy": 0.1424078032374382,
      "num_tokens": 528407.0,
      "step": 230
    },
    {
      "entropy": 6.831379413604736,
      "epoch": 0.022574447646493755,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00011700000000000001,
      "loss": 6.5324,
      "mean_token_accuracy": 0.1359228663146496,
      "num_tokens": 539974.0,
      "step": 235
    },
    {
      "entropy": 6.731849145889282,
      "epoch": 0.023054755043227664,
      "grad_norm": 1.328125,
      "learning_rate": 0.00011949999999999999,
      "loss": 6.3723,
      "mean_token_accuracy": 0.1379515901207924,
      "num_tokens": 551127.0,
      "step": 240
    },
    {
      "entropy": 6.754569816589355,
      "epoch": 0.023535062439961577,
      "grad_norm": 1.234375,
      "learning_rate": 0.000122,
      "loss": 6.4283,
      "mean_token_accuracy": 0.14257267117500305,
      "num_tokens": 561980.0,
      "step": 245
    },
    {
      "entropy": 6.748660755157471,
      "epoch": 0.024015369836695485,
      "grad_norm": 1.28125,
      "learning_rate": 0.0001245,
      "loss": 6.4724,
      "mean_token_accuracy": 0.13603295460343362,
      "num_tokens": 573840.0,
      "step": 250
    },
    {
      "entropy": 6.679523086547851,
      "epoch": 0.024495677233429394,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000127,
      "loss": 6.4282,
      "mean_token_accuracy": 0.1383022792637348,
      "num_tokens": 585110.0,
      "step": 255
    },
    {
      "entropy": 6.642319250106811,
      "epoch": 0.024975984630163303,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0001295,
      "loss": 6.3071,
      "mean_token_accuracy": 0.1404804766178131,
      "num_tokens": 596902.0,
      "step": 260
    },
    {
      "entropy": 6.77508692741394,
      "epoch": 0.025456292026897216,
      "grad_norm": 1.171875,
      "learning_rate": 0.000132,
      "loss": 6.4858,
      "mean_token_accuracy": 0.13461525812745095,
      "num_tokens": 608738.0,
      "step": 265
    },
    {
      "entropy": 6.804401445388794,
      "epoch": 0.025936599423631124,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00013450000000000002,
      "loss": 6.4096,
      "mean_token_accuracy": 0.13826687932014464,
      "num_tokens": 621078.0,
      "step": 270
    },
    {
      "entropy": 6.799535465240479,
      "epoch": 0.026416906820365033,
      "grad_norm": 1.359375,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.5096,
      "mean_token_accuracy": 0.13709565252065659,
      "num_tokens": 632333.0,
      "step": 275
    },
    {
      "entropy": 6.714285564422608,
      "epoch": 0.026897214217098942,
      "grad_norm": 1.3125,
      "learning_rate": 0.0001395,
      "loss": 6.3294,
      "mean_token_accuracy": 0.13480642214417457,
      "num_tokens": 644474.0,
      "step": 280
    },
    {
      "entropy": 6.669364023208618,
      "epoch": 0.027377521613832854,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.4381,
      "mean_token_accuracy": 0.13586120754480363,
      "num_tokens": 655608.0,
      "step": 285
    },
    {
      "entropy": 6.735668182373047,
      "epoch": 0.027857829010566763,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0001445,
      "loss": 6.4302,
      "mean_token_accuracy": 0.13791066855192186,
      "num_tokens": 667371.0,
      "step": 290
    },
    {
      "entropy": 6.750858640670776,
      "epoch": 0.028338136407300672,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000147,
      "loss": 6.4653,
      "mean_token_accuracy": 0.1322570785880089,
      "num_tokens": 679152.0,
      "step": 295
    },
    {
      "entropy": 6.662555599212647,
      "epoch": 0.02881844380403458,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0001495,
      "loss": 6.3797,
      "mean_token_accuracy": 0.14153480380773545,
      "num_tokens": 690359.0,
      "step": 300
    },
    {
      "entropy": 6.780093812942505,
      "epoch": 0.029298751200768493,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000152,
      "loss": 6.3685,
      "mean_token_accuracy": 0.13668907657265664,
      "num_tokens": 701819.0,
      "step": 305
    },
    {
      "entropy": 6.678733968734742,
      "epoch": 0.029779058597502402,
      "grad_norm": 1.375,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.3892,
      "mean_token_accuracy": 0.13896276280283928,
      "num_tokens": 711048.0,
      "step": 310
    },
    {
      "entropy": 6.702047061920166,
      "epoch": 0.03025936599423631,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000157,
      "loss": 6.3094,
      "mean_token_accuracy": 0.144956311583519,
      "num_tokens": 721399.0,
      "step": 315
    },
    {
      "entropy": 6.66398663520813,
      "epoch": 0.03073967339097022,
      "grad_norm": 1.265625,
      "learning_rate": 0.0001595,
      "loss": 6.323,
      "mean_token_accuracy": 0.14815439581871032,
      "num_tokens": 732144.0,
      "step": 320
    },
    {
      "entropy": 6.663816642761231,
      "epoch": 0.03121998078770413,
      "grad_norm": 1.421875,
      "learning_rate": 0.000162,
      "loss": 6.3222,
      "mean_token_accuracy": 0.14062243029475213,
      "num_tokens": 743962.0,
      "step": 325
    },
    {
      "entropy": 6.774586772918701,
      "epoch": 0.03170028818443804,
      "grad_norm": 1.359375,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.4838,
      "mean_token_accuracy": 0.134712003916502,
      "num_tokens": 755641.0,
      "step": 330
    },
    {
      "entropy": 6.539123487472534,
      "epoch": 0.03218059558117195,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.3358,
      "mean_token_accuracy": 0.14351314306259155,
      "num_tokens": 766798.0,
      "step": 335
    },
    {
      "entropy": 6.761009979248047,
      "epoch": 0.03266090297790586,
      "grad_norm": 1.21875,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.4273,
      "mean_token_accuracy": 0.14208256751298903,
      "num_tokens": 778097.0,
      "step": 340
    },
    {
      "entropy": 6.617664432525634,
      "epoch": 0.03314121037463977,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.3365,
      "mean_token_accuracy": 0.14357409998774529,
      "num_tokens": 789775.0,
      "step": 345
    },
    {
      "entropy": 6.800221252441406,
      "epoch": 0.03362151777137368,
      "grad_norm": 1.390625,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.4119,
      "mean_token_accuracy": 0.1336173504590988,
      "num_tokens": 800155.0,
      "step": 350
    },
    {
      "entropy": 6.688700437545776,
      "epoch": 0.034101825168107586,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000177,
      "loss": 6.4309,
      "mean_token_accuracy": 0.14149890020489692,
      "num_tokens": 811213.0,
      "step": 355
    },
    {
      "entropy": 6.706441020965576,
      "epoch": 0.0345821325648415,
      "grad_norm": 1.3125,
      "learning_rate": 0.0001795,
      "loss": 6.4167,
      "mean_token_accuracy": 0.13566643744707108,
      "num_tokens": 822832.0,
      "step": 360
    },
    {
      "entropy": 6.64611029624939,
      "epoch": 0.03506243996157541,
      "grad_norm": 1.625,
      "learning_rate": 0.000182,
      "loss": 6.4001,
      "mean_token_accuracy": 0.14356471225619316,
      "num_tokens": 834824.0,
      "step": 365
    },
    {
      "entropy": 6.767692947387696,
      "epoch": 0.03554274735830932,
      "grad_norm": 1.265625,
      "learning_rate": 0.0001845,
      "loss": 6.4247,
      "mean_token_accuracy": 0.13687072023749353,
      "num_tokens": 846969.0,
      "step": 370
    },
    {
      "entropy": 6.731805896759033,
      "epoch": 0.03602305475504323,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000187,
      "loss": 6.5275,
      "mean_token_accuracy": 0.1303038403391838,
      "num_tokens": 859271.0,
      "step": 375
    },
    {
      "entropy": 6.70332555770874,
      "epoch": 0.03650336215177714,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0001895,
      "loss": 6.397,
      "mean_token_accuracy": 0.13794534802436828,
      "num_tokens": 869754.0,
      "step": 380
    },
    {
      "entropy": 6.676585054397583,
      "epoch": 0.036983669548511046,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000192,
      "loss": 6.3553,
      "mean_token_accuracy": 0.14230940788984298,
      "num_tokens": 880519.0,
      "step": 385
    },
    {
      "entropy": 6.566428184509277,
      "epoch": 0.037463976945244955,
      "grad_norm": 1.375,
      "learning_rate": 0.0001945,
      "loss": 6.3207,
      "mean_token_accuracy": 0.13795292973518372,
      "num_tokens": 890765.0,
      "step": 390
    },
    {
      "entropy": 6.678460311889649,
      "epoch": 0.037944284341978864,
      "grad_norm": 1.3125,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.3213,
      "mean_token_accuracy": 0.13789281025528907,
      "num_tokens": 901717.0,
      "step": 395
    },
    {
      "entropy": 6.771489000320434,
      "epoch": 0.03842459173871278,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.4173,
      "mean_token_accuracy": 0.13703501969575882,
      "num_tokens": 913334.0,
      "step": 400
    },
    {
      "entropy": 6.673534536361695,
      "epoch": 0.03890489913544669,
      "grad_norm": 1.28125,
      "learning_rate": 0.000202,
      "loss": 6.4358,
      "mean_token_accuracy": 0.13748908787965775,
      "num_tokens": 924753.0,
      "step": 405
    },
    {
      "entropy": 6.69045729637146,
      "epoch": 0.0393852065321806,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.3857,
      "mean_token_accuracy": 0.13825666159391403,
      "num_tokens": 935692.0,
      "step": 410
    },
    {
      "entropy": 6.684324502944946,
      "epoch": 0.039865513928914506,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000207,
      "loss": 6.4318,
      "mean_token_accuracy": 0.13761915788054466,
      "num_tokens": 947902.0,
      "step": 415
    },
    {
      "entropy": 6.693652248382568,
      "epoch": 0.040345821325648415,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002095,
      "loss": 6.4061,
      "mean_token_accuracy": 0.14424072802066804,
      "num_tokens": 959437.0,
      "step": 420
    },
    {
      "entropy": 6.5537275791168215,
      "epoch": 0.040826128722382324,
      "grad_norm": 1.328125,
      "learning_rate": 0.000212,
      "loss": 6.2109,
      "mean_token_accuracy": 0.1485251747071743,
      "num_tokens": 970191.0,
      "step": 425
    },
    {
      "entropy": 6.59421329498291,
      "epoch": 0.04130643611911623,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002145,
      "loss": 6.349,
      "mean_token_accuracy": 0.14038845673203468,
      "num_tokens": 981811.0,
      "step": 430
    },
    {
      "entropy": 6.6880528926849365,
      "epoch": 0.04178674351585014,
      "grad_norm": 1.203125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.4534,
      "mean_token_accuracy": 0.13162965103983879,
      "num_tokens": 994034.0,
      "step": 435
    },
    {
      "entropy": 6.761573839187622,
      "epoch": 0.04226705091258405,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002195,
      "loss": 6.4095,
      "mean_token_accuracy": 0.13384866043925286,
      "num_tokens": 1005506.0,
      "step": 440
    },
    {
      "entropy": 6.527941131591797,
      "epoch": 0.042747358309317966,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000222,
      "loss": 6.3822,
      "mean_token_accuracy": 0.13079070374369622,
      "num_tokens": 1017103.0,
      "step": 445
    },
    {
      "entropy": 6.732952213287353,
      "epoch": 0.043227665706051875,
      "grad_norm": 1.125,
      "learning_rate": 0.0002245,
      "loss": 6.4624,
      "mean_token_accuracy": 0.13197413310408593,
      "num_tokens": 1029606.0,
      "step": 450
    },
    {
      "entropy": 6.717301893234253,
      "epoch": 0.043707973102785784,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.4124,
      "mean_token_accuracy": 0.13508057743310928,
      "num_tokens": 1041899.0,
      "step": 455
    },
    {
      "entropy": 6.644534492492676,
      "epoch": 0.04418828049951969,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.3181,
      "mean_token_accuracy": 0.14683899581432341,
      "num_tokens": 1053577.0,
      "step": 460
    },
    {
      "entropy": 6.72195086479187,
      "epoch": 0.0446685878962536,
      "grad_norm": 1.21875,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.3696,
      "mean_token_accuracy": 0.1393313579261303,
      "num_tokens": 1065541.0,
      "step": 465
    },
    {
      "entropy": 6.632449531555176,
      "epoch": 0.04514889529298751,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.3341,
      "mean_token_accuracy": 0.13726358041167258,
      "num_tokens": 1076138.0,
      "step": 470
    },
    {
      "entropy": 6.624440240859985,
      "epoch": 0.04562920268972142,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000237,
      "loss": 6.4125,
      "mean_token_accuracy": 0.138886184245348,
      "num_tokens": 1087620.0,
      "step": 475
    },
    {
      "entropy": 6.616486501693726,
      "epoch": 0.04610951008645533,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002395,
      "loss": 6.2776,
      "mean_token_accuracy": 0.14400001466274262,
      "num_tokens": 1099493.0,
      "step": 480
    },
    {
      "entropy": 6.6268257141113285,
      "epoch": 0.046589817483189244,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000242,
      "loss": 6.3667,
      "mean_token_accuracy": 0.1370286226272583,
      "num_tokens": 1110559.0,
      "step": 485
    },
    {
      "entropy": 6.628957080841064,
      "epoch": 0.04707012487992315,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002445,
      "loss": 6.3286,
      "mean_token_accuracy": 0.1350909538567066,
      "num_tokens": 1121051.0,
      "step": 490
    },
    {
      "entropy": 6.550091123580932,
      "epoch": 0.04755043227665706,
      "grad_norm": 1.25,
      "learning_rate": 0.000247,
      "loss": 6.2996,
      "mean_token_accuracy": 0.14171547815203667,
      "num_tokens": 1132305.0,
      "step": 495
    },
    {
      "entropy": 6.606603384017944,
      "epoch": 0.04803073967339097,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002495,
      "loss": 6.308,
      "mean_token_accuracy": 0.1442063570022583,
      "num_tokens": 1142979.0,
      "step": 500
    },
    {
      "entropy": 6.564784145355224,
      "epoch": 0.04851104707012488,
      "grad_norm": 1.421875,
      "learning_rate": 0.000252,
      "loss": 6.3125,
      "mean_token_accuracy": 0.14522673487663268,
      "num_tokens": 1153889.0,
      "step": 505
    },
    {
      "entropy": 6.613766813278199,
      "epoch": 0.04899135446685879,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002545,
      "loss": 6.3265,
      "mean_token_accuracy": 0.13703470006585122,
      "num_tokens": 1166030.0,
      "step": 510
    },
    {
      "entropy": 6.763839912414551,
      "epoch": 0.0494716618635927,
      "grad_norm": 1.28125,
      "learning_rate": 0.000257,
      "loss": 6.5003,
      "mean_token_accuracy": 0.13086057305336,
      "num_tokens": 1177545.0,
      "step": 515
    },
    {
      "entropy": 6.6000009059906,
      "epoch": 0.049951969260326606,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002595,
      "loss": 6.3245,
      "mean_token_accuracy": 0.14302466735243796,
      "num_tokens": 1188944.0,
      "step": 520
    },
    {
      "entropy": 6.479497623443604,
      "epoch": 0.05043227665706052,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000262,
      "loss": 6.3873,
      "mean_token_accuracy": 0.13586132526397704,
      "num_tokens": 1199517.0,
      "step": 525
    },
    {
      "entropy": 6.65042929649353,
      "epoch": 0.05091258405379443,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.3071,
      "mean_token_accuracy": 0.1335797443985939,
      "num_tokens": 1211536.0,
      "step": 530
    },
    {
      "entropy": 6.517299795150757,
      "epoch": 0.05139289145052834,
      "grad_norm": 1.28125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.2241,
      "mean_token_accuracy": 0.1449120409786701,
      "num_tokens": 1224159.0,
      "step": 535
    },
    {
      "entropy": 6.549662160873413,
      "epoch": 0.05187319884726225,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.3091,
      "mean_token_accuracy": 0.15078746303915977,
      "num_tokens": 1235226.0,
      "step": 540
    },
    {
      "entropy": 6.471032619476318,
      "epoch": 0.05235350624399616,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.2327,
      "mean_token_accuracy": 0.1502161517739296,
      "num_tokens": 1245862.0,
      "step": 545
    },
    {
      "entropy": 6.543156433105469,
      "epoch": 0.052833813640730067,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002745,
      "loss": 6.2496,
      "mean_token_accuracy": 0.14741843491792678,
      "num_tokens": 1257512.0,
      "step": 550
    },
    {
      "entropy": 6.532389879226685,
      "epoch": 0.053314121037463975,
      "grad_norm": 1.09375,
      "learning_rate": 0.000277,
      "loss": 6.3629,
      "mean_token_accuracy": 0.1384106993675232,
      "num_tokens": 1269317.0,
      "step": 555
    },
    {
      "entropy": 6.504622459411621,
      "epoch": 0.053794428434197884,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002795,
      "loss": 6.2464,
      "mean_token_accuracy": 0.13634237721562387,
      "num_tokens": 1281878.0,
      "step": 560
    },
    {
      "entropy": 6.503180837631225,
      "epoch": 0.05427473583093179,
      "grad_norm": 1.140625,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.257,
      "mean_token_accuracy": 0.14371064007282258,
      "num_tokens": 1293178.0,
      "step": 565
    },
    {
      "entropy": 6.489447402954101,
      "epoch": 0.05475504322766571,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002845,
      "loss": 6.2417,
      "mean_token_accuracy": 0.14116708934307098,
      "num_tokens": 1304969.0,
      "step": 570
    },
    {
      "entropy": 6.604181814193725,
      "epoch": 0.05523535062439962,
      "grad_norm": 1.265625,
      "learning_rate": 0.000287,
      "loss": 6.2788,
      "mean_token_accuracy": 0.14233512282371522,
      "num_tokens": 1317301.0,
      "step": 575
    },
    {
      "entropy": 6.477435731887818,
      "epoch": 0.05571565802113353,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002895,
      "loss": 6.265,
      "mean_token_accuracy": 0.1436339274048805,
      "num_tokens": 1327688.0,
      "step": 580
    },
    {
      "entropy": 6.550082349777222,
      "epoch": 0.056195965417867436,
      "grad_norm": 1.5,
      "learning_rate": 0.000292,
      "loss": 6.2146,
      "mean_token_accuracy": 0.1425333023071289,
      "num_tokens": 1338750.0,
      "step": 585
    },
    {
      "entropy": 6.4750957012176515,
      "epoch": 0.056676272814601344,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002945,
      "loss": 6.3745,
      "mean_token_accuracy": 0.12992137521505356,
      "num_tokens": 1350899.0,
      "step": 590
    },
    {
      "entropy": 6.533016681671143,
      "epoch": 0.05715658021133525,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000297,
      "loss": 6.2198,
      "mean_token_accuracy": 0.14716191366314887,
      "num_tokens": 1363184.0,
      "step": 595
    },
    {
      "entropy": 6.425290632247925,
      "epoch": 0.05763688760806916,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002995,
      "loss": 6.1952,
      "mean_token_accuracy": 0.1448550522327423,
      "num_tokens": 1374192.0,
      "step": 600
    },
    {
      "entropy": 6.497121429443359,
      "epoch": 0.05811719500480307,
      "grad_norm": 1.390625,
      "learning_rate": 0.000302,
      "loss": 6.2521,
      "mean_token_accuracy": 0.13990678563714026,
      "num_tokens": 1384888.0,
      "step": 605
    },
    {
      "entropy": 6.52073392868042,
      "epoch": 0.05859750240153699,
      "grad_norm": 1.375,
      "learning_rate": 0.0003045,
      "loss": 6.3405,
      "mean_token_accuracy": 0.14426726028323172,
      "num_tokens": 1395139.0,
      "step": 610
    },
    {
      "entropy": 6.423754501342773,
      "epoch": 0.059077809798270896,
      "grad_norm": 1.4375,
      "learning_rate": 0.000307,
      "loss": 6.1486,
      "mean_token_accuracy": 0.1441607266664505,
      "num_tokens": 1407219.0,
      "step": 615
    },
    {
      "entropy": 6.469784259796143,
      "epoch": 0.059558117195004805,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003095,
      "loss": 6.2751,
      "mean_token_accuracy": 0.13724448457360267,
      "num_tokens": 1418917.0,
      "step": 620
    },
    {
      "entropy": 6.500107717514038,
      "epoch": 0.060038424591738714,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000312,
      "loss": 6.2195,
      "mean_token_accuracy": 0.14356782287359238,
      "num_tokens": 1430792.0,
      "step": 625
    },
    {
      "entropy": 6.331269454956055,
      "epoch": 0.06051873198847262,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003145,
      "loss": 6.229,
      "mean_token_accuracy": 0.14871832877397537,
      "num_tokens": 1442825.0,
      "step": 630
    },
    {
      "entropy": 6.626306200027466,
      "epoch": 0.06099903938520653,
      "grad_norm": 1.390625,
      "learning_rate": 0.000317,
      "loss": 6.3337,
      "mean_token_accuracy": 0.14251203015446662,
      "num_tokens": 1454078.0,
      "step": 635
    },
    {
      "entropy": 6.469714879989624,
      "epoch": 0.06147934678194044,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003195,
      "loss": 6.2641,
      "mean_token_accuracy": 0.14170267581939697,
      "num_tokens": 1465164.0,
      "step": 640
    },
    {
      "entropy": 6.445518112182617,
      "epoch": 0.06195965417867435,
      "grad_norm": 1.140625,
      "learning_rate": 0.000322,
      "loss": 6.2758,
      "mean_token_accuracy": 0.1456916779279709,
      "num_tokens": 1476813.0,
      "step": 645
    },
    {
      "entropy": 6.426507139205933,
      "epoch": 0.06243996157540826,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.2858,
      "mean_token_accuracy": 0.14199284166097642,
      "num_tokens": 1488547.0,
      "step": 650
    },
    {
      "entropy": 6.549951553344727,
      "epoch": 0.06292026897214217,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.2862,
      "mean_token_accuracy": 0.13861191645264626,
      "num_tokens": 1500714.0,
      "step": 655
    },
    {
      "entropy": 6.4686461925506595,
      "epoch": 0.06340057636887608,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.271,
      "mean_token_accuracy": 0.14288542419672012,
      "num_tokens": 1512212.0,
      "step": 660
    },
    {
      "entropy": 6.412826299667358,
      "epoch": 0.06388088376560999,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.2458,
      "mean_token_accuracy": 0.14411089718341827,
      "num_tokens": 1523610.0,
      "step": 665
    },
    {
      "entropy": 6.428916597366333,
      "epoch": 0.0643611911623439,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.207,
      "mean_token_accuracy": 0.14545318782329558,
      "num_tokens": 1535543.0,
      "step": 670
    },
    {
      "entropy": 6.586844778060913,
      "epoch": 0.06484149855907781,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000337,
      "loss": 6.4396,
      "mean_token_accuracy": 0.1306020125746727,
      "num_tokens": 1547183.0,
      "step": 675
    },
    {
      "entropy": 6.379982614517212,
      "epoch": 0.06532180595581172,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003395,
      "loss": 6.201,
      "mean_token_accuracy": 0.1453273430466652,
      "num_tokens": 1559265.0,
      "step": 680
    },
    {
      "entropy": 6.4871210098266605,
      "epoch": 0.06580211335254563,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000342,
      "loss": 6.2149,
      "mean_token_accuracy": 0.1423127330839634,
      "num_tokens": 1569774.0,
      "step": 685
    },
    {
      "entropy": 6.295485496520996,
      "epoch": 0.06628242074927954,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.1273,
      "mean_token_accuracy": 0.14571069031953812,
      "num_tokens": 1579408.0,
      "step": 690
    },
    {
      "entropy": 6.419068622589111,
      "epoch": 0.06676272814601344,
      "grad_norm": 1.265625,
      "learning_rate": 0.000347,
      "loss": 6.2013,
      "mean_token_accuracy": 0.14634275287389756,
      "num_tokens": 1591849.0,
      "step": 695
    },
    {
      "entropy": 6.380165672302246,
      "epoch": 0.06724303554274735,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003495,
      "loss": 6.2564,
      "mean_token_accuracy": 0.1449852168560028,
      "num_tokens": 1603269.0,
      "step": 700
    },
    {
      "entropy": 6.5532173156738285,
      "epoch": 0.06772334293948126,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000352,
      "loss": 6.2875,
      "mean_token_accuracy": 0.14029174968600272,
      "num_tokens": 1613639.0,
      "step": 705
    },
    {
      "entropy": 6.4776379585266115,
      "epoch": 0.06820365033621517,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003545,
      "loss": 6.3104,
      "mean_token_accuracy": 0.13592585399746895,
      "num_tokens": 1625393.0,
      "step": 710
    },
    {
      "entropy": 6.359980726242066,
      "epoch": 0.0686839577329491,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000357,
      "loss": 6.154,
      "mean_token_accuracy": 0.14281153082847595,
      "num_tokens": 1636355.0,
      "step": 715
    },
    {
      "entropy": 6.46089243888855,
      "epoch": 0.069164265129683,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003595,
      "loss": 6.2008,
      "mean_token_accuracy": 0.14487438052892684,
      "num_tokens": 1646157.0,
      "step": 720
    },
    {
      "entropy": 6.30911774635315,
      "epoch": 0.06964457252641691,
      "grad_norm": 1.34375,
      "learning_rate": 0.000362,
      "loss": 6.1967,
      "mean_token_accuracy": 0.1431838683784008,
      "num_tokens": 1657827.0,
      "step": 725
    },
    {
      "entropy": 6.406947374343872,
      "epoch": 0.07012487992315082,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003645,
      "loss": 6.2665,
      "mean_token_accuracy": 0.14254478514194488,
      "num_tokens": 1669023.0,
      "step": 730
    },
    {
      "entropy": 6.439149236679077,
      "epoch": 0.07060518731988473,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000367,
      "loss": 6.2449,
      "mean_token_accuracy": 0.1420585885643959,
      "num_tokens": 1680048.0,
      "step": 735
    },
    {
      "entropy": 6.492261695861816,
      "epoch": 0.07108549471661864,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003695,
      "loss": 6.4099,
      "mean_token_accuracy": 0.13245131075382233,
      "num_tokens": 1690907.0,
      "step": 740
    },
    {
      "entropy": 6.504529047012329,
      "epoch": 0.07156580211335255,
      "grad_norm": 1.296875,
      "learning_rate": 0.000372,
      "loss": 6.2452,
      "mean_token_accuracy": 0.14205930680036544,
      "num_tokens": 1702337.0,
      "step": 745
    },
    {
      "entropy": 6.37948842048645,
      "epoch": 0.07204610951008646,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003745,
      "loss": 6.1913,
      "mean_token_accuracy": 0.14261919558048247,
      "num_tokens": 1713242.0,
      "step": 750
    },
    {
      "entropy": 6.319987010955811,
      "epoch": 0.07252641690682037,
      "grad_norm": 1.4375,
      "learning_rate": 0.000377,
      "loss": 6.1894,
      "mean_token_accuracy": 0.14029277563095094,
      "num_tokens": 1724939.0,
      "step": 755
    },
    {
      "entropy": 6.339571523666382,
      "epoch": 0.07300672430355427,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003795,
      "loss": 6.1217,
      "mean_token_accuracy": 0.15060591399669648,
      "num_tokens": 1737311.0,
      "step": 760
    },
    {
      "entropy": 6.433168601989746,
      "epoch": 0.07348703170028818,
      "grad_norm": 1.25,
      "learning_rate": 0.000382,
      "loss": 6.2744,
      "mean_token_accuracy": 0.14171026945114135,
      "num_tokens": 1749528.0,
      "step": 765
    },
    {
      "entropy": 6.279998445510865,
      "epoch": 0.07396733909702209,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003845,
      "loss": 6.1478,
      "mean_token_accuracy": 0.14513302072882653,
      "num_tokens": 1760304.0,
      "step": 770
    },
    {
      "entropy": 6.366886043548584,
      "epoch": 0.074447646493756,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.1265,
      "mean_token_accuracy": 0.14714072048664092,
      "num_tokens": 1772014.0,
      "step": 775
    },
    {
      "entropy": 6.309161758422851,
      "epoch": 0.07492795389048991,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.1815,
      "mean_token_accuracy": 0.13897588327527047,
      "num_tokens": 1784585.0,
      "step": 780
    },
    {
      "entropy": 6.3078656673431395,
      "epoch": 0.07540826128722382,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.1274,
      "mean_token_accuracy": 0.14889669567346572,
      "num_tokens": 1795246.0,
      "step": 785
    },
    {
      "entropy": 6.25552887916565,
      "epoch": 0.07588856868395773,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.1005,
      "mean_token_accuracy": 0.15239656567573548,
      "num_tokens": 1806826.0,
      "step": 790
    },
    {
      "entropy": 6.338869380950928,
      "epoch": 0.07636887608069164,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.2446,
      "mean_token_accuracy": 0.14244618490338326,
      "num_tokens": 1818766.0,
      "step": 795
    },
    {
      "entropy": 6.325174713134766,
      "epoch": 0.07684918347742556,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003995,
      "loss": 6.1009,
      "mean_token_accuracy": 0.155482742190361,
      "num_tokens": 1830824.0,
      "step": 800
    },
    {
      "entropy": 6.289180946350098,
      "epoch": 0.07732949087415947,
      "grad_norm": 1.21875,
      "learning_rate": 0.000402,
      "loss": 6.1829,
      "mean_token_accuracy": 0.14374645203351974,
      "num_tokens": 1841792.0,
      "step": 805
    },
    {
      "entropy": 6.302430200576782,
      "epoch": 0.07780979827089338,
      "grad_norm": 1.125,
      "learning_rate": 0.0004045,
      "loss": 6.1559,
      "mean_token_accuracy": 0.14021549075841905,
      "num_tokens": 1854683.0,
      "step": 810
    },
    {
      "entropy": 6.337079668045044,
      "epoch": 0.07829010566762729,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.2478,
      "mean_token_accuracy": 0.14183034971356392,
      "num_tokens": 1865826.0,
      "step": 815
    },
    {
      "entropy": 6.36335711479187,
      "epoch": 0.0787704130643612,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004095,
      "loss": 6.173,
      "mean_token_accuracy": 0.1493820831179619,
      "num_tokens": 1876759.0,
      "step": 820
    },
    {
      "entropy": 6.290498876571656,
      "epoch": 0.0792507204610951,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000412,
      "loss": 6.222,
      "mean_token_accuracy": 0.14272527173161506,
      "num_tokens": 1888013.0,
      "step": 825
    },
    {
      "entropy": 6.415748977661133,
      "epoch": 0.07973102785782901,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004145,
      "loss": 6.2222,
      "mean_token_accuracy": 0.14676567614078523,
      "num_tokens": 1898964.0,
      "step": 830
    },
    {
      "entropy": 6.28214316368103,
      "epoch": 0.08021133525456292,
      "grad_norm": 1.28125,
      "learning_rate": 0.000417,
      "loss": 6.149,
      "mean_token_accuracy": 0.1509616807103157,
      "num_tokens": 1909728.0,
      "step": 835
    },
    {
      "entropy": 6.214513111114502,
      "epoch": 0.08069164265129683,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004195,
      "loss": 6.0757,
      "mean_token_accuracy": 0.15560405552387238,
      "num_tokens": 1919895.0,
      "step": 840
    },
    {
      "entropy": 6.2361761093139645,
      "epoch": 0.08117195004803074,
      "grad_norm": 1.25,
      "learning_rate": 0.000422,
      "loss": 6.0097,
      "mean_token_accuracy": 0.1526105895638466,
      "num_tokens": 1930839.0,
      "step": 845
    },
    {
      "entropy": 6.217610692977905,
      "epoch": 0.08165225744476465,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004245,
      "loss": 6.2137,
      "mean_token_accuracy": 0.14339369758963585,
      "num_tokens": 1942605.0,
      "step": 850
    },
    {
      "entropy": 6.317978858947754,
      "epoch": 0.08213256484149856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000427,
      "loss": 6.2029,
      "mean_token_accuracy": 0.13737519904971124,
      "num_tokens": 1954863.0,
      "step": 855
    },
    {
      "entropy": 6.382518625259399,
      "epoch": 0.08261287223823247,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004295,
      "loss": 6.2852,
      "mean_token_accuracy": 0.1441127136349678,
      "num_tokens": 1966088.0,
      "step": 860
    },
    {
      "entropy": 6.296528148651123,
      "epoch": 0.08309317963496637,
      "grad_norm": 1.21875,
      "learning_rate": 0.000432,
      "loss": 6.1112,
      "mean_token_accuracy": 0.14490053951740264,
      "num_tokens": 1978081.0,
      "step": 865
    },
    {
      "entropy": 6.272141361236573,
      "epoch": 0.08357348703170028,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004345,
      "loss": 6.1827,
      "mean_token_accuracy": 0.13639964535832405,
      "num_tokens": 1991212.0,
      "step": 870
    },
    {
      "entropy": 6.341372632980347,
      "epoch": 0.08405379442843419,
      "grad_norm": 1.28125,
      "learning_rate": 0.000437,
      "loss": 6.1709,
      "mean_token_accuracy": 0.14349714294075966,
      "num_tokens": 2002074.0,
      "step": 875
    },
    {
      "entropy": 6.2472961902618405,
      "epoch": 0.0845341018251681,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004395,
      "loss": 6.0965,
      "mean_token_accuracy": 0.15204162746667862,
      "num_tokens": 2013039.0,
      "step": 880
    },
    {
      "entropy": 6.304888582229614,
      "epoch": 0.08501440922190202,
      "grad_norm": 1.03125,
      "learning_rate": 0.000442,
      "loss": 6.2385,
      "mean_token_accuracy": 0.1450421005487442,
      "num_tokens": 2024744.0,
      "step": 885
    },
    {
      "entropy": 6.2084990501403805,
      "epoch": 0.08549471661863593,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004445,
      "loss": 6.108,
      "mean_token_accuracy": 0.15128371939063073,
      "num_tokens": 2035587.0,
      "step": 890
    },
    {
      "entropy": 6.304928970336914,
      "epoch": 0.08597502401536984,
      "grad_norm": 1.140625,
      "learning_rate": 0.000447,
      "loss": 6.1802,
      "mean_token_accuracy": 0.14263538867235184,
      "num_tokens": 2047924.0,
      "step": 895
    },
    {
      "entropy": 6.22371997833252,
      "epoch": 0.08645533141210375,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.126,
      "mean_token_accuracy": 0.14922973960638047,
      "num_tokens": 2060161.0,
      "step": 900
    },
    {
      "entropy": 6.24933967590332,
      "epoch": 0.08693563880883766,
      "grad_norm": 1.25,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.1122,
      "mean_token_accuracy": 0.14982035011053085,
      "num_tokens": 2071674.0,
      "step": 905
    },
    {
      "entropy": 6.230753231048584,
      "epoch": 0.08741594620557157,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.2656,
      "mean_token_accuracy": 0.14044685065746307,
      "num_tokens": 2084025.0,
      "step": 910
    },
    {
      "entropy": 6.335498666763305,
      "epoch": 0.08789625360230548,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.2135,
      "mean_token_accuracy": 0.14406091794371606,
      "num_tokens": 2094501.0,
      "step": 915
    },
    {
      "entropy": 6.295937156677246,
      "epoch": 0.08837656099903939,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.1786,
      "mean_token_accuracy": 0.15006129145622255,
      "num_tokens": 2106207.0,
      "step": 920
    },
    {
      "entropy": 6.271474409103393,
      "epoch": 0.0888568683957733,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000462,
      "loss": 6.0837,
      "mean_token_accuracy": 0.15673873275518418,
      "num_tokens": 2116879.0,
      "step": 925
    },
    {
      "entropy": 6.162377500534058,
      "epoch": 0.0893371757925072,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004645,
      "loss": 6.1608,
      "mean_token_accuracy": 0.14606007188558578,
      "num_tokens": 2128884.0,
      "step": 930
    },
    {
      "entropy": 6.210721969604492,
      "epoch": 0.08981748318924111,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000467,
      "loss": 6.1085,
      "mean_token_accuracy": 0.14014682546257973,
      "num_tokens": 2140810.0,
      "step": 935
    },
    {
      "entropy": 6.217668056488037,
      "epoch": 0.09029779058597502,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004695,
      "loss": 6.141,
      "mean_token_accuracy": 0.14254814609885216,
      "num_tokens": 2152312.0,
      "step": 940
    },
    {
      "entropy": 6.211342811584473,
      "epoch": 0.09077809798270893,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000472,
      "loss": 6.0985,
      "mean_token_accuracy": 0.144034481048584,
      "num_tokens": 2164408.0,
      "step": 945
    },
    {
      "entropy": 6.172352027893067,
      "epoch": 0.09125840537944284,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004745,
      "loss": 6.1289,
      "mean_token_accuracy": 0.1452830523252487,
      "num_tokens": 2176050.0,
      "step": 950
    },
    {
      "entropy": 6.114393424987793,
      "epoch": 0.09173871277617675,
      "grad_norm": 1.0625,
      "learning_rate": 0.000477,
      "loss": 5.9983,
      "mean_token_accuracy": 0.15028768330812453,
      "num_tokens": 2187354.0,
      "step": 955
    },
    {
      "entropy": 6.046187686920166,
      "epoch": 0.09221902017291066,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004795,
      "loss": 6.0429,
      "mean_token_accuracy": 0.14834860116243362,
      "num_tokens": 2198686.0,
      "step": 960
    },
    {
      "entropy": 6.162608814239502,
      "epoch": 0.09269932756964457,
      "grad_norm": 1.21875,
      "learning_rate": 0.000482,
      "loss": 6.0594,
      "mean_token_accuracy": 0.15012447834014891,
      "num_tokens": 2209780.0,
      "step": 965
    },
    {
      "entropy": 6.187800312042237,
      "epoch": 0.09317963496637849,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004845,
      "loss": 6.1874,
      "mean_token_accuracy": 0.1498972252011299,
      "num_tokens": 2221760.0,
      "step": 970
    },
    {
      "entropy": 6.147322368621826,
      "epoch": 0.0936599423631124,
      "grad_norm": 1.1875,
      "learning_rate": 0.000487,
      "loss": 6.1852,
      "mean_token_accuracy": 0.14224128648638726,
      "num_tokens": 2233767.0,
      "step": 975
    },
    {
      "entropy": 6.226958322525024,
      "epoch": 0.0941402497598463,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004895,
      "loss": 6.0957,
      "mean_token_accuracy": 0.14374661892652513,
      "num_tokens": 2244841.0,
      "step": 980
    },
    {
      "entropy": 6.09796028137207,
      "epoch": 0.09462055715658022,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000492,
      "loss": 5.9658,
      "mean_token_accuracy": 0.1552761510014534,
      "num_tokens": 2255620.0,
      "step": 985
    },
    {
      "entropy": 6.203439474105835,
      "epoch": 0.09510086455331412,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004945,
      "loss": 6.1626,
      "mean_token_accuracy": 0.14942158684134482,
      "num_tokens": 2266345.0,
      "step": 990
    },
    {
      "entropy": 6.17238416671753,
      "epoch": 0.09558117195004803,
      "grad_norm": 1.34375,
      "learning_rate": 0.000497,
      "loss": 6.0839,
      "mean_token_accuracy": 0.15113430023193358,
      "num_tokens": 2277190.0,
      "step": 995
    },
    {
      "entropy": 6.099387550354004,
      "epoch": 0.09606147934678194,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995,
      "loss": 5.9749,
      "mean_token_accuracy": 0.15104159116744995,
      "num_tokens": 2288241.0,
      "step": 1000
    },
    {
      "entropy": 6.172607707977295,
      "epoch": 0.09654178674351585,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999999983283737,
      "loss": 6.1383,
      "mean_token_accuracy": 0.14582284688949584,
      "num_tokens": 2301092.0,
      "step": 1005
    },
    {
      "entropy": 6.113104200363159,
      "epoch": 0.09702209414024976,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999999915373924,
      "loss": 6.0215,
      "mean_token_accuracy": 0.1486254870891571,
      "num_tokens": 2313622.0,
      "step": 1010
    },
    {
      "entropy": 6.008445453643799,
      "epoch": 0.09750240153698367,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999795225793,
      "loss": 5.9944,
      "mean_token_accuracy": 0.1521578386425972,
      "num_tokens": 2325699.0,
      "step": 1015
    },
    {
      "entropy": 6.20647497177124,
      "epoch": 0.09798270893371758,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999999622839347,
      "loss": 6.0653,
      "mean_token_accuracy": 0.15288252979516984,
      "num_tokens": 2337816.0,
      "step": 1020
    },
    {
      "entropy": 6.215051984786987,
      "epoch": 0.09846301633045149,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999999398214593,
      "loss": 6.1867,
      "mean_token_accuracy": 0.14325481206178664,
      "num_tokens": 2348793.0,
      "step": 1025
    },
    {
      "entropy": 6.149576854705811,
      "epoch": 0.0989433237271854,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999999121351532,
      "loss": 6.1092,
      "mean_token_accuracy": 0.1484948292374611,
      "num_tokens": 2360209.0,
      "step": 1030
    },
    {
      "entropy": 6.243153047561646,
      "epoch": 0.0994236311239193,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999998792250173,
      "loss": 6.1044,
      "mean_token_accuracy": 0.14303301200270652,
      "num_tokens": 2370597.0,
      "step": 1035
    },
    {
      "entropy": 6.169728374481201,
      "epoch": 0.09990393852065321,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999998410910524,
      "loss": 6.1585,
      "mean_token_accuracy": 0.14206091016530992,
      "num_tokens": 2381466.0,
      "step": 1040
    },
    {
      "entropy": 6.2217827320098875,
      "epoch": 0.10038424591738712,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999997977332592,
      "loss": 6.2478,
      "mean_token_accuracy": 0.1373165175318718,
      "num_tokens": 2394507.0,
      "step": 1045
    },
    {
      "entropy": 6.160476970672607,
      "epoch": 0.10086455331412104,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999997491516389,
      "loss": 6.0899,
      "mean_token_accuracy": 0.1465790644288063,
      "num_tokens": 2406367.0,
      "step": 1050
    },
    {
      "entropy": 6.151379632949829,
      "epoch": 0.10134486071085495,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999996953461925,
      "loss": 6.0559,
      "mean_token_accuracy": 0.1507312148809433,
      "num_tokens": 2416962.0,
      "step": 1055
    },
    {
      "entropy": 6.161687278747559,
      "epoch": 0.10182516810758886,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999996363169212,
      "loss": 6.1202,
      "mean_token_accuracy": 0.14120108187198638,
      "num_tokens": 2429441.0,
      "step": 1060
    },
    {
      "entropy": 6.095824718475342,
      "epoch": 0.10230547550432277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999995720638266,
      "loss": 6.0097,
      "mean_token_accuracy": 0.15365689769387245,
      "num_tokens": 2441360.0,
      "step": 1065
    },
    {
      "entropy": 6.10391526222229,
      "epoch": 0.10278578290105668,
      "grad_norm": 1.15625,
      "learning_rate": 0.00049999950258691,
      "loss": 6.0177,
      "mean_token_accuracy": 0.14536169916391373,
      "num_tokens": 2452928.0,
      "step": 1070
    },
    {
      "entropy": 6.02531418800354,
      "epoch": 0.10326609029779059,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999994278861731,
      "loss": 5.9698,
      "mean_token_accuracy": 0.14933416843414307,
      "num_tokens": 2464218.0,
      "step": 1075
    },
    {
      "entropy": 6.104892635345459,
      "epoch": 0.1037463976945245,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999993479616175,
      "loss": 6.0582,
      "mean_token_accuracy": 0.15327436327934266,
      "num_tokens": 2475084.0,
      "step": 1080
    },
    {
      "entropy": 6.1318131446838375,
      "epoch": 0.1042267050912584,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999992628132451,
      "loss": 6.0505,
      "mean_token_accuracy": 0.1454383134841919,
      "num_tokens": 2486845.0,
      "step": 1085
    },
    {
      "entropy": 6.044811820983886,
      "epoch": 0.10470701248799232,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999991724410582,
      "loss": 6.1225,
      "mean_token_accuracy": 0.14644298553466797,
      "num_tokens": 2498049.0,
      "step": 1090
    },
    {
      "entropy": 6.118540906906128,
      "epoch": 0.10518731988472622,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999990768450583,
      "loss": 6.0521,
      "mean_token_accuracy": 0.14827468991279602,
      "num_tokens": 2509511.0,
      "step": 1095
    },
    {
      "entropy": 6.200696563720703,
      "epoch": 0.10566762728146013,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999989760252482,
      "loss": 6.0115,
      "mean_token_accuracy": 0.15714994966983795,
      "num_tokens": 2521518.0,
      "step": 1100
    },
    {
      "entropy": 5.881813478469849,
      "epoch": 0.10614793467819404,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999988699816299,
      "loss": 5.8771,
      "mean_token_accuracy": 0.15803790241479873,
      "num_tokens": 2533141.0,
      "step": 1105
    },
    {
      "entropy": 5.976910018920899,
      "epoch": 0.10662824207492795,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999987587142058,
      "loss": 5.8731,
      "mean_token_accuracy": 0.15562851428985597,
      "num_tokens": 2542907.0,
      "step": 1110
    },
    {
      "entropy": 6.015887451171875,
      "epoch": 0.10710854947166186,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999986422229789,
      "loss": 6.0172,
      "mean_token_accuracy": 0.1435399830341339,
      "num_tokens": 2554497.0,
      "step": 1115
    },
    {
      "entropy": 6.222254037857056,
      "epoch": 0.10758885686839577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999985205079514,
      "loss": 6.1419,
      "mean_token_accuracy": 0.15260767042636872,
      "num_tokens": 2567035.0,
      "step": 1120
    },
    {
      "entropy": 5.941625452041626,
      "epoch": 0.10806916426512968,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999983935691265,
      "loss": 5.9125,
      "mean_token_accuracy": 0.16007250249385835,
      "num_tokens": 2578003.0,
      "step": 1125
    },
    {
      "entropy": 6.222092819213867,
      "epoch": 0.10854947166186359,
      "grad_norm": 1.1875,
      "learning_rate": 0.000499998261406507,
      "loss": 6.2379,
      "mean_token_accuracy": 0.14466361775994302,
      "num_tokens": 2590272.0,
      "step": 1130
    },
    {
      "entropy": 5.978817510604858,
      "epoch": 0.10902977905859751,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999981240200958,
      "loss": 6.0431,
      "mean_token_accuracy": 0.14697342813014985,
      "num_tokens": 2602874.0,
      "step": 1135
    },
    {
      "entropy": 6.120553874969483,
      "epoch": 0.10951008645533142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999979814098966,
      "loss": 5.9901,
      "mean_token_accuracy": 0.15932974964380264,
      "num_tokens": 2613835.0,
      "step": 1140
    },
    {
      "entropy": 6.091197299957275,
      "epoch": 0.10999039385206533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999978335759121,
      "loss": 6.1007,
      "mean_token_accuracy": 0.14361643120646478,
      "num_tokens": 2624986.0,
      "step": 1145
    },
    {
      "entropy": 6.129514551162719,
      "epoch": 0.11047070124879924,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999976805181461,
      "loss": 6.1063,
      "mean_token_accuracy": 0.14257665798068048,
      "num_tokens": 2637654.0,
      "step": 1150
    },
    {
      "entropy": 6.070494890213013,
      "epoch": 0.11095100864553314,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499997522236602,
      "loss": 6.0191,
      "mean_token_accuracy": 0.14936878979206086,
      "num_tokens": 2648947.0,
      "step": 1155
    },
    {
      "entropy": 6.050145816802979,
      "epoch": 0.11143131604226705,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999973587312837,
      "loss": 6.0003,
      "mean_token_accuracy": 0.15421223491430283,
      "num_tokens": 2659573.0,
      "step": 1160
    },
    {
      "entropy": 5.9971983432769775,
      "epoch": 0.11191162343900096,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999971900021947,
      "loss": 6.0016,
      "mean_token_accuracy": 0.1543707251548767,
      "num_tokens": 2671491.0,
      "step": 1165
    },
    {
      "entropy": 6.009974718093872,
      "epoch": 0.11239193083573487,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999970160493391,
      "loss": 5.9746,
      "mean_token_accuracy": 0.1493075892329216,
      "num_tokens": 2682705.0,
      "step": 1170
    },
    {
      "entropy": 5.976160669326783,
      "epoch": 0.11287223823246878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999968368727209,
      "loss": 5.9202,
      "mean_token_accuracy": 0.15478333532810212,
      "num_tokens": 2693539.0,
      "step": 1175
    },
    {
      "entropy": 6.010011291503906,
      "epoch": 0.11335254562920269,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999966524723442,
      "loss": 5.9729,
      "mean_token_accuracy": 0.15669194161891936,
      "num_tokens": 2705403.0,
      "step": 1180
    },
    {
      "entropy": 6.115589570999146,
      "epoch": 0.1138328530259366,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999964628482135,
      "loss": 6.0684,
      "mean_token_accuracy": 0.14865557104349136,
      "num_tokens": 2717320.0,
      "step": 1185
    },
    {
      "entropy": 6.025648975372315,
      "epoch": 0.1143131604226705,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999962680003328,
      "loss": 6.0295,
      "mean_token_accuracy": 0.14510569497942924,
      "num_tokens": 2730056.0,
      "step": 1190
    },
    {
      "entropy": 6.009522199630737,
      "epoch": 0.11479346781940442,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499996067928707,
      "loss": 5.8953,
      "mean_token_accuracy": 0.15695455223321914,
      "num_tokens": 2741399.0,
      "step": 1195
    },
    {
      "entropy": 6.138769435882568,
      "epoch": 0.11527377521613832,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999958626333406,
      "loss": 6.1321,
      "mean_token_accuracy": 0.14366149455308913,
      "num_tokens": 2752277.0,
      "step": 1200
    },
    {
      "entropy": 6.120853424072266,
      "epoch": 0.11575408261287223,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999956521142383,
      "loss": 6.0335,
      "mean_token_accuracy": 0.14938354790210723,
      "num_tokens": 2763172.0,
      "step": 1205
    },
    {
      "entropy": 6.033753395080566,
      "epoch": 0.11623439000960614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999954363714051,
      "loss": 6.0339,
      "mean_token_accuracy": 0.1517849951982498,
      "num_tokens": 2775096.0,
      "step": 1210
    },
    {
      "entropy": 6.113629102706909,
      "epoch": 0.11671469740634005,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999952154048459,
      "loss": 6.1204,
      "mean_token_accuracy": 0.14458551853895188,
      "num_tokens": 2786007.0,
      "step": 1215
    },
    {
      "entropy": 6.020708131790161,
      "epoch": 0.11719500480307397,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499994989214566,
      "loss": 5.93,
      "mean_token_accuracy": 0.15760562121868132,
      "num_tokens": 2797265.0,
      "step": 1220
    },
    {
      "entropy": 6.061748027801514,
      "epoch": 0.11767531219980788,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999947578005705,
      "loss": 6.079,
      "mean_token_accuracy": 0.15117940902709961,
      "num_tokens": 2809427.0,
      "step": 1225
    },
    {
      "entropy": 6.040156745910645,
      "epoch": 0.11815561959654179,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999945211628648,
      "loss": 6.0195,
      "mean_token_accuracy": 0.15237225145101546,
      "num_tokens": 2822880.0,
      "step": 1230
    },
    {
      "entropy": 6.100583505630493,
      "epoch": 0.1186359269932757,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999942793014544,
      "loss": 6.0854,
      "mean_token_accuracy": 0.14831554368138314,
      "num_tokens": 2834760.0,
      "step": 1235
    },
    {
      "entropy": 6.070679807662964,
      "epoch": 0.11911623439000961,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499994032216345,
      "loss": 6.0407,
      "mean_token_accuracy": 0.1506760910153389,
      "num_tokens": 2846517.0,
      "step": 1240
    },
    {
      "entropy": 5.970952844619751,
      "epoch": 0.11959654178674352,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999937799075422,
      "loss": 5.918,
      "mean_token_accuracy": 0.15903548300266265,
      "num_tokens": 2857280.0,
      "step": 1245
    },
    {
      "entropy": 6.055624532699585,
      "epoch": 0.12007684918347743,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499993522375052,
      "loss": 5.9884,
      "mean_token_accuracy": 0.15432048588991165,
      "num_tokens": 2869735.0,
      "step": 1250
    },
    {
      "entropy": 5.979946804046631,
      "epoch": 0.12055715658021134,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999932596188802,
      "loss": 5.9782,
      "mean_token_accuracy": 0.1547883614897728,
      "num_tokens": 2881681.0,
      "step": 1255
    },
    {
      "entropy": 5.934148931503296,
      "epoch": 0.12103746397694524,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999929916390331,
      "loss": 5.9816,
      "mean_token_accuracy": 0.15643564164638518,
      "num_tokens": 2893828.0,
      "step": 1260
    },
    {
      "entropy": 5.921392345428467,
      "epoch": 0.12151777137367915,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999927184355169,
      "loss": 5.9543,
      "mean_token_accuracy": 0.15064309760928155,
      "num_tokens": 2905804.0,
      "step": 1265
    },
    {
      "entropy": 6.214194297790527,
      "epoch": 0.12199807877041306,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999924400083377,
      "loss": 6.0906,
      "mean_token_accuracy": 0.1528476931154728,
      "num_tokens": 2918709.0,
      "step": 1270
    },
    {
      "entropy": 6.012447738647461,
      "epoch": 0.12247838616714697,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999921563575022,
      "loss": 5.9939,
      "mean_token_accuracy": 0.15287727490067482,
      "num_tokens": 2930589.0,
      "step": 1275
    },
    {
      "entropy": 6.041652250289917,
      "epoch": 0.12295869356388088,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999918674830169,
      "loss": 6.0472,
      "mean_token_accuracy": 0.15353401154279708,
      "num_tokens": 2940635.0,
      "step": 1280
    },
    {
      "entropy": 5.93117880821228,
      "epoch": 0.12343900096061479,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999915733848886,
      "loss": 5.9315,
      "mean_token_accuracy": 0.15750601142644882,
      "num_tokens": 2952731.0,
      "step": 1285
    },
    {
      "entropy": 5.993732738494873,
      "epoch": 0.1239193083573487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499991274063124,
      "loss": 5.9264,
      "mean_token_accuracy": 0.15445593893527984,
      "num_tokens": 2964544.0,
      "step": 1290
    },
    {
      "entropy": 5.881064367294312,
      "epoch": 0.12439961575408261,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999909695177301,
      "loss": 5.8655,
      "mean_token_accuracy": 0.15871210992336274,
      "num_tokens": 2974996.0,
      "step": 1295
    },
    {
      "entropy": 5.977795648574829,
      "epoch": 0.12487992315081652,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499990659748714,
      "loss": 5.8809,
      "mean_token_accuracy": 0.15484783351421355,
      "num_tokens": 2986496.0,
      "step": 1300
    },
    {
      "entropy": 5.925467538833618,
      "epoch": 0.12536023054755044,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999903447560828,
      "loss": 5.9557,
      "mean_token_accuracy": 0.15500155389308928,
      "num_tokens": 2997693.0,
      "step": 1305
    },
    {
      "entropy": 5.94029974937439,
      "epoch": 0.12584053794428435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999900245398439,
      "loss": 5.9248,
      "mean_token_accuracy": 0.15493305549025535,
      "num_tokens": 3009155.0,
      "step": 1310
    },
    {
      "entropy": 5.945185422897339,
      "epoch": 0.12632084534101826,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999896991000047,
      "loss": 5.8471,
      "mean_token_accuracy": 0.16124534010887145,
      "num_tokens": 3020944.0,
      "step": 1315
    },
    {
      "entropy": 5.93646125793457,
      "epoch": 0.12680115273775217,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999893684365729,
      "loss": 5.9099,
      "mean_token_accuracy": 0.15695340186357498,
      "num_tokens": 3032893.0,
      "step": 1320
    },
    {
      "entropy": 5.944838190078736,
      "epoch": 0.12728146013448607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999890325495559,
      "loss": 5.9864,
      "mean_token_accuracy": 0.15219316333532334,
      "num_tokens": 3044937.0,
      "step": 1325
    },
    {
      "entropy": 5.944310712814331,
      "epoch": 0.12776176753121998,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999886914389617,
      "loss": 5.8678,
      "mean_token_accuracy": 0.1587807759642601,
      "num_tokens": 3056071.0,
      "step": 1330
    },
    {
      "entropy": 5.932201910018921,
      "epoch": 0.1282420749279539,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004999883451047981,
      "loss": 6.0432,
      "mean_token_accuracy": 0.15118279233574866,
      "num_tokens": 3067724.0,
      "step": 1335
    },
    {
      "entropy": 6.02701096534729,
      "epoch": 0.1287223823246878,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999879935470733,
      "loss": 5.9518,
      "mean_token_accuracy": 0.158474200963974,
      "num_tokens": 3080779.0,
      "step": 1340
    },
    {
      "entropy": 5.9539937496185305,
      "epoch": 0.1292026897214217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999876367657954,
      "loss": 5.9255,
      "mean_token_accuracy": 0.15368285030126572,
      "num_tokens": 3091991.0,
      "step": 1345
    },
    {
      "entropy": 5.926160478591919,
      "epoch": 0.12968299711815562,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999872747609725,
      "loss": 5.9166,
      "mean_token_accuracy": 0.15522892773151398,
      "num_tokens": 3103713.0,
      "step": 1350
    },
    {
      "entropy": 5.925470161437988,
      "epoch": 0.13016330451488953,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999869075326132,
      "loss": 5.948,
      "mean_token_accuracy": 0.1540072813630104,
      "num_tokens": 3115891.0,
      "step": 1355
    },
    {
      "entropy": 5.85691990852356,
      "epoch": 0.13064361191162344,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499986535080726,
      "loss": 5.8039,
      "mean_token_accuracy": 0.16470443606376647,
      "num_tokens": 3126223.0,
      "step": 1360
    },
    {
      "entropy": 6.008657360076905,
      "epoch": 0.13112391930835735,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999861574053196,
      "loss": 5.9948,
      "mean_token_accuracy": 0.151066617667675,
      "num_tokens": 3137537.0,
      "step": 1365
    },
    {
      "entropy": 6.027498769760132,
      "epoch": 0.13160422670509125,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999857745064027,
      "loss": 5.9965,
      "mean_token_accuracy": 0.15381778329610823,
      "num_tokens": 3150126.0,
      "step": 1370
    },
    {
      "entropy": 5.952863073348999,
      "epoch": 0.13208453410182516,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499985386383984,
      "loss": 5.9663,
      "mean_token_accuracy": 0.1613561525940895,
      "num_tokens": 3161684.0,
      "step": 1375
    },
    {
      "entropy": 5.872996187210083,
      "epoch": 0.13256484149855907,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999849930380729,
      "loss": 5.8446,
      "mean_token_accuracy": 0.1591193348169327,
      "num_tokens": 3172342.0,
      "step": 1380
    },
    {
      "entropy": 6.032765531539917,
      "epoch": 0.13304514889529298,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999845944686781,
      "loss": 5.983,
      "mean_token_accuracy": 0.16015656739473344,
      "num_tokens": 3183681.0,
      "step": 1385
    },
    {
      "entropy": 5.799786615371704,
      "epoch": 0.1335254562920269,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999841906758093,
      "loss": 5.7942,
      "mean_token_accuracy": 0.16851347237825393,
      "num_tokens": 3194122.0,
      "step": 1390
    },
    {
      "entropy": 5.937053966522217,
      "epoch": 0.1340057636887608,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999837816594757,
      "loss": 5.8463,
      "mean_token_accuracy": 0.16169387549161912,
      "num_tokens": 3205719.0,
      "step": 1395
    },
    {
      "entropy": 5.814732265472412,
      "epoch": 0.1344860710854947,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999833674196865,
      "loss": 5.8652,
      "mean_token_accuracy": 0.15869019478559493,
      "num_tokens": 3216262.0,
      "step": 1400
    },
    {
      "entropy": 6.002225685119629,
      "epoch": 0.13496637848222862,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999829479564518,
      "loss": 5.9264,
      "mean_token_accuracy": 0.14966704398393632,
      "num_tokens": 3228035.0,
      "step": 1405
    },
    {
      "entropy": 5.857042741775513,
      "epoch": 0.13544668587896252,
      "grad_norm": 1.0,
      "learning_rate": 0.000499982523269781,
      "loss": 5.8011,
      "mean_token_accuracy": 0.16352371722459794,
      "num_tokens": 3239643.0,
      "step": 1410
    },
    {
      "entropy": 5.89963698387146,
      "epoch": 0.13592699327569643,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999820933596842,
      "loss": 5.881,
      "mean_token_accuracy": 0.159575891494751,
      "num_tokens": 3250293.0,
      "step": 1415
    },
    {
      "entropy": 5.967745113372803,
      "epoch": 0.13640730067243034,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499981658226171,
      "loss": 6.0057,
      "mean_token_accuracy": 0.14675760343670846,
      "num_tokens": 3262448.0,
      "step": 1420
    },
    {
      "entropy": 5.867046260833741,
      "epoch": 0.13688760806916425,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499981217869252,
      "loss": 5.8731,
      "mean_token_accuracy": 0.1589981123805046,
      "num_tokens": 3274041.0,
      "step": 1425
    },
    {
      "entropy": 5.931325817108155,
      "epoch": 0.1373679154658982,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499980772288937,
      "loss": 5.9072,
      "mean_token_accuracy": 0.15695446729660034,
      "num_tokens": 3285063.0,
      "step": 1430
    },
    {
      "entropy": 5.940659904479981,
      "epoch": 0.1378482228626321,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999803214852367,
      "loss": 5.9238,
      "mean_token_accuracy": 0.15722591131925584,
      "num_tokens": 3297031.0,
      "step": 1435
    },
    {
      "entropy": 5.956357479095459,
      "epoch": 0.138328530259366,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999798654581613,
      "loss": 5.9031,
      "mean_token_accuracy": 0.1554128795862198,
      "num_tokens": 3308187.0,
      "step": 1440
    },
    {
      "entropy": 5.911098432540894,
      "epoch": 0.13880883765609991,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999794042077214,
      "loss": 5.9062,
      "mean_token_accuracy": 0.15348758697509765,
      "num_tokens": 3319439.0,
      "step": 1445
    },
    {
      "entropy": 5.934011363983155,
      "epoch": 0.13928914505283382,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999789377339279,
      "loss": 5.8996,
      "mean_token_accuracy": 0.15216536745429038,
      "num_tokens": 3331715.0,
      "step": 1450
    },
    {
      "entropy": 5.9182721138000485,
      "epoch": 0.13976945244956773,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999784660367915,
      "loss": 5.8609,
      "mean_token_accuracy": 0.16844206303358078,
      "num_tokens": 3342281.0,
      "step": 1455
    },
    {
      "entropy": 6.01653094291687,
      "epoch": 0.14024975984630164,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999779891163231,
      "loss": 5.9946,
      "mean_token_accuracy": 0.15272812992334367,
      "num_tokens": 3353530.0,
      "step": 1460
    },
    {
      "entropy": 5.90668683052063,
      "epoch": 0.14073006724303555,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999775069725339,
      "loss": 5.9582,
      "mean_token_accuracy": 0.1540558233857155,
      "num_tokens": 3365083.0,
      "step": 1465
    },
    {
      "entropy": 5.972508382797241,
      "epoch": 0.14121037463976946,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499977019605435,
      "loss": 5.9154,
      "mean_token_accuracy": 0.15416828840970992,
      "num_tokens": 3378143.0,
      "step": 1470
    },
    {
      "entropy": 5.9383855819702145,
      "epoch": 0.14169068203650337,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999765270150378,
      "loss": 5.882,
      "mean_token_accuracy": 0.1628871500492096,
      "num_tokens": 3388438.0,
      "step": 1475
    },
    {
      "entropy": 5.881742811203003,
      "epoch": 0.14217098943323728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999760292013536,
      "loss": 5.8832,
      "mean_token_accuracy": 0.15267159640789033,
      "num_tokens": 3400150.0,
      "step": 1480
    },
    {
      "entropy": 5.778295660018921,
      "epoch": 0.14265129682997119,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999755261643941,
      "loss": 5.765,
      "mean_token_accuracy": 0.16744979172945024,
      "num_tokens": 3411073.0,
      "step": 1485
    },
    {
      "entropy": 5.8690554141998295,
      "epoch": 0.1431316042267051,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999750179041709,
      "loss": 5.8341,
      "mean_token_accuracy": 0.1667776271700859,
      "num_tokens": 3422196.0,
      "step": 1490
    },
    {
      "entropy": 5.88673849105835,
      "epoch": 0.143611911623439,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999745044206959,
      "loss": 5.8586,
      "mean_token_accuracy": 0.15091366693377495,
      "num_tokens": 3432568.0,
      "step": 1495
    },
    {
      "entropy": 5.872237205505371,
      "epoch": 0.1440922190201729,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999739857139809,
      "loss": 5.8889,
      "mean_token_accuracy": 0.15535980463027954,
      "num_tokens": 3444096.0,
      "step": 1500
    },
    {
      "entropy": 5.795199871063232,
      "epoch": 0.14457252641690682,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499973461784038,
      "loss": 5.7672,
      "mean_token_accuracy": 0.1684318333864212,
      "num_tokens": 3455502.0,
      "step": 1505
    },
    {
      "entropy": 5.938459205627441,
      "epoch": 0.14505283381364073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999729326308792,
      "loss": 5.9119,
      "mean_token_accuracy": 0.15085936039686204,
      "num_tokens": 3467753.0,
      "step": 1510
    },
    {
      "entropy": 5.837009334564209,
      "epoch": 0.14553314121037464,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499972398254517,
      "loss": 5.7789,
      "mean_token_accuracy": 0.1624188169836998,
      "num_tokens": 3479206.0,
      "step": 1515
    },
    {
      "entropy": 5.872801876068115,
      "epoch": 0.14601344860710855,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499971858654964,
      "loss": 5.8213,
      "mean_token_accuracy": 0.1598054826259613,
      "num_tokens": 3490914.0,
      "step": 1520
    },
    {
      "entropy": 5.859324884414673,
      "epoch": 0.14649375600384246,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999713138322321,
      "loss": 5.8751,
      "mean_token_accuracy": 0.15338242948055267,
      "num_tokens": 3502932.0,
      "step": 1525
    },
    {
      "entropy": 5.787054777145386,
      "epoch": 0.14697406340057637,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999707637863346,
      "loss": 5.7941,
      "mean_token_accuracy": 0.16474374681711196,
      "num_tokens": 3515071.0,
      "step": 1530
    },
    {
      "entropy": 5.964262914657593,
      "epoch": 0.14745437079731027,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999702085172838,
      "loss": 5.9125,
      "mean_token_accuracy": 0.15990954861044884,
      "num_tokens": 3526845.0,
      "step": 1535
    },
    {
      "entropy": 5.844745063781739,
      "epoch": 0.14793467819404418,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999696480250929,
      "loss": 5.8904,
      "mean_token_accuracy": 0.15008662939071654,
      "num_tokens": 3538810.0,
      "step": 1540
    },
    {
      "entropy": 5.913762807846069,
      "epoch": 0.1484149855907781,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999690823097747,
      "loss": 5.8619,
      "mean_token_accuracy": 0.168930259346962,
      "num_tokens": 3549570.0,
      "step": 1545
    },
    {
      "entropy": 5.6661440372467045,
      "epoch": 0.148895292987512,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999685113713426,
      "loss": 5.7885,
      "mean_token_accuracy": 0.16731411069631577,
      "num_tokens": 3560937.0,
      "step": 1550
    },
    {
      "entropy": 5.910660028457642,
      "epoch": 0.1493756003842459,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999679352098096,
      "loss": 5.8428,
      "mean_token_accuracy": 0.15711865276098252,
      "num_tokens": 3573603.0,
      "step": 1555
    },
    {
      "entropy": 5.797772836685181,
      "epoch": 0.14985590778097982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999673538251891,
      "loss": 5.8481,
      "mean_token_accuracy": 0.15457407534122466,
      "num_tokens": 3586034.0,
      "step": 1560
    },
    {
      "entropy": 5.83761682510376,
      "epoch": 0.15033621517771373,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999667672174947,
      "loss": 5.7873,
      "mean_token_accuracy": 0.16625763177871705,
      "num_tokens": 3597305.0,
      "step": 1565
    },
    {
      "entropy": 5.9048542976379395,
      "epoch": 0.15081652257444764,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049996617538674,
      "loss": 5.9291,
      "mean_token_accuracy": 0.1503527745604515,
      "num_tokens": 3608758.0,
      "step": 1570
    },
    {
      "entropy": 5.960012817382813,
      "epoch": 0.15129682997118155,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999655783329386,
      "loss": 5.9154,
      "mean_token_accuracy": 0.15208876878023148,
      "num_tokens": 3621508.0,
      "step": 1575
    },
    {
      "entropy": 5.895190334320068,
      "epoch": 0.15177713736791545,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999649760561046,
      "loss": 5.8637,
      "mean_token_accuracy": 0.1636648565530777,
      "num_tokens": 3634826.0,
      "step": 1580
    },
    {
      "entropy": 5.734798717498779,
      "epoch": 0.15225744476464936,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999643685562519,
      "loss": 5.7017,
      "mean_token_accuracy": 0.18017627894878388,
      "num_tokens": 3646427.0,
      "step": 1585
    },
    {
      "entropy": 5.792345857620239,
      "epoch": 0.15273775216138327,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999637558333945,
      "loss": 5.833,
      "mean_token_accuracy": 0.16134492605924605,
      "num_tokens": 3658157.0,
      "step": 1590
    },
    {
      "entropy": 5.953826999664306,
      "epoch": 0.15321805955811718,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999631378875467,
      "loss": 5.8724,
      "mean_token_accuracy": 0.1592890664935112,
      "num_tokens": 3670695.0,
      "step": 1595
    },
    {
      "entropy": 5.788533782958984,
      "epoch": 0.15369836695485112,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999625147187228,
      "loss": 5.8006,
      "mean_token_accuracy": 0.15868576020002365,
      "num_tokens": 3683379.0,
      "step": 1600
    },
    {
      "entropy": 5.749165391921997,
      "epoch": 0.15417867435158503,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999618863269373,
      "loss": 5.6955,
      "mean_token_accuracy": 0.17023993730545045,
      "num_tokens": 3694411.0,
      "step": 1605
    },
    {
      "entropy": 5.775392723083496,
      "epoch": 0.15465898174831894,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999612527122049,
      "loss": 5.7396,
      "mean_token_accuracy": 0.16863008290529252,
      "num_tokens": 3703833.0,
      "step": 1610
    },
    {
      "entropy": 5.853102731704712,
      "epoch": 0.15513928914505284,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999606138745402,
      "loss": 5.8203,
      "mean_token_accuracy": 0.1702694848179817,
      "num_tokens": 3714702.0,
      "step": 1615
    },
    {
      "entropy": 5.865390205383301,
      "epoch": 0.15561959654178675,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999599698139581,
      "loss": 5.9266,
      "mean_token_accuracy": 0.16280574798583985,
      "num_tokens": 3726000.0,
      "step": 1620
    },
    {
      "entropy": 5.882253932952881,
      "epoch": 0.15609990393852066,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999593205304734,
      "loss": 5.8091,
      "mean_token_accuracy": 0.15844953060150146,
      "num_tokens": 3737681.0,
      "step": 1625
    },
    {
      "entropy": 5.8566688060760494,
      "epoch": 0.15658021133525457,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999586660241012,
      "loss": 5.9001,
      "mean_token_accuracy": 0.15781862735748292,
      "num_tokens": 3749477.0,
      "step": 1630
    },
    {
      "entropy": 5.952614831924438,
      "epoch": 0.15706051873198848,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999580062948569,
      "loss": 5.8735,
      "mean_token_accuracy": 0.16022729724645615,
      "num_tokens": 3759946.0,
      "step": 1635
    },
    {
      "entropy": 5.780766010284424,
      "epoch": 0.1575408261287224,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999573413427556,
      "loss": 5.7901,
      "mean_token_accuracy": 0.16178222000598907,
      "num_tokens": 3770147.0,
      "step": 1640
    },
    {
      "entropy": 5.765101289749145,
      "epoch": 0.1580211335254563,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999566711678128,
      "loss": 5.7431,
      "mean_token_accuracy": 0.17101316899061203,
      "num_tokens": 3781028.0,
      "step": 1645
    },
    {
      "entropy": 5.815601539611817,
      "epoch": 0.1585014409221902,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999559957700442,
      "loss": 5.817,
      "mean_token_accuracy": 0.1586659610271454,
      "num_tokens": 3792529.0,
      "step": 1650
    },
    {
      "entropy": 5.89880256652832,
      "epoch": 0.15898174831892412,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999553151494653,
      "loss": 5.9256,
      "mean_token_accuracy": 0.16100257486104966,
      "num_tokens": 3804630.0,
      "step": 1655
    },
    {
      "entropy": 5.736216402053833,
      "epoch": 0.15946205571565802,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999546293060919,
      "loss": 5.6871,
      "mean_token_accuracy": 0.1763335272669792,
      "num_tokens": 3815019.0,
      "step": 1660
    },
    {
      "entropy": 5.719022655487061,
      "epoch": 0.15994236311239193,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00049995393823994,
      "loss": 5.7902,
      "mean_token_accuracy": 0.15792783051729203,
      "num_tokens": 3825648.0,
      "step": 1665
    },
    {
      "entropy": 5.86321792602539,
      "epoch": 0.16042267050912584,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999532419510255,
      "loss": 5.7728,
      "mean_token_accuracy": 0.16869686394929886,
      "num_tokens": 3836779.0,
      "step": 1670
    },
    {
      "entropy": 5.7344707489013675,
      "epoch": 0.16090297790585975,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499952540439365,
      "loss": 5.8289,
      "mean_token_accuracy": 0.16161244064569474,
      "num_tokens": 3848350.0,
      "step": 1675
    },
    {
      "entropy": 5.933605194091797,
      "epoch": 0.16138328530259366,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999518337049743,
      "loss": 5.8159,
      "mean_token_accuracy": 0.16452196836471558,
      "num_tokens": 3859792.0,
      "step": 1680
    },
    {
      "entropy": 5.732853984832763,
      "epoch": 0.16186359269932757,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049995112174787,
      "loss": 5.7598,
      "mean_token_accuracy": 0.16248514652252197,
      "num_tokens": 3871539.0,
      "step": 1685
    },
    {
      "entropy": 5.743935489654541,
      "epoch": 0.16234390009606148,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999504045680687,
      "loss": 5.8142,
      "mean_token_accuracy": 0.1657908469438553,
      "num_tokens": 3885025.0,
      "step": 1690
    },
    {
      "entropy": 5.881377410888672,
      "epoch": 0.1628242074927954,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999496821655869,
      "loss": 5.9101,
      "mean_token_accuracy": 0.15664341002702714,
      "num_tokens": 3896519.0,
      "step": 1695
    },
    {
      "entropy": 5.774726724624633,
      "epoch": 0.1633045148895293,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999489545404414,
      "loss": 5.6679,
      "mean_token_accuracy": 0.16605159938335418,
      "num_tokens": 3907636.0,
      "step": 1700
    },
    {
      "entropy": 5.7758361339569095,
      "epoch": 0.1637848222862632,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999482216926493,
      "loss": 5.7873,
      "mean_token_accuracy": 0.16288612633943558,
      "num_tokens": 3917746.0,
      "step": 1705
    },
    {
      "entropy": 5.775166130065918,
      "epoch": 0.1642651296829971,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999474836222273,
      "loss": 5.7129,
      "mean_token_accuracy": 0.17258119732141494,
      "num_tokens": 3928901.0,
      "step": 1710
    },
    {
      "entropy": 5.706221103668213,
      "epoch": 0.16474543707973102,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999467403291928,
      "loss": 5.8489,
      "mean_token_accuracy": 0.16663677543401717,
      "num_tokens": 3940094.0,
      "step": 1715
    },
    {
      "entropy": 5.823417139053345,
      "epoch": 0.16522574447646493,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999459918135628,
      "loss": 5.7544,
      "mean_token_accuracy": 0.16869125217199327,
      "num_tokens": 3952714.0,
      "step": 1720
    },
    {
      "entropy": 5.729210901260376,
      "epoch": 0.16570605187319884,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499945238075355,
      "loss": 5.7318,
      "mean_token_accuracy": 0.1667981579899788,
      "num_tokens": 3963107.0,
      "step": 1725
    },
    {
      "entropy": 5.833112668991089,
      "epoch": 0.16618635926993275,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999444791145865,
      "loss": 5.6788,
      "mean_token_accuracy": 0.16824763119220734,
      "num_tokens": 3973454.0,
      "step": 1730
    },
    {
      "entropy": 5.756386947631836,
      "epoch": 0.16666666666666666,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999437149312754,
      "loss": 5.8065,
      "mean_token_accuracy": 0.1626473069190979,
      "num_tokens": 3985352.0,
      "step": 1735
    },
    {
      "entropy": 5.8663242816925045,
      "epoch": 0.16714697406340057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499942945525439,
      "loss": 5.7607,
      "mean_token_accuracy": 0.16653031706809998,
      "num_tokens": 3997088.0,
      "step": 1740
    },
    {
      "entropy": 5.762393426895142,
      "epoch": 0.16762728146013448,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999421708970954,
      "loss": 5.7977,
      "mean_token_accuracy": 0.17101218104362487,
      "num_tokens": 4008551.0,
      "step": 1745
    },
    {
      "entropy": 5.807009887695313,
      "epoch": 0.16810758885686838,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999413910462625,
      "loss": 5.827,
      "mean_token_accuracy": 0.16373823583126068,
      "num_tokens": 4020886.0,
      "step": 1750
    },
    {
      "entropy": 5.802869558334351,
      "epoch": 0.1685878962536023,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999406059729586,
      "loss": 5.7631,
      "mean_token_accuracy": 0.1604061797261238,
      "num_tokens": 4032811.0,
      "step": 1755
    },
    {
      "entropy": 5.792250299453736,
      "epoch": 0.1690682036503362,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999398156772016,
      "loss": 5.7608,
      "mean_token_accuracy": 0.16829253733158112,
      "num_tokens": 4044102.0,
      "step": 1760
    },
    {
      "entropy": 5.774747133255005,
      "epoch": 0.16954851104707014,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049993902015901,
      "loss": 5.7878,
      "mean_token_accuracy": 0.16452797800302504,
      "num_tokens": 4056849.0,
      "step": 1765
    },
    {
      "entropy": 5.772494602203369,
      "epoch": 0.17002881844380405,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999382194184023,
      "loss": 5.8186,
      "mean_token_accuracy": 0.16248029321432114,
      "num_tokens": 4068392.0,
      "step": 1770
    },
    {
      "entropy": 5.777360963821411,
      "epoch": 0.17050912584053796,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999374134553972,
      "loss": 5.646,
      "mean_token_accuracy": 0.17268918007612227,
      "num_tokens": 4078949.0,
      "step": 1775
    },
    {
      "entropy": 5.744507360458374,
      "epoch": 0.17098943323727187,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999366022700131,
      "loss": 5.8057,
      "mean_token_accuracy": 0.1625596508383751,
      "num_tokens": 4091432.0,
      "step": 1780
    },
    {
      "entropy": 5.717019987106323,
      "epoch": 0.17146974063400577,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999357858622691,
      "loss": 5.6587,
      "mean_token_accuracy": 0.171124404668808,
      "num_tokens": 4103001.0,
      "step": 1785
    },
    {
      "entropy": 5.755549430847168,
      "epoch": 0.17195004803073968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999349642321842,
      "loss": 5.7557,
      "mean_token_accuracy": 0.16994501501321793,
      "num_tokens": 4113570.0,
      "step": 1790
    },
    {
      "entropy": 5.760085391998291,
      "epoch": 0.1724303554274736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999341373797772,
      "loss": 5.8149,
      "mean_token_accuracy": 0.15936170816421508,
      "num_tokens": 4125658.0,
      "step": 1795
    },
    {
      "entropy": 5.720655012130737,
      "epoch": 0.1729106628242075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999333053050675,
      "loss": 5.7078,
      "mean_token_accuracy": 0.16616615355014802,
      "num_tokens": 4137424.0,
      "step": 1800
    },
    {
      "entropy": 5.744824743270874,
      "epoch": 0.1733909702209414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999324680080744,
      "loss": 5.7518,
      "mean_token_accuracy": 0.16471749246120454,
      "num_tokens": 4149772.0,
      "step": 1805
    },
    {
      "entropy": 5.733233070373535,
      "epoch": 0.17387127761767532,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999316254888172,
      "loss": 5.7438,
      "mean_token_accuracy": 0.16947544515132903,
      "num_tokens": 4162246.0,
      "step": 1810
    },
    {
      "entropy": 5.736325740814209,
      "epoch": 0.17435158501440923,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999307777473157,
      "loss": 5.7074,
      "mean_token_accuracy": 0.174015836417675,
      "num_tokens": 4173449.0,
      "step": 1815
    },
    {
      "entropy": 5.735329103469849,
      "epoch": 0.17483189241114314,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999299247835893,
      "loss": 5.7338,
      "mean_token_accuracy": 0.16482603400945664,
      "num_tokens": 4185560.0,
      "step": 1820
    },
    {
      "entropy": 5.752186679840088,
      "epoch": 0.17531219980787704,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499929066597658,
      "loss": 5.814,
      "mean_token_accuracy": 0.15707085877656937,
      "num_tokens": 4197843.0,
      "step": 1825
    },
    {
      "entropy": 5.825539398193359,
      "epoch": 0.17579250720461095,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999282031895418,
      "loss": 5.6427,
      "mean_token_accuracy": 0.16972217261791228,
      "num_tokens": 4210009.0,
      "step": 1830
    },
    {
      "entropy": 5.74087233543396,
      "epoch": 0.17627281460134486,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999273345592604,
      "loss": 5.9079,
      "mean_token_accuracy": 0.15466363951563836,
      "num_tokens": 4222213.0,
      "step": 1835
    },
    {
      "entropy": 5.786208820343018,
      "epoch": 0.17675312199807877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999264607068343,
      "loss": 5.7524,
      "mean_token_accuracy": 0.16184426844120026,
      "num_tokens": 4234159.0,
      "step": 1840
    },
    {
      "entropy": 5.770721483230591,
      "epoch": 0.17723342939481268,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999255816322837,
      "loss": 5.8244,
      "mean_token_accuracy": 0.15809911489486694,
      "num_tokens": 4246358.0,
      "step": 1845
    },
    {
      "entropy": 5.925276517868042,
      "epoch": 0.1777137367915466,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499924697335629,
      "loss": 5.8836,
      "mean_token_accuracy": 0.1607755497097969,
      "num_tokens": 4256698.0,
      "step": 1850
    },
    {
      "entropy": 5.714931774139404,
      "epoch": 0.1781940441882805,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999238078168906,
      "loss": 5.6716,
      "mean_token_accuracy": 0.1786502331495285,
      "num_tokens": 4268007.0,
      "step": 1855
    },
    {
      "entropy": 5.753628253936768,
      "epoch": 0.1786743515850144,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999229130760894,
      "loss": 5.7698,
      "mean_token_accuracy": 0.16579428613185881,
      "num_tokens": 4278583.0,
      "step": 1860
    },
    {
      "entropy": 5.856677532196045,
      "epoch": 0.17915465898174832,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499922013113246,
      "loss": 5.7914,
      "mean_token_accuracy": 0.16347781121730803,
      "num_tokens": 4290267.0,
      "step": 1865
    },
    {
      "entropy": 5.739168548583985,
      "epoch": 0.17963496637848222,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999211079283814,
      "loss": 5.7111,
      "mean_token_accuracy": 0.16955134123563767,
      "num_tokens": 4302177.0,
      "step": 1870
    },
    {
      "entropy": 5.694641065597534,
      "epoch": 0.18011527377521613,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999201975215164,
      "loss": 5.6553,
      "mean_token_accuracy": 0.16879584789276122,
      "num_tokens": 4312420.0,
      "step": 1875
    },
    {
      "entropy": 5.815468692779541,
      "epoch": 0.18059558117195004,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999192818926725,
      "loss": 5.8052,
      "mean_token_accuracy": 0.15983834564685823,
      "num_tokens": 4324050.0,
      "step": 1880
    },
    {
      "entropy": 5.708034896850586,
      "epoch": 0.18107588856868395,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999183610418706,
      "loss": 5.6484,
      "mean_token_accuracy": 0.16869973838329316,
      "num_tokens": 4335336.0,
      "step": 1885
    },
    {
      "entropy": 5.616009616851807,
      "epoch": 0.18155619596541786,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999174349691322,
      "loss": 5.662,
      "mean_token_accuracy": 0.17169511318206787,
      "num_tokens": 4346239.0,
      "step": 1890
    },
    {
      "entropy": 5.76420841217041,
      "epoch": 0.18203650336215177,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999165036744788,
      "loss": 5.7368,
      "mean_token_accuracy": 0.16637959629297255,
      "num_tokens": 4357195.0,
      "step": 1895
    },
    {
      "entropy": 5.639016723632812,
      "epoch": 0.18251681075888568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999155671579322,
      "loss": 5.6806,
      "mean_token_accuracy": 0.17036481946706772,
      "num_tokens": 4368905.0,
      "step": 1900
    },
    {
      "entropy": 5.747797107696533,
      "epoch": 0.1829971181556196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499914625419514,
      "loss": 5.7451,
      "mean_token_accuracy": 0.16570276916027069,
      "num_tokens": 4381708.0,
      "step": 1905
    },
    {
      "entropy": 5.6733973026275635,
      "epoch": 0.1834774255523535,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999136784592459,
      "loss": 5.591,
      "mean_token_accuracy": 0.18222368210554124,
      "num_tokens": 4393199.0,
      "step": 1910
    },
    {
      "entropy": 5.731898021697998,
      "epoch": 0.1839577329490874,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999127262771502,
      "loss": 5.8355,
      "mean_token_accuracy": 0.15900960564613342,
      "num_tokens": 4403986.0,
      "step": 1915
    },
    {
      "entropy": 5.8278138637542725,
      "epoch": 0.1844380403458213,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999117688732487,
      "loss": 5.6824,
      "mean_token_accuracy": 0.16800769567489623,
      "num_tokens": 4415574.0,
      "step": 1920
    },
    {
      "entropy": 5.6885899066925045,
      "epoch": 0.18491834774255522,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999108062475638,
      "loss": 5.7163,
      "mean_token_accuracy": 0.1703812912106514,
      "num_tokens": 4427579.0,
      "step": 1925
    },
    {
      "entropy": 5.7626372337341305,
      "epoch": 0.18539865513928913,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499909838400118,
      "loss": 5.7116,
      "mean_token_accuracy": 0.16777720302343369,
      "num_tokens": 4440045.0,
      "step": 1930
    },
    {
      "entropy": 5.7155296325683596,
      "epoch": 0.18587896253602307,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999088653309334,
      "loss": 5.8093,
      "mean_token_accuracy": 0.16413577646017075,
      "num_tokens": 4450129.0,
      "step": 1935
    },
    {
      "entropy": 5.716533088684082,
      "epoch": 0.18635926993275698,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999078870400329,
      "loss": 5.6497,
      "mean_token_accuracy": 0.17000699192285537,
      "num_tokens": 4461378.0,
      "step": 1940
    },
    {
      "entropy": 5.688685703277588,
      "epoch": 0.18683957732949089,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999069035274391,
      "loss": 5.7,
      "mean_token_accuracy": 0.17359329164028167,
      "num_tokens": 4472249.0,
      "step": 1945
    },
    {
      "entropy": 5.717999982833862,
      "epoch": 0.1873198847262248,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999059147931747,
      "loss": 5.7387,
      "mean_token_accuracy": 0.166311414539814,
      "num_tokens": 4484243.0,
      "step": 1950
    },
    {
      "entropy": 5.771619606018066,
      "epoch": 0.1878001921229587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999049208372629,
      "loss": 5.6593,
      "mean_token_accuracy": 0.16572581231594086,
      "num_tokens": 4495776.0,
      "step": 1955
    },
    {
      "entropy": 5.670734643936157,
      "epoch": 0.1882804995196926,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999039216597267,
      "loss": 5.7501,
      "mean_token_accuracy": 0.16451041847467424,
      "num_tokens": 4508581.0,
      "step": 1960
    },
    {
      "entropy": 5.775052642822265,
      "epoch": 0.18876080691642652,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999029172605892,
      "loss": 5.6696,
      "mean_token_accuracy": 0.16900223344564438,
      "num_tokens": 4519753.0,
      "step": 1965
    },
    {
      "entropy": 5.719663810729981,
      "epoch": 0.18924111431316043,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999019076398738,
      "loss": 5.75,
      "mean_token_accuracy": 0.17347084879875183,
      "num_tokens": 4529957.0,
      "step": 1970
    },
    {
      "entropy": 5.759815454483032,
      "epoch": 0.18972142170989434,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499900892797604,
      "loss": 5.7093,
      "mean_token_accuracy": 0.16766101717948914,
      "num_tokens": 4541181.0,
      "step": 1975
    },
    {
      "entropy": 5.702226018905639,
      "epoch": 0.19020172910662825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998998727338031,
      "loss": 5.6776,
      "mean_token_accuracy": 0.173050557076931,
      "num_tokens": 4552235.0,
      "step": 1980
    },
    {
      "entropy": 5.690057182312012,
      "epoch": 0.19068203650336216,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998988474484952,
      "loss": 5.6913,
      "mean_token_accuracy": 0.17268367409706115,
      "num_tokens": 4564045.0,
      "step": 1985
    },
    {
      "entropy": 5.639818572998047,
      "epoch": 0.19116234390009607,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998978169417038,
      "loss": 5.586,
      "mean_token_accuracy": 0.17723700404167175,
      "num_tokens": 4575141.0,
      "step": 1990
    },
    {
      "entropy": 5.673330593109131,
      "epoch": 0.19164265129682997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998967812134529,
      "loss": 5.7166,
      "mean_token_accuracy": 0.1675497755408287,
      "num_tokens": 4586491.0,
      "step": 1995
    },
    {
      "entropy": 5.74068775177002,
      "epoch": 0.19212295869356388,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998957402637664,
      "loss": 5.697,
      "mean_token_accuracy": 0.16967157423496246,
      "num_tokens": 4598421.0,
      "step": 2000
    },
    {
      "entropy": 5.713365173339843,
      "epoch": 0.1926032660902978,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998946940926687,
      "loss": 5.5956,
      "mean_token_accuracy": 0.18232411295175552,
      "num_tokens": 4609688.0,
      "step": 2005
    },
    {
      "entropy": 5.647695112228393,
      "epoch": 0.1930835734870317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499893642700184,
      "loss": 5.7603,
      "mean_token_accuracy": 0.16484474539756774,
      "num_tokens": 4621169.0,
      "step": 2010
    },
    {
      "entropy": 5.75804328918457,
      "epoch": 0.1935638808837656,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998925860863368,
      "loss": 5.7089,
      "mean_token_accuracy": 0.16970116794109344,
      "num_tokens": 4632512.0,
      "step": 2015
    },
    {
      "entropy": 5.767560434341431,
      "epoch": 0.19404418828049952,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998915242511516,
      "loss": 5.763,
      "mean_token_accuracy": 0.16098930090665817,
      "num_tokens": 4644356.0,
      "step": 2020
    },
    {
      "entropy": 5.748520946502685,
      "epoch": 0.19452449567723343,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998904571946528,
      "loss": 5.7372,
      "mean_token_accuracy": 0.16949849873781203,
      "num_tokens": 4655601.0,
      "step": 2025
    },
    {
      "entropy": 5.7134335994720455,
      "epoch": 0.19500480307396734,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998893849168655,
      "loss": 5.7485,
      "mean_token_accuracy": 0.16730379909276963,
      "num_tokens": 4666813.0,
      "step": 2030
    },
    {
      "entropy": 5.751007986068726,
      "epoch": 0.19548511047070125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998883074178144,
      "loss": 5.7471,
      "mean_token_accuracy": 0.16477151960134506,
      "num_tokens": 4679123.0,
      "step": 2035
    },
    {
      "entropy": 5.712923908233643,
      "epoch": 0.19596541786743515,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998872246975247,
      "loss": 5.6323,
      "mean_token_accuracy": 0.17063149064779282,
      "num_tokens": 4690485.0,
      "step": 2040
    },
    {
      "entropy": 5.578420925140381,
      "epoch": 0.19644572526416906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998861367560213,
      "loss": 5.6523,
      "mean_token_accuracy": 0.17362430989742278,
      "num_tokens": 4702573.0,
      "step": 2045
    },
    {
      "entropy": 5.7771985054016115,
      "epoch": 0.19692603266090297,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998850435933296,
      "loss": 5.7034,
      "mean_token_accuracy": 0.16219049245119094,
      "num_tokens": 4715653.0,
      "step": 2050
    },
    {
      "entropy": 5.648106813430786,
      "epoch": 0.19740634005763688,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998839452094749,
      "loss": 5.6953,
      "mean_token_accuracy": 0.1705027177929878,
      "num_tokens": 4726051.0,
      "step": 2055
    },
    {
      "entropy": 5.749042558670044,
      "epoch": 0.1978866474543708,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998828416044829,
      "loss": 5.8326,
      "mean_token_accuracy": 0.16298552751541137,
      "num_tokens": 4738176.0,
      "step": 2060
    },
    {
      "entropy": 5.764062976837158,
      "epoch": 0.1983669548511047,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499881732778379,
      "loss": 5.7244,
      "mean_token_accuracy": 0.16867188662290572,
      "num_tokens": 4749547.0,
      "step": 2065
    },
    {
      "entropy": 5.670080184936523,
      "epoch": 0.1988472622478386,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499880618731189,
      "loss": 5.6588,
      "mean_token_accuracy": 0.1658295750617981,
      "num_tokens": 4761967.0,
      "step": 2070
    },
    {
      "entropy": 5.638352346420288,
      "epoch": 0.19932756964457252,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998794994629388,
      "loss": 5.6486,
      "mean_token_accuracy": 0.17256570607423782,
      "num_tokens": 4772847.0,
      "step": 2075
    },
    {
      "entropy": 5.743561458587647,
      "epoch": 0.19980787704130643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998783749736545,
      "loss": 5.7147,
      "mean_token_accuracy": 0.16527553051710128,
      "num_tokens": 4785049.0,
      "step": 2080
    },
    {
      "entropy": 5.694084978103637,
      "epoch": 0.20028818443804033,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998772452633619,
      "loss": 5.7413,
      "mean_token_accuracy": 0.16240385919809341,
      "num_tokens": 4797484.0,
      "step": 2085
    },
    {
      "entropy": 5.689326190948487,
      "epoch": 0.20076849183477424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998761103320876,
      "loss": 5.6549,
      "mean_token_accuracy": 0.1645843580365181,
      "num_tokens": 4809422.0,
      "step": 2090
    },
    {
      "entropy": 5.750069236755371,
      "epoch": 0.20124879923150815,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998749701798577,
      "loss": 5.8479,
      "mean_token_accuracy": 0.16339512243866922,
      "num_tokens": 4822340.0,
      "step": 2095
    },
    {
      "entropy": 5.713565874099731,
      "epoch": 0.2017291066282421,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998738248066986,
      "loss": 5.6261,
      "mean_token_accuracy": 0.17067924439907073,
      "num_tokens": 4833382.0,
      "step": 2100
    },
    {
      "entropy": 5.642834043502807,
      "epoch": 0.202209414024976,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998726742126372,
      "loss": 5.6054,
      "mean_token_accuracy": 0.179599392414093,
      "num_tokens": 4843916.0,
      "step": 2105
    },
    {
      "entropy": 5.686129951477051,
      "epoch": 0.2026897214217099,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998715183976999,
      "loss": 5.6313,
      "mean_token_accuracy": 0.17115692496299745,
      "num_tokens": 4855420.0,
      "step": 2110
    },
    {
      "entropy": 5.717159843444824,
      "epoch": 0.20317002881844382,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998703573619137,
      "loss": 5.7091,
      "mean_token_accuracy": 0.16790662705898285,
      "num_tokens": 4866540.0,
      "step": 2115
    },
    {
      "entropy": 5.737949991226197,
      "epoch": 0.20365033621517772,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998691911053056,
      "loss": 5.6503,
      "mean_token_accuracy": 0.1651172399520874,
      "num_tokens": 4878860.0,
      "step": 2120
    },
    {
      "entropy": 5.814427518844605,
      "epoch": 0.20413064361191163,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998680196279026,
      "loss": 5.775,
      "mean_token_accuracy": 0.1618766739964485,
      "num_tokens": 4889692.0,
      "step": 2125
    },
    {
      "entropy": 5.733552122116089,
      "epoch": 0.20461095100864554,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998668429297319,
      "loss": 5.7168,
      "mean_token_accuracy": 0.1730179101228714,
      "num_tokens": 4901337.0,
      "step": 2130
    },
    {
      "entropy": 5.628330326080322,
      "epoch": 0.20509125840537945,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998656610108208,
      "loss": 5.6162,
      "mean_token_accuracy": 0.1758786827325821,
      "num_tokens": 4912340.0,
      "step": 2135
    },
    {
      "entropy": 5.650088310241699,
      "epoch": 0.20557156580211336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998644738711969,
      "loss": 5.6037,
      "mean_token_accuracy": 0.18037814944982528,
      "num_tokens": 4923429.0,
      "step": 2140
    },
    {
      "entropy": 5.692839908599853,
      "epoch": 0.20605187319884727,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998632815108874,
      "loss": 5.672,
      "mean_token_accuracy": 0.1628017544746399,
      "num_tokens": 4934826.0,
      "step": 2145
    },
    {
      "entropy": 5.705739879608155,
      "epoch": 0.20653218059558118,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998620839299203,
      "loss": 5.7485,
      "mean_token_accuracy": 0.1703076645731926,
      "num_tokens": 4945189.0,
      "step": 2150
    },
    {
      "entropy": 5.772362995147705,
      "epoch": 0.2070124879923151,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998608811283233,
      "loss": 5.6281,
      "mean_token_accuracy": 0.17795057147741317,
      "num_tokens": 4956060.0,
      "step": 2155
    },
    {
      "entropy": 5.667694616317749,
      "epoch": 0.207492795389049,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998596731061244,
      "loss": 5.6475,
      "mean_token_accuracy": 0.17147868126630783,
      "num_tokens": 4967710.0,
      "step": 2160
    },
    {
      "entropy": 5.711640691757202,
      "epoch": 0.2079731027857829,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998584598633516,
      "loss": 5.675,
      "mean_token_accuracy": 0.1658002734184265,
      "num_tokens": 4978716.0,
      "step": 2165
    },
    {
      "entropy": 5.694676256179809,
      "epoch": 0.2084534101825168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998572414000329,
      "loss": 5.7477,
      "mean_token_accuracy": 0.16772115379571914,
      "num_tokens": 4990742.0,
      "step": 2170
    },
    {
      "entropy": 5.771293210983276,
      "epoch": 0.20893371757925072,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998560177161969,
      "loss": 5.6868,
      "mean_token_accuracy": 0.17304315716028212,
      "num_tokens": 5001217.0,
      "step": 2175
    },
    {
      "entropy": 5.671742153167725,
      "epoch": 0.20941402497598463,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998547888118718,
      "loss": 5.6906,
      "mean_token_accuracy": 0.16428728848695756,
      "num_tokens": 5012750.0,
      "step": 2180
    },
    {
      "entropy": 5.71360969543457,
      "epoch": 0.20989433237271854,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998535546870862,
      "loss": 5.6916,
      "mean_token_accuracy": 0.17811540961265565,
      "num_tokens": 5023818.0,
      "step": 2185
    },
    {
      "entropy": 5.748793077468872,
      "epoch": 0.21037463976945245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998523153418687,
      "loss": 5.7013,
      "mean_token_accuracy": 0.16991553604602813,
      "num_tokens": 5035060.0,
      "step": 2190
    },
    {
      "entropy": 5.751564168930054,
      "epoch": 0.21085494716618636,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998510707762481,
      "loss": 5.7382,
      "mean_token_accuracy": 0.15955581441521643,
      "num_tokens": 5046780.0,
      "step": 2195
    },
    {
      "entropy": 5.7015965461730955,
      "epoch": 0.21133525456292027,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998498209902533,
      "loss": 5.7148,
      "mean_token_accuracy": 0.16586878895759583,
      "num_tokens": 5059152.0,
      "step": 2200
    },
    {
      "entropy": 5.687603616714478,
      "epoch": 0.21181556195965417,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998485659839134,
      "loss": 5.6789,
      "mean_token_accuracy": 0.1731492057442665,
      "num_tokens": 5069971.0,
      "step": 2205
    },
    {
      "entropy": 5.678986549377441,
      "epoch": 0.21229586935638808,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998473057572575,
      "loss": 5.632,
      "mean_token_accuracy": 0.17143055498600007,
      "num_tokens": 5080640.0,
      "step": 2210
    },
    {
      "entropy": 5.662276315689087,
      "epoch": 0.212776176753122,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998460403103146,
      "loss": 5.6432,
      "mean_token_accuracy": 0.1736779436469078,
      "num_tokens": 5091240.0,
      "step": 2215
    },
    {
      "entropy": 5.623275947570801,
      "epoch": 0.2132564841498559,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998447696431146,
      "loss": 5.6744,
      "mean_token_accuracy": 0.16834193468093872,
      "num_tokens": 5102920.0,
      "step": 2220
    },
    {
      "entropy": 5.729692602157593,
      "epoch": 0.2137367915465898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998434937556865,
      "loss": 5.6973,
      "mean_token_accuracy": 0.1643601253628731,
      "num_tokens": 5114240.0,
      "step": 2225
    },
    {
      "entropy": 5.849734830856323,
      "epoch": 0.21421709894332372,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998422126480602,
      "loss": 5.916,
      "mean_token_accuracy": 0.1578417658805847,
      "num_tokens": 5126653.0,
      "step": 2230
    },
    {
      "entropy": 5.786351490020752,
      "epoch": 0.21469740634005763,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998409263202653,
      "loss": 5.7602,
      "mean_token_accuracy": 0.1627764031291008,
      "num_tokens": 5138778.0,
      "step": 2235
    },
    {
      "entropy": 5.689749240875244,
      "epoch": 0.21517771373679154,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998396347723318,
      "loss": 5.613,
      "mean_token_accuracy": 0.17623931765556336,
      "num_tokens": 5148670.0,
      "step": 2240
    },
    {
      "entropy": 5.536856698989868,
      "epoch": 0.21565802113352545,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998383380042895,
      "loss": 5.6654,
      "mean_token_accuracy": 0.16913559436798095,
      "num_tokens": 5159617.0,
      "step": 2245
    },
    {
      "entropy": 5.721935510635376,
      "epoch": 0.21613832853025935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998370360161688,
      "loss": 5.7085,
      "mean_token_accuracy": 0.1725509762763977,
      "num_tokens": 5171225.0,
      "step": 2250
    },
    {
      "entropy": 5.699740695953369,
      "epoch": 0.21661863592699326,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998357288079996,
      "loss": 5.7162,
      "mean_token_accuracy": 0.16764118522405624,
      "num_tokens": 5184586.0,
      "step": 2255
    },
    {
      "entropy": 5.628728437423706,
      "epoch": 0.21709894332372717,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998344163798125,
      "loss": 5.6851,
      "mean_token_accuracy": 0.16584176570177078,
      "num_tokens": 5196089.0,
      "step": 2260
    },
    {
      "entropy": 5.7385218143463135,
      "epoch": 0.21757925072046108,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998330987316379,
      "loss": 5.7088,
      "mean_token_accuracy": 0.16616473644971846,
      "num_tokens": 5207581.0,
      "step": 2265
    },
    {
      "entropy": 5.65544490814209,
      "epoch": 0.21805955811719502,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004998317758635062,
      "loss": 5.668,
      "mean_token_accuracy": 0.1719151794910431,
      "num_tokens": 5219108.0,
      "step": 2270
    },
    {
      "entropy": 5.698003768920898,
      "epoch": 0.21853986551392893,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998304477754484,
      "loss": 5.6096,
      "mean_token_accuracy": 0.17026123851537706,
      "num_tokens": 5231137.0,
      "step": 2275
    },
    {
      "entropy": 5.793634605407715,
      "epoch": 0.21902017291066284,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998291144674952,
      "loss": 5.6957,
      "mean_token_accuracy": 0.17130818516016005,
      "num_tokens": 5243528.0,
      "step": 2280
    },
    {
      "entropy": 5.544963455200195,
      "epoch": 0.21950048030739674,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998277759396776,
      "loss": 5.6366,
      "mean_token_accuracy": 0.17359607815742492,
      "num_tokens": 5254869.0,
      "step": 2285
    },
    {
      "entropy": 5.613512754440308,
      "epoch": 0.21998078770413065,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998264321920265,
      "loss": 5.6196,
      "mean_token_accuracy": 0.17148103266954423,
      "num_tokens": 5265370.0,
      "step": 2290
    },
    {
      "entropy": 5.6785135746002195,
      "epoch": 0.22046109510086456,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998250832245734,
      "loss": 5.5957,
      "mean_token_accuracy": 0.17259823232889177,
      "num_tokens": 5277812.0,
      "step": 2295
    },
    {
      "entropy": 5.57698540687561,
      "epoch": 0.22094140249759847,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998237290373494,
      "loss": 5.5758,
      "mean_token_accuracy": 0.1758878469467163,
      "num_tokens": 5289107.0,
      "step": 2300
    },
    {
      "entropy": 5.679394817352295,
      "epoch": 0.22142170989433238,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499822369630386,
      "loss": 5.6315,
      "mean_token_accuracy": 0.16491704434156418,
      "num_tokens": 5301489.0,
      "step": 2305
    },
    {
      "entropy": 5.549877023696899,
      "epoch": 0.2219020172910663,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998210050037148,
      "loss": 5.6098,
      "mean_token_accuracy": 0.17484947741031648,
      "num_tokens": 5313273.0,
      "step": 2310
    },
    {
      "entropy": 5.7064066410064695,
      "epoch": 0.2223823246878002,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004998196351573674,
      "loss": 5.6389,
      "mean_token_accuracy": 0.17079515755176544,
      "num_tokens": 5324227.0,
      "step": 2315
    },
    {
      "entropy": 5.61033296585083,
      "epoch": 0.2228626320845341,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998182600913757,
      "loss": 5.5885,
      "mean_token_accuracy": 0.17890511006116866,
      "num_tokens": 5335138.0,
      "step": 2320
    },
    {
      "entropy": 5.5808428764343265,
      "epoch": 0.22334293948126802,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998168798057715,
      "loss": 5.6627,
      "mean_token_accuracy": 0.1708785980939865,
      "num_tokens": 5345652.0,
      "step": 2325
    },
    {
      "entropy": 5.7352948665618895,
      "epoch": 0.22382324687800192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499815494300587,
      "loss": 5.749,
      "mean_token_accuracy": 0.16825126260519027,
      "num_tokens": 5357999.0,
      "step": 2330
    },
    {
      "entropy": 5.732254123687744,
      "epoch": 0.22430355427473583,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998141035758542,
      "loss": 5.645,
      "mean_token_accuracy": 0.1687160015106201,
      "num_tokens": 5369032.0,
      "step": 2335
    },
    {
      "entropy": 5.6413782119750975,
      "epoch": 0.22478386167146974,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998127076316054,
      "loss": 5.5879,
      "mean_token_accuracy": 0.17352873533964158,
      "num_tokens": 5380692.0,
      "step": 2340
    },
    {
      "entropy": 5.730327844619751,
      "epoch": 0.22526416906820365,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998113064678734,
      "loss": 5.7181,
      "mean_token_accuracy": 0.16110651195049286,
      "num_tokens": 5392320.0,
      "step": 2345
    },
    {
      "entropy": 5.694684886932373,
      "epoch": 0.22574447646493756,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998099000846901,
      "loss": 5.7082,
      "mean_token_accuracy": 0.170077383518219,
      "num_tokens": 5404374.0,
      "step": 2350
    },
    {
      "entropy": 5.719666624069214,
      "epoch": 0.22622478386167147,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998084884820887,
      "loss": 5.6904,
      "mean_token_accuracy": 0.17066318392753602,
      "num_tokens": 5415471.0,
      "step": 2355
    },
    {
      "entropy": 5.532975578308106,
      "epoch": 0.22670509125840538,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004998070716601016,
      "loss": 5.5269,
      "mean_token_accuracy": 0.1728779211640358,
      "num_tokens": 5425971.0,
      "step": 2360
    },
    {
      "entropy": 5.670989894866944,
      "epoch": 0.2271853986551393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998056496187618,
      "loss": 5.6391,
      "mean_token_accuracy": 0.17769685834646226,
      "num_tokens": 5437638.0,
      "step": 2365
    },
    {
      "entropy": 5.6903704643249515,
      "epoch": 0.2276657060518732,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998042223581025,
      "loss": 5.6898,
      "mean_token_accuracy": 0.1719474747776985,
      "num_tokens": 5449463.0,
      "step": 2370
    },
    {
      "entropy": 5.664535570144653,
      "epoch": 0.2281460134486071,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004998027898781565,
      "loss": 5.7406,
      "mean_token_accuracy": 0.1684306815266609,
      "num_tokens": 5460972.0,
      "step": 2375
    },
    {
      "entropy": 5.866941213607788,
      "epoch": 0.228626320845341,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998013521789574,
      "loss": 5.7553,
      "mean_token_accuracy": 0.15989520251750947,
      "num_tokens": 5471575.0,
      "step": 2380
    },
    {
      "entropy": 5.589054918289184,
      "epoch": 0.22910662824207492,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997999092605384,
      "loss": 5.6089,
      "mean_token_accuracy": 0.17029613703489305,
      "num_tokens": 5483041.0,
      "step": 2385
    },
    {
      "entropy": 5.63406343460083,
      "epoch": 0.22958693563880883,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499798461122933,
      "loss": 5.6321,
      "mean_token_accuracy": 0.1724291518330574,
      "num_tokens": 5494040.0,
      "step": 2390
    },
    {
      "entropy": 5.6091015338897705,
      "epoch": 0.23006724303554274,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997970077661748,
      "loss": 5.5201,
      "mean_token_accuracy": 0.1848677799105644,
      "num_tokens": 5504217.0,
      "step": 2395
    },
    {
      "entropy": 5.669242858886719,
      "epoch": 0.23054755043227665,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997955491902977,
      "loss": 5.6951,
      "mean_token_accuracy": 0.16823576092720033,
      "num_tokens": 5517180.0,
      "step": 2400
    },
    {
      "entropy": 5.640048980712891,
      "epoch": 0.23102785782901056,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997940853953354,
      "loss": 5.5985,
      "mean_token_accuracy": 0.16914743930101395,
      "num_tokens": 5529626.0,
      "step": 2405
    },
    {
      "entropy": 5.701624059677124,
      "epoch": 0.23150816522574447,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499792616381322,
      "loss": 5.6212,
      "mean_token_accuracy": 0.17812359482049941,
      "num_tokens": 5540844.0,
      "step": 2410
    },
    {
      "entropy": 5.684215545654297,
      "epoch": 0.23198847262247838,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997911421482914,
      "loss": 5.5587,
      "mean_token_accuracy": 0.17999959141016006,
      "num_tokens": 5551732.0,
      "step": 2415
    },
    {
      "entropy": 5.6654558181762695,
      "epoch": 0.23246878001921228,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499789662696278,
      "loss": 5.7098,
      "mean_token_accuracy": 0.17055163383483887,
      "num_tokens": 5562692.0,
      "step": 2420
    },
    {
      "entropy": 5.7184652328491214,
      "epoch": 0.2329490874159462,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997881780253162,
      "loss": 5.6944,
      "mean_token_accuracy": 0.1724271610379219,
      "num_tokens": 5574884.0,
      "step": 2425
    },
    {
      "entropy": 5.526378154754639,
      "epoch": 0.2334293948126801,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997866881354403,
      "loss": 5.5406,
      "mean_token_accuracy": 0.17680294513702394,
      "num_tokens": 5585335.0,
      "step": 2430
    },
    {
      "entropy": 5.715866374969482,
      "epoch": 0.23390970220941404,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499785193026685,
      "loss": 5.7062,
      "mean_token_accuracy": 0.16400452703237534,
      "num_tokens": 5597621.0,
      "step": 2435
    },
    {
      "entropy": 5.559733724594116,
      "epoch": 0.23439000960614795,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997836926990851,
      "loss": 5.5486,
      "mean_token_accuracy": 0.17132115513086318,
      "num_tokens": 5609795.0,
      "step": 2440
    },
    {
      "entropy": 5.626978397369385,
      "epoch": 0.23487031700288186,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997821871526752,
      "loss": 5.6476,
      "mean_token_accuracy": 0.16910365521907805,
      "num_tokens": 5621473.0,
      "step": 2445
    },
    {
      "entropy": 5.741739797592163,
      "epoch": 0.23535062439961577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997806763874905,
      "loss": 5.7391,
      "mean_token_accuracy": 0.16913674920797347,
      "num_tokens": 5635064.0,
      "step": 2450
    },
    {
      "entropy": 5.771869468688965,
      "epoch": 0.23583093179634967,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997791604035659,
      "loss": 5.7399,
      "mean_token_accuracy": 0.16541620939970017,
      "num_tokens": 5647382.0,
      "step": 2455
    },
    {
      "entropy": 5.632869815826416,
      "epoch": 0.23631123919308358,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997776392009366,
      "loss": 5.6043,
      "mean_token_accuracy": 0.17495698034763335,
      "num_tokens": 5658417.0,
      "step": 2460
    },
    {
      "entropy": 5.617389917373657,
      "epoch": 0.2367915465898175,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997761127796381,
      "loss": 5.5897,
      "mean_token_accuracy": 0.174434295296669,
      "num_tokens": 5672190.0,
      "step": 2465
    },
    {
      "entropy": 5.6352379322052,
      "epoch": 0.2372718539865514,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997745811397056,
      "loss": 5.6077,
      "mean_token_accuracy": 0.1803595706820488,
      "num_tokens": 5684259.0,
      "step": 2470
    },
    {
      "entropy": 5.635552072525025,
      "epoch": 0.2377521613832853,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997730442811748,
      "loss": 5.6732,
      "mean_token_accuracy": 0.1712580442428589,
      "num_tokens": 5696505.0,
      "step": 2475
    },
    {
      "entropy": 5.591545057296753,
      "epoch": 0.23823246878001922,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997715022040814,
      "loss": 5.546,
      "mean_token_accuracy": 0.17534102201461793,
      "num_tokens": 5707638.0,
      "step": 2480
    },
    {
      "entropy": 5.5858382225036625,
      "epoch": 0.23871277617675313,
      "grad_norm": 1.0,
      "learning_rate": 0.000499769954908461,
      "loss": 5.532,
      "mean_token_accuracy": 0.181480710208416,
      "num_tokens": 5719098.0,
      "step": 2485
    },
    {
      "entropy": 5.7313658714294435,
      "epoch": 0.23919308357348704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997684023943498,
      "loss": 5.7639,
      "mean_token_accuracy": 0.15622588247060776,
      "num_tokens": 5731345.0,
      "step": 2490
    },
    {
      "entropy": 5.546728086471558,
      "epoch": 0.23967339097022095,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997668446617837,
      "loss": 5.5085,
      "mean_token_accuracy": 0.17273759096860886,
      "num_tokens": 5742604.0,
      "step": 2495
    },
    {
      "entropy": 5.542533683776855,
      "epoch": 0.24015369836695485,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997652817107989,
      "loss": 5.549,
      "mean_token_accuracy": 0.17851854413747786,
      "num_tokens": 5754620.0,
      "step": 2500
    },
    {
      "entropy": 5.704870271682739,
      "epoch": 0.24063400576368876,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997637135414315,
      "loss": 5.6163,
      "mean_token_accuracy": 0.17889242321252824,
      "num_tokens": 5764764.0,
      "step": 2505
    },
    {
      "entropy": 5.53765640258789,
      "epoch": 0.24111431316042267,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997621401537183,
      "loss": 5.5202,
      "mean_token_accuracy": 0.18003057241439818,
      "num_tokens": 5776341.0,
      "step": 2510
    },
    {
      "entropy": 5.672939825057983,
      "epoch": 0.24159462055715658,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997605615476955,
      "loss": 5.743,
      "mean_token_accuracy": 0.16112136393785476,
      "num_tokens": 5788604.0,
      "step": 2515
    },
    {
      "entropy": 5.670176792144775,
      "epoch": 0.2420749279538905,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004997589777234,
      "loss": 5.62,
      "mean_token_accuracy": 0.17324668765068055,
      "num_tokens": 5800194.0,
      "step": 2520
    },
    {
      "entropy": 5.625196886062622,
      "epoch": 0.2425552353506244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997573886808684,
      "loss": 5.6197,
      "mean_token_accuracy": 0.17728133350610734,
      "num_tokens": 5812978.0,
      "step": 2525
    },
    {
      "entropy": 5.587188100814819,
      "epoch": 0.2430355427473583,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997557944201375,
      "loss": 5.5326,
      "mean_token_accuracy": 0.18073128014802933,
      "num_tokens": 5824591.0,
      "step": 2530
    },
    {
      "entropy": 5.583196496963501,
      "epoch": 0.24351585014409222,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997541949412445,
      "loss": 5.6288,
      "mean_token_accuracy": 0.1718851387500763,
      "num_tokens": 5835970.0,
      "step": 2535
    },
    {
      "entropy": 5.70057430267334,
      "epoch": 0.24399615754082613,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997525902442266,
      "loss": 5.6803,
      "mean_token_accuracy": 0.1663990870118141,
      "num_tokens": 5848687.0,
      "step": 2540
    },
    {
      "entropy": 5.654277038574219,
      "epoch": 0.24447646493756003,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997509803291207,
      "loss": 5.6426,
      "mean_token_accuracy": 0.1717447817325592,
      "num_tokens": 5860820.0,
      "step": 2545
    },
    {
      "entropy": 5.606211185455322,
      "epoch": 0.24495677233429394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997493651959647,
      "loss": 5.579,
      "mean_token_accuracy": 0.1743954598903656,
      "num_tokens": 5872172.0,
      "step": 2550
    },
    {
      "entropy": 5.641741323471069,
      "epoch": 0.24543707973102785,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004997477448447955,
      "loss": 5.6171,
      "mean_token_accuracy": 0.17213639616966248,
      "num_tokens": 5883574.0,
      "step": 2555
    },
    {
      "entropy": 5.679425430297852,
      "epoch": 0.24591738712776176,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997461192756512,
      "loss": 5.7187,
      "mean_token_accuracy": 0.1697825610637665,
      "num_tokens": 5895223.0,
      "step": 2560
    },
    {
      "entropy": 5.592727375030518,
      "epoch": 0.24639769452449567,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997444884885694,
      "loss": 5.5358,
      "mean_token_accuracy": 0.17672908753156663,
      "num_tokens": 5907448.0,
      "step": 2565
    },
    {
      "entropy": 5.6109534740448,
      "epoch": 0.24687800192122958,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997428524835879,
      "loss": 5.6404,
      "mean_token_accuracy": 0.17392444908618926,
      "num_tokens": 5918906.0,
      "step": 2570
    },
    {
      "entropy": 5.628335285186767,
      "epoch": 0.2473583093179635,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997412112607446,
      "loss": 5.515,
      "mean_token_accuracy": 0.18007228374481202,
      "num_tokens": 5930619.0,
      "step": 2575
    },
    {
      "entropy": 5.718863439559937,
      "epoch": 0.2478386167146974,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004997395648200778,
      "loss": 5.7062,
      "mean_token_accuracy": 0.16648458540439606,
      "num_tokens": 5941554.0,
      "step": 2580
    },
    {
      "entropy": 5.687685155868531,
      "epoch": 0.2483189241114313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997379131616257,
      "loss": 5.6831,
      "mean_token_accuracy": 0.17195172160863875,
      "num_tokens": 5953309.0,
      "step": 2585
    },
    {
      "entropy": 5.610647296905517,
      "epoch": 0.24879923150816521,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004997362562854266,
      "loss": 5.6294,
      "mean_token_accuracy": 0.17338041216135025,
      "num_tokens": 5965286.0,
      "step": 2590
    },
    {
      "entropy": 5.6528027057647705,
      "epoch": 0.24927953890489912,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004997345941915187,
      "loss": 5.5986,
      "mean_token_accuracy": 0.17616538554430008,
      "num_tokens": 5976558.0,
      "step": 2595
    },
    {
      "entropy": 5.646711730957032,
      "epoch": 0.24975984630163303,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004997329268799412,
      "loss": 5.6446,
      "mean_token_accuracy": 0.16481584012508393,
      "num_tokens": 5988793.0,
      "step": 2600
    },
    {
      "entropy": 5.680540704727173,
      "epoch": 0.25024015369836694,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004997312543507322,
      "loss": 5.6778,
      "mean_token_accuracy": 0.16258761882781983,
      "num_tokens": 5999825.0,
      "step": 2605
    },
    {
      "entropy": 5.660827779769898,
      "epoch": 0.2507204610951009,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997295766039309,
      "loss": 5.6168,
      "mean_token_accuracy": 0.1700749561190605,
      "num_tokens": 6010408.0,
      "step": 2610
    },
    {
      "entropy": 5.694224882125854,
      "epoch": 0.25120076849183476,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004997278936395761,
      "loss": 5.6605,
      "mean_token_accuracy": 0.1695483595132828,
      "num_tokens": 6021161.0,
      "step": 2615
    },
    {
      "entropy": 5.579520511627197,
      "epoch": 0.2516810758885687,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997262054577071,
      "loss": 5.4949,
      "mean_token_accuracy": 0.1799078106880188,
      "num_tokens": 6032646.0,
      "step": 2620
    },
    {
      "entropy": 5.61772780418396,
      "epoch": 0.2521613832853026,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997245120583627,
      "loss": 5.6832,
      "mean_token_accuracy": 0.17070280015468597,
      "num_tokens": 6045019.0,
      "step": 2625
    },
    {
      "entropy": 5.547396659851074,
      "epoch": 0.2526416906820365,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997228134415825,
      "loss": 5.5513,
      "mean_token_accuracy": 0.17881007641553878,
      "num_tokens": 6055644.0,
      "step": 2630
    },
    {
      "entropy": 5.676703310012817,
      "epoch": 0.2531219980787704,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997211096074059,
      "loss": 5.5686,
      "mean_token_accuracy": 0.1793759658932686,
      "num_tokens": 6066488.0,
      "step": 2635
    },
    {
      "entropy": 5.61687445640564,
      "epoch": 0.25360230547550433,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997194005558722,
      "loss": 5.6107,
      "mean_token_accuracy": 0.17230986505746843,
      "num_tokens": 6078300.0,
      "step": 2640
    },
    {
      "entropy": 5.612687253952027,
      "epoch": 0.2540826128722382,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004997176862870216,
      "loss": 5.5767,
      "mean_token_accuracy": 0.17045443654060363,
      "num_tokens": 6089180.0,
      "step": 2645
    },
    {
      "entropy": 5.659189224243164,
      "epoch": 0.25456292026897215,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997159668008933,
      "loss": 5.6136,
      "mean_token_accuracy": 0.18232642710208893,
      "num_tokens": 6099093.0,
      "step": 2650
    },
    {
      "entropy": 5.612885761260986,
      "epoch": 0.25504322766570603,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004997142420975277,
      "loss": 5.6459,
      "mean_token_accuracy": 0.1759412705898285,
      "num_tokens": 6109161.0,
      "step": 2655
    },
    {
      "entropy": 5.7561407566070555,
      "epoch": 0.25552353506243997,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997125121769647,
      "loss": 5.7539,
      "mean_token_accuracy": 0.16455692201852798,
      "num_tokens": 6120031.0,
      "step": 2660
    },
    {
      "entropy": 5.593028593063354,
      "epoch": 0.25600384245917385,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997107770392444,
      "loss": 5.6017,
      "mean_token_accuracy": 0.17366934418678284,
      "num_tokens": 6132398.0,
      "step": 2665
    },
    {
      "entropy": 5.646261167526245,
      "epoch": 0.2564841498559078,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499709036684407,
      "loss": 5.5896,
      "mean_token_accuracy": 0.16633834168314934,
      "num_tokens": 6143099.0,
      "step": 2670
    },
    {
      "entropy": 5.615106105804443,
      "epoch": 0.25696445725264166,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004997072911124932,
      "loss": 5.5636,
      "mean_token_accuracy": 0.17667461335659027,
      "num_tokens": 6154815.0,
      "step": 2675
    },
    {
      "entropy": 5.633945465087891,
      "epoch": 0.2574447646493756,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004997055403235432,
      "loss": 5.6097,
      "mean_token_accuracy": 0.17703237533569335,
      "num_tokens": 6165696.0,
      "step": 2680
    },
    {
      "entropy": 5.583321189880371,
      "epoch": 0.2579250720461095,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997037843175978,
      "loss": 5.6479,
      "mean_token_accuracy": 0.1697259709239006,
      "num_tokens": 6177042.0,
      "step": 2685
    },
    {
      "entropy": 5.723847389221191,
      "epoch": 0.2584053794428434,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997020230946978,
      "loss": 5.6299,
      "mean_token_accuracy": 0.1708068698644638,
      "num_tokens": 6188691.0,
      "step": 2690
    },
    {
      "entropy": 5.480944967269897,
      "epoch": 0.25888568683957736,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004997002566548841,
      "loss": 5.4628,
      "mean_token_accuracy": 0.18257904648780823,
      "num_tokens": 6199931.0,
      "step": 2695
    },
    {
      "entropy": 5.554953145980835,
      "epoch": 0.25936599423631124,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996984849981976,
      "loss": 5.5978,
      "mean_token_accuracy": 0.18077756613492965,
      "num_tokens": 6211493.0,
      "step": 2700
    },
    {
      "entropy": 5.7185674667358395,
      "epoch": 0.2598463016330452,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004996967081246794,
      "loss": 5.5288,
      "mean_token_accuracy": 0.17527295053005218,
      "num_tokens": 6221505.0,
      "step": 2705
    },
    {
      "entropy": 5.5221052169799805,
      "epoch": 0.26032660902977905,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996949260343711,
      "loss": 5.5012,
      "mean_token_accuracy": 0.1818551629781723,
      "num_tokens": 6233058.0,
      "step": 2710
    },
    {
      "entropy": 5.628518390655517,
      "epoch": 0.260806916426513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996931387273137,
      "loss": 5.6465,
      "mean_token_accuracy": 0.17151581346988679,
      "num_tokens": 6244713.0,
      "step": 2715
    },
    {
      "entropy": 5.612377023696899,
      "epoch": 0.2612872238232469,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996913462035487,
      "loss": 5.6046,
      "mean_token_accuracy": 0.16954844444990158,
      "num_tokens": 6256624.0,
      "step": 2720
    },
    {
      "entropy": 5.582393217086792,
      "epoch": 0.2617675312199808,
      "grad_norm": 1.0,
      "learning_rate": 0.000499689548463118,
      "loss": 5.6041,
      "mean_token_accuracy": 0.16892476230859757,
      "num_tokens": 6271134.0,
      "step": 2725
    },
    {
      "entropy": 5.718617343902588,
      "epoch": 0.2622478386167147,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996877455060631,
      "loss": 5.698,
      "mean_token_accuracy": 0.17103526890277862,
      "num_tokens": 6282082.0,
      "step": 2730
    },
    {
      "entropy": 5.666055965423584,
      "epoch": 0.2627281460134486,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996859373324259,
      "loss": 5.6319,
      "mean_token_accuracy": 0.17113145142793657,
      "num_tokens": 6293869.0,
      "step": 2735
    },
    {
      "entropy": 5.63692798614502,
      "epoch": 0.2632084534101825,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004996841239422485,
      "loss": 5.6633,
      "mean_token_accuracy": 0.16283004432916642,
      "num_tokens": 6304858.0,
      "step": 2740
    },
    {
      "entropy": 5.64309549331665,
      "epoch": 0.26368876080691644,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996823053355729,
      "loss": 5.5963,
      "mean_token_accuracy": 0.17556946128606796,
      "num_tokens": 6316738.0,
      "step": 2745
    },
    {
      "entropy": 5.568399381637573,
      "epoch": 0.2641690682036503,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996804815124413,
      "loss": 5.556,
      "mean_token_accuracy": 0.17209173142910003,
      "num_tokens": 6327691.0,
      "step": 2750
    },
    {
      "entropy": 5.66407265663147,
      "epoch": 0.26464937560038426,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004996786524728962,
      "loss": 5.5718,
      "mean_token_accuracy": 0.16894693523645402,
      "num_tokens": 6337964.0,
      "step": 2755
    },
    {
      "entropy": 5.56714882850647,
      "epoch": 0.26512968299711814,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996768182169797,
      "loss": 5.6163,
      "mean_token_accuracy": 0.17449129968881608,
      "num_tokens": 6349764.0,
      "step": 2760
    },
    {
      "entropy": 5.662902736663819,
      "epoch": 0.2656099903938521,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004996749787447349,
      "loss": 5.5711,
      "mean_token_accuracy": 0.17478797137737273,
      "num_tokens": 6360538.0,
      "step": 2765
    },
    {
      "entropy": 5.595956087112427,
      "epoch": 0.26609029779058596,
      "grad_norm": 1.125,
      "learning_rate": 0.000499673134056204,
      "loss": 5.6144,
      "mean_token_accuracy": 0.17373100519180298,
      "num_tokens": 6372489.0,
      "step": 2770
    },
    {
      "entropy": 5.443460750579834,
      "epoch": 0.2665706051873199,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996712841514303,
      "loss": 5.4845,
      "mean_token_accuracy": 0.1831083118915558,
      "num_tokens": 6383080.0,
      "step": 2775
    },
    {
      "entropy": 5.692691278457642,
      "epoch": 0.2670509125840538,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996694290304563,
      "loss": 5.6041,
      "mean_token_accuracy": 0.17385849058628083,
      "num_tokens": 6394033.0,
      "step": 2780
    },
    {
      "entropy": 5.572810697555542,
      "epoch": 0.2675312199807877,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996675686933255,
      "loss": 5.611,
      "mean_token_accuracy": 0.17484720796346664,
      "num_tokens": 6404990.0,
      "step": 2785
    },
    {
      "entropy": 5.576703500747681,
      "epoch": 0.2680115273775216,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996657031400807,
      "loss": 5.4782,
      "mean_token_accuracy": 0.181576107442379,
      "num_tokens": 6416337.0,
      "step": 2790
    },
    {
      "entropy": 5.643291425704956,
      "epoch": 0.26849183477425553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996638323707655,
      "loss": 5.6961,
      "mean_token_accuracy": 0.17010248452425003,
      "num_tokens": 6427892.0,
      "step": 2795
    },
    {
      "entropy": 5.650494956970215,
      "epoch": 0.2689721421709894,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996619563854232,
      "loss": 5.596,
      "mean_token_accuracy": 0.1760585606098175,
      "num_tokens": 6440300.0,
      "step": 2800
    },
    {
      "entropy": 5.595854425430298,
      "epoch": 0.26945244956772335,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996600751840974,
      "loss": 5.4853,
      "mean_token_accuracy": 0.17765479534864426,
      "num_tokens": 6451499.0,
      "step": 2805
    },
    {
      "entropy": 5.594804906845093,
      "epoch": 0.26993275696445723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996581887668317,
      "loss": 5.5832,
      "mean_token_accuracy": 0.17601191103458405,
      "num_tokens": 6462882.0,
      "step": 2810
    },
    {
      "entropy": 5.572175931930542,
      "epoch": 0.27041306436119117,
      "grad_norm": 1.203125,
      "learning_rate": 0.00049965629713367,
      "loss": 5.5569,
      "mean_token_accuracy": 0.17694126069545746,
      "num_tokens": 6475042.0,
      "step": 2815
    },
    {
      "entropy": 5.570115184783935,
      "epoch": 0.27089337175792505,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996544002846561,
      "loss": 5.5648,
      "mean_token_accuracy": 0.17592734843492508,
      "num_tokens": 6486245.0,
      "step": 2820
    },
    {
      "entropy": 5.579746532440185,
      "epoch": 0.271373679154659,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996524982198343,
      "loss": 5.4745,
      "mean_token_accuracy": 0.178873273730278,
      "num_tokens": 6497705.0,
      "step": 2825
    },
    {
      "entropy": 5.532494258880615,
      "epoch": 0.27185398655139287,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004996505909392485,
      "loss": 5.5101,
      "mean_token_accuracy": 0.18047289401292801,
      "num_tokens": 6508193.0,
      "step": 2830
    },
    {
      "entropy": 5.649497365951538,
      "epoch": 0.2723342939481268,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004996486784429429,
      "loss": 5.6201,
      "mean_token_accuracy": 0.17633747458457946,
      "num_tokens": 6519228.0,
      "step": 2835
    },
    {
      "entropy": 5.627618598937988,
      "epoch": 0.2728146013448607,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004996467607309622,
      "loss": 5.587,
      "mean_token_accuracy": 0.1716805502772331,
      "num_tokens": 6530743.0,
      "step": 2840
    },
    {
      "entropy": 5.575333595275879,
      "epoch": 0.2732949087415946,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996448378033507,
      "loss": 5.4738,
      "mean_token_accuracy": 0.17245458513498307,
      "num_tokens": 6542588.0,
      "step": 2845
    },
    {
      "entropy": 5.695534610748291,
      "epoch": 0.2737752161383285,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996429096601532,
      "loss": 5.674,
      "mean_token_accuracy": 0.16420495510101318,
      "num_tokens": 6554002.0,
      "step": 2850
    },
    {
      "entropy": 5.686659669876098,
      "epoch": 0.27425552353506244,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004996409763014144,
      "loss": 5.6925,
      "mean_token_accuracy": 0.1730800151824951,
      "num_tokens": 6563358.0,
      "step": 2855
    },
    {
      "entropy": 5.598821449279785,
      "epoch": 0.2747358309317964,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004996390377271791,
      "loss": 5.5962,
      "mean_token_accuracy": 0.16940404772758483,
      "num_tokens": 6574986.0,
      "step": 2860
    },
    {
      "entropy": 5.531180191040039,
      "epoch": 0.27521613832853026,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004996370939374924,
      "loss": 5.5498,
      "mean_token_accuracy": 0.17848169654607773,
      "num_tokens": 6587407.0,
      "step": 2865
    },
    {
      "entropy": 5.562257719039917,
      "epoch": 0.2756964457252642,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004996351449323994,
      "loss": 5.5087,
      "mean_token_accuracy": 0.18236653059720992,
      "num_tokens": 6598483.0,
      "step": 2870
    },
    {
      "entropy": 5.533087635040284,
      "epoch": 0.2761767531219981,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996331907119455,
      "loss": 5.5975,
      "mean_token_accuracy": 0.1755070611834526,
      "num_tokens": 6610749.0,
      "step": 2875
    },
    {
      "entropy": 5.659117031097412,
      "epoch": 0.276657060518732,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004996312312761758,
      "loss": 5.5745,
      "mean_token_accuracy": 0.17036049515008928,
      "num_tokens": 6621762.0,
      "step": 2880
    },
    {
      "entropy": 5.612740898132325,
      "epoch": 0.2771373679154659,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000499629266625136,
      "loss": 5.6441,
      "mean_token_accuracy": 0.17175253927707673,
      "num_tokens": 6632296.0,
      "step": 2885
    },
    {
      "entropy": 5.629729986190796,
      "epoch": 0.27761767531219983,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996272967588715,
      "loss": 5.5947,
      "mean_token_accuracy": 0.17308499813079833,
      "num_tokens": 6644413.0,
      "step": 2890
    },
    {
      "entropy": 5.616733264923096,
      "epoch": 0.2780979827089337,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004996253216774283,
      "loss": 5.5385,
      "mean_token_accuracy": 0.17536325752735138,
      "num_tokens": 6655634.0,
      "step": 2895
    },
    {
      "entropy": 5.566686773300171,
      "epoch": 0.27857829010566765,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004996233413808521,
      "loss": 5.4757,
      "mean_token_accuracy": 0.18005252480506898,
      "num_tokens": 6665977.0,
      "step": 2900
    },
    {
      "entropy": 5.589098834991455,
      "epoch": 0.27905859750240153,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996213558691889,
      "loss": 5.5451,
      "mean_token_accuracy": 0.16986947655677795,
      "num_tokens": 6677165.0,
      "step": 2905
    },
    {
      "entropy": 5.5928606510162355,
      "epoch": 0.27953890489913547,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996193651424848,
      "loss": 5.6578,
      "mean_token_accuracy": 0.17219153493642808,
      "num_tokens": 6688182.0,
      "step": 2910
    },
    {
      "entropy": 5.562531757354736,
      "epoch": 0.28001921229586935,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499617369200786,
      "loss": 5.5547,
      "mean_token_accuracy": 0.17558752298355101,
      "num_tokens": 6699259.0,
      "step": 2915
    },
    {
      "entropy": 5.630407905578613,
      "epoch": 0.2804995196926033,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996153680441389,
      "loss": 5.5647,
      "mean_token_accuracy": 0.17240963131189346,
      "num_tokens": 6710071.0,
      "step": 2920
    },
    {
      "entropy": 5.601401424407959,
      "epoch": 0.28097982708933716,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00049961336167259,
      "loss": 5.61,
      "mean_token_accuracy": 0.17204567939043044,
      "num_tokens": 6720267.0,
      "step": 2925
    },
    {
      "entropy": 5.683440780639648,
      "epoch": 0.2814601344860711,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004996113500861857,
      "loss": 5.5973,
      "mean_token_accuracy": 0.16914477646350862,
      "num_tokens": 6730919.0,
      "step": 2930
    },
    {
      "entropy": 5.553895092010498,
      "epoch": 0.281940441882805,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996093332849729,
      "loss": 5.5829,
      "mean_token_accuracy": 0.17506259977817534,
      "num_tokens": 6741480.0,
      "step": 2935
    },
    {
      "entropy": 5.604211473464966,
      "epoch": 0.2824207492795389,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004996073112689983,
      "loss": 5.5304,
      "mean_token_accuracy": 0.1787044510245323,
      "num_tokens": 6751816.0,
      "step": 2940
    },
    {
      "entropy": 5.535765361785889,
      "epoch": 0.2829010566762728,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004996052840383088,
      "loss": 5.5574,
      "mean_token_accuracy": 0.17378600239753722,
      "num_tokens": 6763645.0,
      "step": 2945
    },
    {
      "entropy": 5.56224799156189,
      "epoch": 0.28338136407300674,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004996032515929516,
      "loss": 5.5275,
      "mean_token_accuracy": 0.17407883554697037,
      "num_tokens": 6774795.0,
      "step": 2950
    },
    {
      "entropy": 5.566365337371826,
      "epoch": 0.2838616714697406,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996012139329738,
      "loss": 5.4567,
      "mean_token_accuracy": 0.18206687271595,
      "num_tokens": 6786833.0,
      "step": 2955
    },
    {
      "entropy": 5.460562705993652,
      "epoch": 0.28434197886647455,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004995991710584228,
      "loss": 5.4785,
      "mean_token_accuracy": 0.17664896845817565,
      "num_tokens": 6798134.0,
      "step": 2960
    },
    {
      "entropy": 5.614284372329712,
      "epoch": 0.28482228626320844,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004995971229693459,
      "loss": 5.5657,
      "mean_token_accuracy": 0.1824972078204155,
      "num_tokens": 6807918.0,
      "step": 2965
    },
    {
      "entropy": 5.6419624328613285,
      "epoch": 0.28530259365994237,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004995950696657909,
      "loss": 5.6505,
      "mean_token_accuracy": 0.17275601327419282,
      "num_tokens": 6818689.0,
      "step": 2970
    },
    {
      "entropy": 5.5683265209198,
      "epoch": 0.28578290105667625,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004995930111478051,
      "loss": 5.5226,
      "mean_token_accuracy": 0.17747195959091186,
      "num_tokens": 6830871.0,
      "step": 2975
    },
    {
      "entropy": 5.584042739868164,
      "epoch": 0.2862632084534102,
      "grad_norm": 1.25,
      "learning_rate": 0.0004995909474154365,
      "loss": 5.5118,
      "mean_token_accuracy": 0.18256898522377013,
      "num_tokens": 6842857.0,
      "step": 2980
    },
    {
      "entropy": 5.512867593765259,
      "epoch": 0.28674351585014407,
      "grad_norm": 1.25,
      "learning_rate": 0.0004995888784687331,
      "loss": 5.5396,
      "mean_token_accuracy": 0.1776072397828102,
      "num_tokens": 6853727.0,
      "step": 2985
    },
    {
      "entropy": 5.5820722579956055,
      "epoch": 0.287223823246878,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004995868043077428,
      "loss": 5.5806,
      "mean_token_accuracy": 0.17571521997451783,
      "num_tokens": 6864728.0,
      "step": 2990
    },
    {
      "entropy": 5.627726793289185,
      "epoch": 0.2877041306436119,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004995847249325137,
      "loss": 5.6271,
      "mean_token_accuracy": 0.17045981287956238,
      "num_tokens": 6875001.0,
      "step": 2995
    },
    {
      "entropy": 5.526087427139283,
      "epoch": 0.2881844380403458,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004995826403430942,
      "loss": 5.5175,
      "mean_token_accuracy": 0.18246698677539824,
      "num_tokens": 6887026.0,
      "step": 3000
    },
    {
      "epoch": 0.2881844380403458,
      "eval_entropy": 5.487682455286986,
      "eval_loss": 5.587414264678955,
      "eval_mean_token_accuracy": 0.18148705224778552,
      "eval_num_tokens": 6887026.0,
      "eval_runtime": 27.0434,
      "eval_samples_per_second": 1213.422,
      "eval_steps_per_second": 151.682,
      "step": 3000
    },
    {
      "entropy": 5.784534358978272,
      "epoch": 0.2886647454370797,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004995805505395328,
      "loss": 5.634,
      "mean_token_accuracy": 0.17297943606972693,
      "num_tokens": 6900118.0,
      "step": 3005
    },
    {
      "entropy": 5.592379331588745,
      "epoch": 0.28914505283381364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995784555218778,
      "loss": 5.5457,
      "mean_token_accuracy": 0.177190899848938,
      "num_tokens": 6912135.0,
      "step": 3010
    },
    {
      "entropy": 5.497978210449219,
      "epoch": 0.2896253602305475,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004995763552901779,
      "loss": 5.4652,
      "mean_token_accuracy": 0.1853455811738968,
      "num_tokens": 6923781.0,
      "step": 3015
    },
    {
      "entropy": 5.664725828170776,
      "epoch": 0.29010566762728146,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995742498444818,
      "loss": 5.5343,
      "mean_token_accuracy": 0.17282128781080247,
      "num_tokens": 6935804.0,
      "step": 3020
    },
    {
      "entropy": 5.571994972229004,
      "epoch": 0.2905859750240154,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995721391848387,
      "loss": 5.5897,
      "mean_token_accuracy": 0.17708188891410828,
      "num_tokens": 6948289.0,
      "step": 3025
    },
    {
      "entropy": 5.632468414306641,
      "epoch": 0.2910662824207493,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004995700233112972,
      "loss": 5.5201,
      "mean_token_accuracy": 0.1799446865916252,
      "num_tokens": 6958719.0,
      "step": 3030
    },
    {
      "entropy": 5.5489743709564205,
      "epoch": 0.2915465898174832,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995679022239066,
      "loss": 5.5794,
      "mean_token_accuracy": 0.18038478195667268,
      "num_tokens": 6969041.0,
      "step": 3035
    },
    {
      "entropy": 5.5848000049591064,
      "epoch": 0.2920268972142171,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004995657759227162,
      "loss": 5.6493,
      "mean_token_accuracy": 0.17737395614385604,
      "num_tokens": 6981206.0,
      "step": 3040
    },
    {
      "entropy": 5.696254730224609,
      "epoch": 0.29250720461095103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995636444077751,
      "loss": 5.538,
      "mean_token_accuracy": 0.1819950446486473,
      "num_tokens": 6993169.0,
      "step": 3045
    },
    {
      "entropy": 5.5612489700317385,
      "epoch": 0.2929875120076849,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995615076791333,
      "loss": 5.6105,
      "mean_token_accuracy": 0.17608709931373595,
      "num_tokens": 7005175.0,
      "step": 3050
    },
    {
      "entropy": 5.648107385635376,
      "epoch": 0.29346781940441885,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995593657368399,
      "loss": 5.6127,
      "mean_token_accuracy": 0.17430707216262817,
      "num_tokens": 7016738.0,
      "step": 3055
    },
    {
      "entropy": 5.608227968215942,
      "epoch": 0.29394812680115273,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499557218580945,
      "loss": 5.5006,
      "mean_token_accuracy": 0.18451340198516847,
      "num_tokens": 7028399.0,
      "step": 3060
    },
    {
      "entropy": 5.553318691253662,
      "epoch": 0.29442843419788667,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004995550662114981,
      "loss": 5.6546,
      "mean_token_accuracy": 0.17278157025575638,
      "num_tokens": 7039528.0,
      "step": 3065
    },
    {
      "entropy": 5.479732799530029,
      "epoch": 0.29490874159462055,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995529086285495,
      "loss": 5.434,
      "mean_token_accuracy": 0.18942976742982864,
      "num_tokens": 7051399.0,
      "step": 3070
    },
    {
      "entropy": 5.4366058826446535,
      "epoch": 0.2953890489913545,
      "grad_norm": 1.296875,
      "learning_rate": 0.000499550745832149,
      "loss": 5.4343,
      "mean_token_accuracy": 0.18675332069396972,
      "num_tokens": 7062452.0,
      "step": 3075
    },
    {
      "entropy": 5.553073835372925,
      "epoch": 0.29586935638808837,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004995485778223471,
      "loss": 5.5675,
      "mean_token_accuracy": 0.17742615193128586,
      "num_tokens": 7074052.0,
      "step": 3080
    },
    {
      "entropy": 5.721508979797363,
      "epoch": 0.2963496637848223,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004995464045991939,
      "loss": 5.6161,
      "mean_token_accuracy": 0.17061588913202286,
      "num_tokens": 7086692.0,
      "step": 3085
    },
    {
      "entropy": 5.601110982894897,
      "epoch": 0.2968299711815562,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00049954422616274,
      "loss": 5.5736,
      "mean_token_accuracy": 0.1758839413523674,
      "num_tokens": 7098221.0,
      "step": 3090
    },
    {
      "entropy": 5.455524539947509,
      "epoch": 0.2973102785782901,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004995420425130359,
      "loss": 5.5355,
      "mean_token_accuracy": 0.182497076690197,
      "num_tokens": 7110129.0,
      "step": 3095
    },
    {
      "entropy": 5.684749460220337,
      "epoch": 0.297790585975024,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004995398536501324,
      "loss": 5.6159,
      "mean_token_accuracy": 0.17379121631383895,
      "num_tokens": 7121523.0,
      "step": 3100
    },
    {
      "entropy": 5.6178731441497805,
      "epoch": 0.29827089337175794,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995376595740801,
      "loss": 5.543,
      "mean_token_accuracy": 0.18315426707267762,
      "num_tokens": 7133482.0,
      "step": 3105
    },
    {
      "entropy": 5.5218054294586185,
      "epoch": 0.2987512007684918,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004995354602849302,
      "loss": 5.5217,
      "mean_token_accuracy": 0.18059220016002656,
      "num_tokens": 7144031.0,
      "step": 3110
    },
    {
      "entropy": 5.6512165546417235,
      "epoch": 0.29923150816522576,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995332557827337,
      "loss": 5.6016,
      "mean_token_accuracy": 0.1730698049068451,
      "num_tokens": 7155469.0,
      "step": 3115
    },
    {
      "entropy": 5.630453109741211,
      "epoch": 0.29971181556195964,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004995310460675416,
      "loss": 5.5397,
      "mean_token_accuracy": 0.1743144854903221,
      "num_tokens": 7167077.0,
      "step": 3120
    },
    {
      "entropy": 5.560130977630616,
      "epoch": 0.3001921229586936,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004995288311394053,
      "loss": 5.4868,
      "mean_token_accuracy": 0.1782015010714531,
      "num_tokens": 7178917.0,
      "step": 3125
    },
    {
      "entropy": 5.596506404876709,
      "epoch": 0.30067243035542746,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995266109983764,
      "loss": 5.5632,
      "mean_token_accuracy": 0.176068215072155,
      "num_tokens": 7190284.0,
      "step": 3130
    },
    {
      "entropy": 5.533182668685913,
      "epoch": 0.3011527377521614,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004995243856445062,
      "loss": 5.5829,
      "mean_token_accuracy": 0.1797621876001358,
      "num_tokens": 7202391.0,
      "step": 3135
    },
    {
      "entropy": 5.574808263778687,
      "epoch": 0.3016330451488953,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004995221550778466,
      "loss": 5.4884,
      "mean_token_accuracy": 0.18022829294204712,
      "num_tokens": 7212882.0,
      "step": 3140
    },
    {
      "entropy": 5.509401130676269,
      "epoch": 0.3021133525456292,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995199192984491,
      "loss": 5.4847,
      "mean_token_accuracy": 0.18269667625427247,
      "num_tokens": 7224102.0,
      "step": 3145
    },
    {
      "entropy": 5.554200077056885,
      "epoch": 0.3025936599423631,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004995176783063657,
      "loss": 5.4433,
      "mean_token_accuracy": 0.17980438470840454,
      "num_tokens": 7234579.0,
      "step": 3150
    },
    {
      "entropy": 5.574840784072876,
      "epoch": 0.30307396733909703,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004995154321016487,
      "loss": 5.5879,
      "mean_token_accuracy": 0.17145463675260544,
      "num_tokens": 7246065.0,
      "step": 3155
    },
    {
      "entropy": 5.650038957595825,
      "epoch": 0.3035542747358309,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004995131806843499,
      "loss": 5.5459,
      "mean_token_accuracy": 0.16908762753009796,
      "num_tokens": 7258899.0,
      "step": 3160
    },
    {
      "entropy": 5.630798196792602,
      "epoch": 0.30403458213256485,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995109240545218,
      "loss": 5.6372,
      "mean_token_accuracy": 0.15845271199941635,
      "num_tokens": 7271887.0,
      "step": 3165
    },
    {
      "entropy": 5.6133870601654055,
      "epoch": 0.3045148895292987,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995086622122167,
      "loss": 5.5628,
      "mean_token_accuracy": 0.17348585575819014,
      "num_tokens": 7282773.0,
      "step": 3170
    },
    {
      "entropy": 5.527261400222779,
      "epoch": 0.30499519692603266,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995063951574871,
      "loss": 5.4046,
      "mean_token_accuracy": 0.18299773186445237,
      "num_tokens": 7294460.0,
      "step": 3175
    },
    {
      "entropy": 5.462077379226685,
      "epoch": 0.30547550432276654,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004995041228903856,
      "loss": 5.4661,
      "mean_token_accuracy": 0.18556461483240128,
      "num_tokens": 7305125.0,
      "step": 3180
    },
    {
      "entropy": 5.476806926727295,
      "epoch": 0.3059558117195005,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499501845410965,
      "loss": 5.4731,
      "mean_token_accuracy": 0.17339503020048141,
      "num_tokens": 7317738.0,
      "step": 3185
    },
    {
      "entropy": 5.590009593963623,
      "epoch": 0.30643611911623436,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994995627192781,
      "loss": 5.4933,
      "mean_token_accuracy": 0.18183531910181044,
      "num_tokens": 7329820.0,
      "step": 3190
    },
    {
      "entropy": 5.583674764633178,
      "epoch": 0.3069164265129683,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994972748153781,
      "loss": 5.5449,
      "mean_token_accuracy": 0.1705733358860016,
      "num_tokens": 7341471.0,
      "step": 3195
    },
    {
      "entropy": 5.594673109054566,
      "epoch": 0.30739673390970224,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000499494981699318,
      "loss": 5.6723,
      "mean_token_accuracy": 0.1798601508140564,
      "num_tokens": 7352459.0,
      "step": 3200
    },
    {
      "entropy": 5.549898624420166,
      "epoch": 0.3078770413064361,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499492683371151,
      "loss": 5.4468,
      "mean_token_accuracy": 0.18590417206287385,
      "num_tokens": 7363129.0,
      "step": 3205
    },
    {
      "entropy": 5.5865977764129635,
      "epoch": 0.30835734870317005,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994903798309306,
      "loss": 5.5361,
      "mean_token_accuracy": 0.17652668207883834,
      "num_tokens": 7373726.0,
      "step": 3210
    },
    {
      "entropy": 5.575078392028809,
      "epoch": 0.30883765609990393,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004994880710787102,
      "loss": 5.5808,
      "mean_token_accuracy": 0.17862567454576492,
      "num_tokens": 7383847.0,
      "step": 3215
    },
    {
      "entropy": 5.550208473205567,
      "epoch": 0.30931796349663787,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994857571145432,
      "loss": 5.5744,
      "mean_token_accuracy": 0.1706383302807808,
      "num_tokens": 7396425.0,
      "step": 3220
    },
    {
      "entropy": 5.712774705886841,
      "epoch": 0.30979827089337175,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994834379384837,
      "loss": 5.6351,
      "mean_token_accuracy": 0.1724379241466522,
      "num_tokens": 7407591.0,
      "step": 3225
    },
    {
      "entropy": 5.496092796325684,
      "epoch": 0.3102785782901057,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994811135505851,
      "loss": 5.4903,
      "mean_token_accuracy": 0.17818500250577926,
      "num_tokens": 7419668.0,
      "step": 3230
    },
    {
      "entropy": 5.6027021408081055,
      "epoch": 0.31075888568683957,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994787839509018,
      "loss": 5.6,
      "mean_token_accuracy": 0.1715978592634201,
      "num_tokens": 7431969.0,
      "step": 3235
    },
    {
      "entropy": 5.718282890319824,
      "epoch": 0.3112391930835735,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004994764491394876,
      "loss": 5.6002,
      "mean_token_accuracy": 0.16918736100196838,
      "num_tokens": 7443467.0,
      "step": 3240
    },
    {
      "entropy": 5.618907451629639,
      "epoch": 0.3117195004803074,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004994741091163969,
      "loss": 5.5802,
      "mean_token_accuracy": 0.18071286231279374,
      "num_tokens": 7455220.0,
      "step": 3245
    },
    {
      "entropy": 5.535886287689209,
      "epoch": 0.3121998078770413,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499471763881684,
      "loss": 5.4891,
      "mean_token_accuracy": 0.17951259315013884,
      "num_tokens": 7466452.0,
      "step": 3250
    },
    {
      "entropy": 5.520643424987793,
      "epoch": 0.3126801152737752,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004994694134354031,
      "loss": 5.5523,
      "mean_token_accuracy": 0.18383537232875824,
      "num_tokens": 7478798.0,
      "step": 3255
    },
    {
      "entropy": 5.656300401687622,
      "epoch": 0.31316042267050914,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000499467057777609,
      "loss": 5.6244,
      "mean_token_accuracy": 0.17627301812171936,
      "num_tokens": 7491039.0,
      "step": 3260
    },
    {
      "entropy": 5.56907434463501,
      "epoch": 0.313640730067243,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004994646969083565,
      "loss": 5.4836,
      "mean_token_accuracy": 0.18003961741924285,
      "num_tokens": 7501455.0,
      "step": 3265
    },
    {
      "entropy": 5.550380516052246,
      "epoch": 0.31412103746397696,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994623308277002,
      "loss": 5.5563,
      "mean_token_accuracy": 0.17807594686746597,
      "num_tokens": 7513484.0,
      "step": 3270
    },
    {
      "entropy": 5.624225521087647,
      "epoch": 0.31460134486071084,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499459959535695,
      "loss": 5.5295,
      "mean_token_accuracy": 0.17495962828397751,
      "num_tokens": 7524868.0,
      "step": 3275
    },
    {
      "entropy": 5.458921718597412,
      "epoch": 0.3150816522574448,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004994575830323962,
      "loss": 5.4201,
      "mean_token_accuracy": 0.18480057418346404,
      "num_tokens": 7536290.0,
      "step": 3280
    },
    {
      "entropy": 5.473631525039673,
      "epoch": 0.31556195965417866,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994552013178586,
      "loss": 5.5184,
      "mean_token_accuracy": 0.18113575130701065,
      "num_tokens": 7546994.0,
      "step": 3285
    },
    {
      "entropy": 5.576454162597656,
      "epoch": 0.3160422670509126,
      "grad_norm": 1.25,
      "learning_rate": 0.000499452814392138,
      "loss": 5.5111,
      "mean_token_accuracy": 0.1771087422966957,
      "num_tokens": 7559231.0,
      "step": 3290
    },
    {
      "entropy": 5.5458738803863525,
      "epoch": 0.3165225744476465,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994504222552894,
      "loss": 5.4728,
      "mean_token_accuracy": 0.18084817826747895,
      "num_tokens": 7570894.0,
      "step": 3295
    },
    {
      "entropy": 5.593354415893555,
      "epoch": 0.3170028818443804,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004994480249073684,
      "loss": 5.6033,
      "mean_token_accuracy": 0.17107368111610413,
      "num_tokens": 7581765.0,
      "step": 3300
    },
    {
      "entropy": 5.584808492660523,
      "epoch": 0.3174831892411143,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994456223484308,
      "loss": 5.4479,
      "mean_token_accuracy": 0.1847134992480278,
      "num_tokens": 7591828.0,
      "step": 3305
    },
    {
      "entropy": 5.496738052368164,
      "epoch": 0.31796349663784823,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994432145785323,
      "loss": 5.5205,
      "mean_token_accuracy": 0.17210932746529578,
      "num_tokens": 7604230.0,
      "step": 3310
    },
    {
      "entropy": 5.508804702758789,
      "epoch": 0.3184438040345821,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004994408015977288,
      "loss": 5.4806,
      "mean_token_accuracy": 0.17956049144268035,
      "num_tokens": 7615780.0,
      "step": 3315
    },
    {
      "entropy": 5.559992456436158,
      "epoch": 0.31892411143131605,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004994383834060764,
      "loss": 5.4228,
      "mean_token_accuracy": 0.18018749207258225,
      "num_tokens": 7627633.0,
      "step": 3320
    },
    {
      "entropy": 5.54780011177063,
      "epoch": 0.31940441882804993,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994359600036311,
      "loss": 5.5638,
      "mean_token_accuracy": 0.17372141033411026,
      "num_tokens": 7639487.0,
      "step": 3325
    },
    {
      "entropy": 5.53131685256958,
      "epoch": 0.31988472622478387,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004994335313904493,
      "loss": 5.4787,
      "mean_token_accuracy": 0.18689664602279663,
      "num_tokens": 7649976.0,
      "step": 3330
    },
    {
      "entropy": 5.652995538711548,
      "epoch": 0.32036503362151775,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004994310975665873,
      "loss": 5.6624,
      "mean_token_accuracy": 0.17187715470790862,
      "num_tokens": 7662555.0,
      "step": 3335
    },
    {
      "entropy": 5.516408157348633,
      "epoch": 0.3208453410182517,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994286585321017,
      "loss": 5.4497,
      "mean_token_accuracy": 0.18426660150289537,
      "num_tokens": 7673869.0,
      "step": 3340
    },
    {
      "entropy": 5.440064382553101,
      "epoch": 0.32132564841498557,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499426214287049,
      "loss": 5.5567,
      "mean_token_accuracy": 0.1737172707915306,
      "num_tokens": 7687149.0,
      "step": 3345
    },
    {
      "entropy": 5.638825702667236,
      "epoch": 0.3218059558117195,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004994237648314862,
      "loss": 5.5771,
      "mean_token_accuracy": 0.172631074488163,
      "num_tokens": 7698352.0,
      "step": 3350
    },
    {
      "entropy": 5.527461862564087,
      "epoch": 0.3222862632084534,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004994213101654697,
      "loss": 5.4253,
      "mean_token_accuracy": 0.18670456856489182,
      "num_tokens": 7709117.0,
      "step": 3355
    },
    {
      "entropy": 5.505496597290039,
      "epoch": 0.3227665706051873,
      "grad_norm": 1.125,
      "learning_rate": 0.000499418850289057,
      "loss": 5.4613,
      "mean_token_accuracy": 0.18491367399692535,
      "num_tokens": 7721731.0,
      "step": 3360
    },
    {
      "entropy": 5.465352964401245,
      "epoch": 0.32324687800192126,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004994163852023048,
      "loss": 5.4271,
      "mean_token_accuracy": 0.1779892385005951,
      "num_tokens": 7733010.0,
      "step": 3365
    },
    {
      "entropy": 5.502725553512573,
      "epoch": 0.32372718539865514,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004994139149052706,
      "loss": 5.4128,
      "mean_token_accuracy": 0.1885485202074051,
      "num_tokens": 7743905.0,
      "step": 3370
    },
    {
      "entropy": 5.549247980117798,
      "epoch": 0.3242074927953891,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994114393980117,
      "loss": 5.4575,
      "mean_token_accuracy": 0.1853503555059433,
      "num_tokens": 7754891.0,
      "step": 3375
    },
    {
      "entropy": 5.602116250991822,
      "epoch": 0.32468780019212296,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994089586805856,
      "loss": 5.5862,
      "mean_token_accuracy": 0.1704148158431053,
      "num_tokens": 7766831.0,
      "step": 3380
    },
    {
      "entropy": 5.567165756225586,
      "epoch": 0.3251681075888569,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994064727530496,
      "loss": 5.6394,
      "mean_token_accuracy": 0.17241787314414977,
      "num_tokens": 7778239.0,
      "step": 3385
    },
    {
      "entropy": 5.570478391647339,
      "epoch": 0.3256484149855908,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994039816154618,
      "loss": 5.4442,
      "mean_token_accuracy": 0.18829160332679748,
      "num_tokens": 7789449.0,
      "step": 3390
    },
    {
      "entropy": 5.566330051422119,
      "epoch": 0.3261287223823247,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00049940148526788,
      "loss": 5.5365,
      "mean_token_accuracy": 0.17587794810533525,
      "num_tokens": 7799507.0,
      "step": 3395
    },
    {
      "entropy": 5.594382429122925,
      "epoch": 0.3266090297790586,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004993989837103618,
      "loss": 5.5278,
      "mean_token_accuracy": 0.1818511962890625,
      "num_tokens": 7811730.0,
      "step": 3400
    },
    {
      "entropy": 5.5170355319976805,
      "epoch": 0.3270893371757925,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993964769429657,
      "loss": 5.3913,
      "mean_token_accuracy": 0.19005172401666642,
      "num_tokens": 7823108.0,
      "step": 3405
    },
    {
      "entropy": 5.479337692260742,
      "epoch": 0.3275696445725264,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004993939649657498,
      "loss": 5.4954,
      "mean_token_accuracy": 0.18005067110061646,
      "num_tokens": 7835384.0,
      "step": 3410
    },
    {
      "entropy": 5.633117008209228,
      "epoch": 0.32804995196926034,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004993914477787721,
      "loss": 5.5289,
      "mean_token_accuracy": 0.18406772017478942,
      "num_tokens": 7845860.0,
      "step": 3415
    },
    {
      "entropy": 5.571261167526245,
      "epoch": 0.3285302593659942,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004993889253820915,
      "loss": 5.5867,
      "mean_token_accuracy": 0.17345319539308549,
      "num_tokens": 7856861.0,
      "step": 3420
    },
    {
      "entropy": 5.547126531600952,
      "epoch": 0.32901056676272816,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004993863977757663,
      "loss": 5.5318,
      "mean_token_accuracy": 0.1789025694131851,
      "num_tokens": 7868902.0,
      "step": 3425
    },
    {
      "entropy": 5.592067241668701,
      "epoch": 0.32949087415946204,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004993838649598552,
      "loss": 5.5241,
      "mean_token_accuracy": 0.17713177651166917,
      "num_tokens": 7880369.0,
      "step": 3430
    },
    {
      "entropy": 5.600041341781616,
      "epoch": 0.329971181556196,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004993813269344171,
      "loss": 5.5442,
      "mean_token_accuracy": 0.1734185978770256,
      "num_tokens": 7892041.0,
      "step": 3435
    },
    {
      "entropy": 5.557718801498413,
      "epoch": 0.33045148895292986,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004993787836995108,
      "loss": 5.4887,
      "mean_token_accuracy": 0.18128235787153243,
      "num_tokens": 7903644.0,
      "step": 3440
    },
    {
      "entropy": 5.55188684463501,
      "epoch": 0.3309317963496638,
      "grad_norm": 1.25,
      "learning_rate": 0.0004993762352551954,
      "loss": 5.5356,
      "mean_token_accuracy": 0.17435626536607743,
      "num_tokens": 7915679.0,
      "step": 3445
    },
    {
      "entropy": 5.591059875488281,
      "epoch": 0.3314121037463977,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004993736816015301,
      "loss": 5.5674,
      "mean_token_accuracy": 0.1774520233273506,
      "num_tokens": 7928020.0,
      "step": 3450
    },
    {
      "entropy": 5.646395540237426,
      "epoch": 0.3318924111431316,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004993711227385742,
      "loss": 5.6181,
      "mean_token_accuracy": 0.17499688863754273,
      "num_tokens": 7940521.0,
      "step": 3455
    },
    {
      "entropy": 5.59263768196106,
      "epoch": 0.3323727185398655,
      "grad_norm": 1.25,
      "learning_rate": 0.0004993685586663871,
      "loss": 5.5587,
      "mean_token_accuracy": 0.17106352746486664,
      "num_tokens": 7952234.0,
      "step": 3460
    },
    {
      "entropy": 5.6110433578491214,
      "epoch": 0.33285302593659943,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004993659893850281,
      "loss": 5.6388,
      "mean_token_accuracy": 0.17529400438070297,
      "num_tokens": 7964900.0,
      "step": 3465
    },
    {
      "entropy": 5.642898845672607,
      "epoch": 0.3333333333333333,
      "grad_norm": 1.375,
      "learning_rate": 0.0004993634148945573,
      "loss": 5.63,
      "mean_token_accuracy": 0.17343595176935195,
      "num_tokens": 7976661.0,
      "step": 3470
    },
    {
      "entropy": 5.529425048828125,
      "epoch": 0.33381364073006725,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004993608351950341,
      "loss": 5.4542,
      "mean_token_accuracy": 0.18070945143699646,
      "num_tokens": 7987605.0,
      "step": 3475
    },
    {
      "entropy": 5.487625551223755,
      "epoch": 0.33429394812680113,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004993582502865185,
      "loss": 5.5197,
      "mean_token_accuracy": 0.17941949963569642,
      "num_tokens": 7998879.0,
      "step": 3480
    },
    {
      "entropy": 5.516484785079956,
      "epoch": 0.33477425552353507,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004993556601690706,
      "loss": 5.5174,
      "mean_token_accuracy": 0.1826210230588913,
      "num_tokens": 8010634.0,
      "step": 3485
    },
    {
      "entropy": 5.6183764934539795,
      "epoch": 0.33525456292026895,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004993530648427505,
      "loss": 5.5108,
      "mean_token_accuracy": 0.17249346971511842,
      "num_tokens": 8022134.0,
      "step": 3490
    },
    {
      "entropy": 5.5445984363555905,
      "epoch": 0.3357348703170029,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004993504643076184,
      "loss": 5.4782,
      "mean_token_accuracy": 0.17687141448259353,
      "num_tokens": 8034454.0,
      "step": 3495
    },
    {
      "entropy": 5.613416385650635,
      "epoch": 0.33621517771373677,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004993478585637347,
      "loss": 5.5539,
      "mean_token_accuracy": 0.17275482267141343,
      "num_tokens": 8046175.0,
      "step": 3500
    },
    {
      "entropy": 5.645603847503662,
      "epoch": 0.3366954851104707,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004993452476111599,
      "loss": 5.5959,
      "mean_token_accuracy": 0.1748912051320076,
      "num_tokens": 8057335.0,
      "step": 3505
    },
    {
      "entropy": 5.4952959537506105,
      "epoch": 0.3371757925072046,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004993426314499546,
      "loss": 5.4204,
      "mean_token_accuracy": 0.1863348513841629,
      "num_tokens": 8068860.0,
      "step": 3510
    },
    {
      "entropy": 5.541372632980346,
      "epoch": 0.3376560999039385,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004993400100801796,
      "loss": 5.4994,
      "mean_token_accuracy": 0.18107041716575623,
      "num_tokens": 8078957.0,
      "step": 3515
    },
    {
      "entropy": 5.620969152450561,
      "epoch": 0.3381364073006724,
      "grad_norm": 1.25,
      "learning_rate": 0.0004993373835018956,
      "loss": 5.6113,
      "mean_token_accuracy": 0.1709742322564125,
      "num_tokens": 8090143.0,
      "step": 3520
    },
    {
      "entropy": 5.475292491912842,
      "epoch": 0.33861671469740634,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004993347517151638,
      "loss": 5.4294,
      "mean_token_accuracy": 0.1815132901072502,
      "num_tokens": 8101961.0,
      "step": 3525
    },
    {
      "entropy": 5.399444246292115,
      "epoch": 0.3390970220941403,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004993321147200452,
      "loss": 5.3475,
      "mean_token_accuracy": 0.1859508275985718,
      "num_tokens": 8112635.0,
      "step": 3530
    },
    {
      "entropy": 5.527603101730347,
      "epoch": 0.33957732949087416,
      "grad_norm": 1.390625,
      "learning_rate": 0.000499329472516601,
      "loss": 5.5301,
      "mean_token_accuracy": 0.17721263766288758,
      "num_tokens": 8123730.0,
      "step": 3535
    },
    {
      "entropy": 5.588210105895996,
      "epoch": 0.3400576368876081,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004993268251048925,
      "loss": 5.5535,
      "mean_token_accuracy": 0.17419885396957396,
      "num_tokens": 8135258.0,
      "step": 3540
    },
    {
      "entropy": 5.5484428882598875,
      "epoch": 0.340537944284342,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004993241724849814,
      "loss": 5.4141,
      "mean_token_accuracy": 0.18497108668088913,
      "num_tokens": 8145946.0,
      "step": 3545
    },
    {
      "entropy": 5.568217039108276,
      "epoch": 0.3410182516810759,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499321514656929,
      "loss": 5.6227,
      "mean_token_accuracy": 0.17725986391305923,
      "num_tokens": 8158422.0,
      "step": 3550
    },
    {
      "entropy": 5.59538516998291,
      "epoch": 0.3414985590778098,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004993188516207972,
      "loss": 5.5336,
      "mean_token_accuracy": 0.16808454543352128,
      "num_tokens": 8170849.0,
      "step": 3555
    },
    {
      "entropy": 5.472879934310913,
      "epoch": 0.34197886647454373,
      "grad_norm": 1.25,
      "learning_rate": 0.0004993161833766478,
      "loss": 5.4718,
      "mean_token_accuracy": 0.17750526070594788,
      "num_tokens": 8180907.0,
      "step": 3560
    },
    {
      "entropy": 5.5876367568969725,
      "epoch": 0.3424591738712776,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004993135099245426,
      "loss": 5.5691,
      "mean_token_accuracy": 0.17272659689188002,
      "num_tokens": 8193418.0,
      "step": 3565
    },
    {
      "entropy": 5.638572835922242,
      "epoch": 0.34293948126801155,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004993108312645438,
      "loss": 5.6212,
      "mean_token_accuracy": 0.17432851642370223,
      "num_tokens": 8203873.0,
      "step": 3570
    },
    {
      "entropy": 5.689888572692871,
      "epoch": 0.34341978866474543,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004993081473967135,
      "loss": 5.6237,
      "mean_token_accuracy": 0.1738373428583145,
      "num_tokens": 8215150.0,
      "step": 3575
    },
    {
      "entropy": 5.633081579208374,
      "epoch": 0.34390009606147937,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004993054583211143,
      "loss": 5.5772,
      "mean_token_accuracy": 0.17401220202445983,
      "num_tokens": 8225732.0,
      "step": 3580
    },
    {
      "entropy": 5.580939245223999,
      "epoch": 0.34438040345821325,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004993027640378081,
      "loss": 5.5413,
      "mean_token_accuracy": 0.17461099326610566,
      "num_tokens": 8237523.0,
      "step": 3585
    },
    {
      "entropy": 5.578071880340576,
      "epoch": 0.3448607108549472,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000499300064546858,
      "loss": 5.5276,
      "mean_token_accuracy": 0.18483215868473052,
      "num_tokens": 8249660.0,
      "step": 3590
    },
    {
      "entropy": 5.528390693664551,
      "epoch": 0.34534101825168106,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004992973598483264,
      "loss": 5.361,
      "mean_token_accuracy": 0.18607661873102188,
      "num_tokens": 8261224.0,
      "step": 3595
    },
    {
      "entropy": 5.4553515911102295,
      "epoch": 0.345821325648415,
      "grad_norm": 1.328125,
      "learning_rate": 0.000499294649942276,
      "loss": 5.5911,
      "mean_token_accuracy": 0.17476649433374405,
      "num_tokens": 8272177.0,
      "step": 3600
    },
    {
      "entropy": 5.509897232055664,
      "epoch": 0.3463016330451489,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004992919348287699,
      "loss": 5.4926,
      "mean_token_accuracy": 0.17783477455377578,
      "num_tokens": 8284449.0,
      "step": 3605
    },
    {
      "entropy": 5.580263137817383,
      "epoch": 0.3467819404418828,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004992892145078711,
      "loss": 5.5644,
      "mean_token_accuracy": 0.18046885281801223,
      "num_tokens": 8297028.0,
      "step": 3610
    },
    {
      "entropy": 5.550862407684326,
      "epoch": 0.3472622478386167,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004992864889796427,
      "loss": 5.4434,
      "mean_token_accuracy": 0.1867825835943222,
      "num_tokens": 8308348.0,
      "step": 3615
    },
    {
      "entropy": 5.5030101299285885,
      "epoch": 0.34774255523535064,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004992837582441481,
      "loss": 5.4419,
      "mean_token_accuracy": 0.18680350333452225,
      "num_tokens": 8318951.0,
      "step": 3620
    },
    {
      "entropy": 5.570378112792969,
      "epoch": 0.3482228626320845,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004992810223014506,
      "loss": 5.5567,
      "mean_token_accuracy": 0.1736852467060089,
      "num_tokens": 8329979.0,
      "step": 3625
    },
    {
      "entropy": 5.6074597358703615,
      "epoch": 0.34870317002881845,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004992782811516137,
      "loss": 5.4958,
      "mean_token_accuracy": 0.18209828138351442,
      "num_tokens": 8340332.0,
      "step": 3630
    },
    {
      "entropy": 5.504545783996582,
      "epoch": 0.34918347742555234,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004992755347947011,
      "loss": 5.5036,
      "mean_token_accuracy": 0.17444830089807511,
      "num_tokens": 8350376.0,
      "step": 3635
    },
    {
      "entropy": 5.5022422790527346,
      "epoch": 0.34966378482228627,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004992727832307766,
      "loss": 5.447,
      "mean_token_accuracy": 0.1850413426756859,
      "num_tokens": 8361884.0,
      "step": 3640
    },
    {
      "entropy": 5.412858676910401,
      "epoch": 0.35014409221902015,
      "grad_norm": 1.25,
      "learning_rate": 0.0004992700264599039,
      "loss": 5.3681,
      "mean_token_accuracy": 0.1885715678334236,
      "num_tokens": 8372914.0,
      "step": 3645
    },
    {
      "entropy": 5.607450914382935,
      "epoch": 0.3506243996157541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992672644821473,
      "loss": 5.541,
      "mean_token_accuracy": 0.17206971049308778,
      "num_tokens": 8385800.0,
      "step": 3650
    },
    {
      "entropy": 5.460282897949218,
      "epoch": 0.35110470701248797,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004992644972975707,
      "loss": 5.4328,
      "mean_token_accuracy": 0.18788817524909973,
      "num_tokens": 8397783.0,
      "step": 3655
    },
    {
      "entropy": 5.495180940628051,
      "epoch": 0.3515850144092219,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004992617249062383,
      "loss": 5.4815,
      "mean_token_accuracy": 0.1792609453201294,
      "num_tokens": 8409096.0,
      "step": 3660
    },
    {
      "entropy": 5.487185144424439,
      "epoch": 0.3520653218059558,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004992589473082147,
      "loss": 5.3823,
      "mean_token_accuracy": 0.18282747715711595,
      "num_tokens": 8420833.0,
      "step": 3665
    },
    {
      "entropy": 5.518129491806031,
      "epoch": 0.3525456292026897,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004992561645035641,
      "loss": 5.484,
      "mean_token_accuracy": 0.1755655437707901,
      "num_tokens": 8431848.0,
      "step": 3670
    },
    {
      "entropy": 5.674550867080688,
      "epoch": 0.3530259365994236,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004992533764923515,
      "loss": 5.581,
      "mean_token_accuracy": 0.17016951739788055,
      "num_tokens": 8443620.0,
      "step": 3675
    },
    {
      "entropy": 5.523875331878662,
      "epoch": 0.35350624399615754,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004992505832746412,
      "loss": 5.5753,
      "mean_token_accuracy": 0.17916093170642852,
      "num_tokens": 8455339.0,
      "step": 3680
    },
    {
      "entropy": 5.41377534866333,
      "epoch": 0.3539865513928914,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004992477848504983,
      "loss": 5.4173,
      "mean_token_accuracy": 0.1903923347592354,
      "num_tokens": 8466342.0,
      "step": 3685
    },
    {
      "entropy": 5.558801507949829,
      "epoch": 0.35446685878962536,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004992449812199877,
      "loss": 5.4341,
      "mean_token_accuracy": 0.18054721355438233,
      "num_tokens": 8476955.0,
      "step": 3690
    },
    {
      "entropy": 5.4651518821716305,
      "epoch": 0.3549471661863593,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004992421723831745,
      "loss": 5.4261,
      "mean_token_accuracy": 0.18249859064817428,
      "num_tokens": 8488148.0,
      "step": 3695
    },
    {
      "entropy": 5.50803542137146,
      "epoch": 0.3554274735830932,
      "grad_norm": 1.375,
      "learning_rate": 0.0004992393583401239,
      "loss": 5.44,
      "mean_token_accuracy": 0.1850776731967926,
      "num_tokens": 8499014.0,
      "step": 3700
    },
    {
      "entropy": 5.592900276184082,
      "epoch": 0.3559077809798271,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004992365390909014,
      "loss": 5.5041,
      "mean_token_accuracy": 0.1760295107960701,
      "num_tokens": 8510191.0,
      "step": 3705
    },
    {
      "entropy": 5.4998047828674315,
      "epoch": 0.356388088376561,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004992337146355721,
      "loss": 5.382,
      "mean_token_accuracy": 0.19070065915584564,
      "num_tokens": 8522368.0,
      "step": 3710
    },
    {
      "entropy": 5.478042936325073,
      "epoch": 0.35686839577329493,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004992308849742019,
      "loss": 5.414,
      "mean_token_accuracy": 0.18542258441448212,
      "num_tokens": 8533692.0,
      "step": 3715
    },
    {
      "entropy": 5.532752656936646,
      "epoch": 0.3573487031700288,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004992280501068563,
      "loss": 5.4923,
      "mean_token_accuracy": 0.18073118031024932,
      "num_tokens": 8545283.0,
      "step": 3720
    },
    {
      "entropy": 5.548139572143555,
      "epoch": 0.35782901056676275,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004992252100336012,
      "loss": 5.4623,
      "mean_token_accuracy": 0.18284210115671157,
      "num_tokens": 8555094.0,
      "step": 3725
    },
    {
      "entropy": 5.493295383453369,
      "epoch": 0.35830931796349663,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004992223647545027,
      "loss": 5.5104,
      "mean_token_accuracy": 0.17202253490686417,
      "num_tokens": 8565179.0,
      "step": 3730
    },
    {
      "entropy": 5.458832216262818,
      "epoch": 0.35878962536023057,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004992195142696266,
      "loss": 5.4144,
      "mean_token_accuracy": 0.17971008121967316,
      "num_tokens": 8575764.0,
      "step": 3735
    },
    {
      "entropy": 5.402663707733154,
      "epoch": 0.35926993275696445,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004992166585790391,
      "loss": 5.3574,
      "mean_token_accuracy": 0.1906619504094124,
      "num_tokens": 8586052.0,
      "step": 3740
    },
    {
      "entropy": 5.586157178878784,
      "epoch": 0.3597502401536984,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004992137976828067,
      "loss": 5.5437,
      "mean_token_accuracy": 0.1729776903986931,
      "num_tokens": 8597832.0,
      "step": 3745
    },
    {
      "entropy": 5.630362415313721,
      "epoch": 0.36023054755043227,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004992109315809955,
      "loss": 5.522,
      "mean_token_accuracy": 0.1815161645412445,
      "num_tokens": 8609917.0,
      "step": 3750
    },
    {
      "entropy": 5.622105073928833,
      "epoch": 0.3607108549471662,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004992080602736725,
      "loss": 5.531,
      "mean_token_accuracy": 0.17484886199235916,
      "num_tokens": 8620137.0,
      "step": 3755
    },
    {
      "entropy": 5.485766124725342,
      "epoch": 0.3611911623439001,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004992051837609039,
      "loss": 5.5359,
      "mean_token_accuracy": 0.17977116703987123,
      "num_tokens": 8630620.0,
      "step": 3760
    },
    {
      "entropy": 5.506924819946289,
      "epoch": 0.361671469740634,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004992023020427568,
      "loss": 5.4226,
      "mean_token_accuracy": 0.186400842666626,
      "num_tokens": 8642820.0,
      "step": 3765
    },
    {
      "entropy": 5.526337003707885,
      "epoch": 0.3621517771373679,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004991994151192979,
      "loss": 5.4057,
      "mean_token_accuracy": 0.18802926689386368,
      "num_tokens": 8654081.0,
      "step": 3770
    },
    {
      "entropy": 5.496154022216797,
      "epoch": 0.36263208453410184,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004991965229905943,
      "loss": 5.4325,
      "mean_token_accuracy": 0.18222517222166063,
      "num_tokens": 8665097.0,
      "step": 3775
    },
    {
      "entropy": 5.496727418899536,
      "epoch": 0.3631123919308357,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004991936256567133,
      "loss": 5.4288,
      "mean_token_accuracy": 0.18645656406879424,
      "num_tokens": 8676576.0,
      "step": 3780
    },
    {
      "entropy": 5.522012424468994,
      "epoch": 0.36359269932756966,
      "grad_norm": 1.875,
      "learning_rate": 0.000499190723117722,
      "loss": 5.4352,
      "mean_token_accuracy": 0.1772135764360428,
      "num_tokens": 8689217.0,
      "step": 3785
    },
    {
      "entropy": 5.579939079284668,
      "epoch": 0.36407300672430354,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004991878153736877,
      "loss": 5.5608,
      "mean_token_accuracy": 0.17048364728689194,
      "num_tokens": 8701101.0,
      "step": 3790
    },
    {
      "entropy": 5.5359409809112545,
      "epoch": 0.3645533141210375,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004991849024246781,
      "loss": 5.4834,
      "mean_token_accuracy": 0.1785448133945465,
      "num_tokens": 8712769.0,
      "step": 3795
    },
    {
      "entropy": 5.5004900932312015,
      "epoch": 0.36503362151777136,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004991819842707608,
      "loss": 5.4207,
      "mean_token_accuracy": 0.18760534077882768,
      "num_tokens": 8724691.0,
      "step": 3800
    },
    {
      "entropy": 5.416055679321289,
      "epoch": 0.3655139289145053,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004991790609120035,
      "loss": 5.3822,
      "mean_token_accuracy": 0.18623173832893372,
      "num_tokens": 8738008.0,
      "step": 3805
    },
    {
      "entropy": 5.621143198013305,
      "epoch": 0.3659942363112392,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000499176132348474,
      "loss": 5.5939,
      "mean_token_accuracy": 0.17312756925821304,
      "num_tokens": 8750353.0,
      "step": 3810
    },
    {
      "entropy": 5.495223045349121,
      "epoch": 0.3664745437079731,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004991731985802405,
      "loss": 5.4073,
      "mean_token_accuracy": 0.1862175390124321,
      "num_tokens": 8761319.0,
      "step": 3815
    },
    {
      "entropy": 5.491991853713989,
      "epoch": 0.366954851104707,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004991702596073708,
      "loss": 5.4479,
      "mean_token_accuracy": 0.18171467781066894,
      "num_tokens": 8772722.0,
      "step": 3820
    },
    {
      "entropy": 5.680236911773681,
      "epoch": 0.36743515850144093,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004991673154299335,
      "loss": 5.5401,
      "mean_token_accuracy": 0.17654287219047546,
      "num_tokens": 8783886.0,
      "step": 3825
    },
    {
      "entropy": 5.507202577590943,
      "epoch": 0.3679154658981748,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004991643660479967,
      "loss": 5.4677,
      "mean_token_accuracy": 0.182183638215065,
      "num_tokens": 8795161.0,
      "step": 3830
    },
    {
      "entropy": 5.455057001113891,
      "epoch": 0.36839577329490875,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004991614114616289,
      "loss": 5.4381,
      "mean_token_accuracy": 0.18425257205963136,
      "num_tokens": 8805417.0,
      "step": 3835
    },
    {
      "entropy": 5.536152696609497,
      "epoch": 0.3688760806916426,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004991584516708988,
      "loss": 5.5076,
      "mean_token_accuracy": 0.17491607666015624,
      "num_tokens": 8817980.0,
      "step": 3840
    },
    {
      "entropy": 5.479372930526734,
      "epoch": 0.36935638808837656,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004991554866758751,
      "loss": 5.3064,
      "mean_token_accuracy": 0.19501214772462844,
      "num_tokens": 8829814.0,
      "step": 3845
    },
    {
      "entropy": 5.439425039291382,
      "epoch": 0.36983669548511044,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004991525164766265,
      "loss": 5.4003,
      "mean_token_accuracy": 0.1816820025444031,
      "num_tokens": 8840220.0,
      "step": 3850
    },
    {
      "entropy": 5.535809803009033,
      "epoch": 0.3703170028818444,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004991495410732222,
      "loss": 5.544,
      "mean_token_accuracy": 0.17503760904073715,
      "num_tokens": 8852156.0,
      "step": 3855
    },
    {
      "entropy": 5.564910411834717,
      "epoch": 0.37079731027857826,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004991465604657311,
      "loss": 5.4649,
      "mean_token_accuracy": 0.1742330566048622,
      "num_tokens": 8863362.0,
      "step": 3860
    },
    {
      "entropy": 5.473694896697998,
      "epoch": 0.3712776176753122,
      "grad_norm": 1.75,
      "learning_rate": 0.0004991435746542224,
      "loss": 5.4766,
      "mean_token_accuracy": 0.1813788130879402,
      "num_tokens": 8874957.0,
      "step": 3865
    },
    {
      "entropy": 5.491585159301758,
      "epoch": 0.37175792507204614,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004991405836387655,
      "loss": 5.4097,
      "mean_token_accuracy": 0.19009595215320588,
      "num_tokens": 8886171.0,
      "step": 3870
    },
    {
      "entropy": 5.552527809143067,
      "epoch": 0.37223823246878,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004991375874194298,
      "loss": 5.5307,
      "mean_token_accuracy": 0.18047902584075928,
      "num_tokens": 8898103.0,
      "step": 3875
    },
    {
      "entropy": 5.532804012298584,
      "epoch": 0.37271853986551395,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000499134585996285,
      "loss": 5.4914,
      "mean_token_accuracy": 0.17974178791046141,
      "num_tokens": 8910195.0,
      "step": 3880
    },
    {
      "entropy": 5.467819118499756,
      "epoch": 0.37319884726224783,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004991315793694004,
      "loss": 5.4644,
      "mean_token_accuracy": 0.18020269870758057,
      "num_tokens": 8922865.0,
      "step": 3885
    },
    {
      "entropy": 5.512961053848267,
      "epoch": 0.37367915465898177,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004991285675388463,
      "loss": 5.4669,
      "mean_token_accuracy": 0.18156256675720214,
      "num_tokens": 8933913.0,
      "step": 3890
    },
    {
      "entropy": 5.496314191818238,
      "epoch": 0.37415946205571565,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004991255505046922,
      "loss": 5.409,
      "mean_token_accuracy": 0.17870626449584961,
      "num_tokens": 8945749.0,
      "step": 3895
    },
    {
      "entropy": 5.534526300430298,
      "epoch": 0.3746397694524496,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004991225282670083,
      "loss": 5.4206,
      "mean_token_accuracy": 0.1765512451529503,
      "num_tokens": 8957454.0,
      "step": 3900
    },
    {
      "entropy": 5.525162124633789,
      "epoch": 0.37512007684918347,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000499119500825865,
      "loss": 5.474,
      "mean_token_accuracy": 0.1824617013335228,
      "num_tokens": 8968173.0,
      "step": 3905
    },
    {
      "entropy": 5.45548300743103,
      "epoch": 0.3756003842459174,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004991164681813323,
      "loss": 5.4823,
      "mean_token_accuracy": 0.18043213784694673,
      "num_tokens": 8979187.0,
      "step": 3910
    },
    {
      "entropy": 5.501256608963013,
      "epoch": 0.3760806916426513,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004991134303334807,
      "loss": 5.3909,
      "mean_token_accuracy": 0.19173142462968826,
      "num_tokens": 8990007.0,
      "step": 3915
    },
    {
      "entropy": 5.502903604507447,
      "epoch": 0.3765609990393852,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004991103872823807,
      "loss": 5.5311,
      "mean_token_accuracy": 0.17736413031816484,
      "num_tokens": 9001108.0,
      "step": 3920
    },
    {
      "entropy": 5.58303918838501,
      "epoch": 0.3770413064361191,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000499107339028103,
      "loss": 5.5154,
      "mean_token_accuracy": 0.1774510145187378,
      "num_tokens": 9012704.0,
      "step": 3925
    },
    {
      "entropy": 5.534247255325317,
      "epoch": 0.37752161383285304,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004991042855707184,
      "loss": 5.467,
      "mean_token_accuracy": 0.18156566768884658,
      "num_tokens": 9024351.0,
      "step": 3930
    },
    {
      "entropy": 5.403720951080322,
      "epoch": 0.3780019212295869,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004991012269102977,
      "loss": 5.3985,
      "mean_token_accuracy": 0.18490844815969468,
      "num_tokens": 9035099.0,
      "step": 3935
    },
    {
      "entropy": 5.571021938323975,
      "epoch": 0.37848222862632086,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004990981630469119,
      "loss": 5.4305,
      "mean_token_accuracy": 0.18070338517427445,
      "num_tokens": 9045675.0,
      "step": 3940
    },
    {
      "entropy": 5.531776332855225,
      "epoch": 0.37896253602305474,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004990950939806323,
      "loss": 5.4006,
      "mean_token_accuracy": 0.18625857383012773,
      "num_tokens": 9057133.0,
      "step": 3945
    },
    {
      "entropy": 5.4402061939239506,
      "epoch": 0.3794428434197887,
      "grad_norm": 1.25,
      "learning_rate": 0.00049909201971153,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18193181306123735,
      "num_tokens": 9067806.0,
      "step": 3950
    },
    {
      "entropy": 5.379292678833008,
      "epoch": 0.37992315081652256,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004990889402396763,
      "loss": 5.4337,
      "mean_token_accuracy": 0.17841372936964034,
      "num_tokens": 9079893.0,
      "step": 3955
    },
    {
      "entropy": 5.518050765991211,
      "epoch": 0.3804034582132565,
      "grad_norm": 1.25,
      "learning_rate": 0.0004990858555651431,
      "loss": 5.4534,
      "mean_token_accuracy": 0.1778656631708145,
      "num_tokens": 9091583.0,
      "step": 3960
    },
    {
      "entropy": 5.542593479156494,
      "epoch": 0.3808837656099904,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004990827656880015,
      "loss": 5.4588,
      "mean_token_accuracy": 0.18845251202583313,
      "num_tokens": 9102919.0,
      "step": 3965
    },
    {
      "entropy": 5.516092967987061,
      "epoch": 0.3813640730067243,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004990796706083235,
      "loss": 5.4236,
      "mean_token_accuracy": 0.18649644553661346,
      "num_tokens": 9113876.0,
      "step": 3970
    },
    {
      "entropy": 5.600319576263428,
      "epoch": 0.3818443804034582,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004990765703261809,
      "loss": 5.5733,
      "mean_token_accuracy": 0.1720522865653038,
      "num_tokens": 9125630.0,
      "step": 3975
    },
    {
      "entropy": 5.546241998672485,
      "epoch": 0.38232468780019213,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004990734648416458,
      "loss": 5.4872,
      "mean_token_accuracy": 0.18249279856681824,
      "num_tokens": 9137302.0,
      "step": 3980
    },
    {
      "entropy": 5.517662239074707,
      "epoch": 0.382804995196926,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004990703541547901,
      "loss": 5.4385,
      "mean_token_accuracy": 0.19126711189746856,
      "num_tokens": 9148048.0,
      "step": 3985
    },
    {
      "entropy": 5.575773859024048,
      "epoch": 0.38328530259365995,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004990672382656863,
      "loss": 5.5034,
      "mean_token_accuracy": 0.17943598479032516,
      "num_tokens": 9158656.0,
      "step": 3990
    },
    {
      "entropy": 5.525322914123535,
      "epoch": 0.38376560999039383,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004990641171744064,
      "loss": 5.3982,
      "mean_token_accuracy": 0.1827353596687317,
      "num_tokens": 9169324.0,
      "step": 3995
    },
    {
      "entropy": 5.381278705596924,
      "epoch": 0.38424591738712777,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004990609908810231,
      "loss": 5.3556,
      "mean_token_accuracy": 0.18749811947345735,
      "num_tokens": 9180238.0,
      "step": 4000
    },
    {
      "entropy": 5.343673324584961,
      "epoch": 0.38472622478386165,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990578593856089,
      "loss": 5.2113,
      "mean_token_accuracy": 0.1928326591849327,
      "num_tokens": 9190766.0,
      "step": 4005
    },
    {
      "entropy": 5.380656909942627,
      "epoch": 0.3852065321805956,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990547226882366,
      "loss": 5.3859,
      "mean_token_accuracy": 0.18694805949926377,
      "num_tokens": 9202445.0,
      "step": 4010
    },
    {
      "entropy": 5.476647901535034,
      "epoch": 0.38568683957732947,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004990515807889788,
      "loss": 5.4586,
      "mean_token_accuracy": 0.1870392620563507,
      "num_tokens": 9212383.0,
      "step": 4015
    },
    {
      "entropy": 5.44529356956482,
      "epoch": 0.3861671469740634,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004990484336879087,
      "loss": 5.349,
      "mean_token_accuracy": 0.1862738847732544,
      "num_tokens": 9223619.0,
      "step": 4020
    },
    {
      "entropy": 5.496671152114868,
      "epoch": 0.3866474543707973,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004990452813850992,
      "loss": 5.5074,
      "mean_token_accuracy": 0.17846391946077347,
      "num_tokens": 9234137.0,
      "step": 4025
    },
    {
      "entropy": 5.539433526992798,
      "epoch": 0.3871277617675312,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004990421238806236,
      "loss": 5.4297,
      "mean_token_accuracy": 0.18255595862865448,
      "num_tokens": 9246093.0,
      "step": 4030
    },
    {
      "entropy": 5.590913152694702,
      "epoch": 0.38760806916426516,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004990389611745551,
      "loss": 5.5424,
      "mean_token_accuracy": 0.16726417541503907,
      "num_tokens": 9258021.0,
      "step": 4035
    },
    {
      "entropy": 5.551828765869141,
      "epoch": 0.38808837656099904,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004990357932669672,
      "loss": 5.5542,
      "mean_token_accuracy": 0.18153435066342355,
      "num_tokens": 9267668.0,
      "step": 4040
    },
    {
      "entropy": 5.57155327796936,
      "epoch": 0.388568683957733,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990326201579335,
      "loss": 5.4729,
      "mean_token_accuracy": 0.17921186834573746,
      "num_tokens": 9279381.0,
      "step": 4045
    },
    {
      "entropy": 5.574979829788208,
      "epoch": 0.38904899135446686,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004990294418475274,
      "loss": 5.5614,
      "mean_token_accuracy": 0.1721788078546524,
      "num_tokens": 9292578.0,
      "step": 4050
    },
    {
      "entropy": 5.47789044380188,
      "epoch": 0.3895292987512008,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004990262583358231,
      "loss": 5.3212,
      "mean_token_accuracy": 0.18749912977218627,
      "num_tokens": 9304116.0,
      "step": 4055
    },
    {
      "entropy": 5.588890504837036,
      "epoch": 0.3900096061479347,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004990230696228943,
      "loss": 5.5638,
      "mean_token_accuracy": 0.18039464950561523,
      "num_tokens": 9315877.0,
      "step": 4060
    },
    {
      "entropy": 5.451913595199585,
      "epoch": 0.3904899135446686,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004990198757088149,
      "loss": 5.4479,
      "mean_token_accuracy": 0.18427368402481079,
      "num_tokens": 9326859.0,
      "step": 4065
    },
    {
      "entropy": 5.570649194717407,
      "epoch": 0.3909702209414025,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004990166765936593,
      "loss": 5.5218,
      "mean_token_accuracy": 0.17401764392852784,
      "num_tokens": 9338067.0,
      "step": 4070
    },
    {
      "entropy": 5.551594066619873,
      "epoch": 0.3914505283381364,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004990134722775016,
      "loss": 5.5548,
      "mean_token_accuracy": 0.17771293371915817,
      "num_tokens": 9349564.0,
      "step": 4075
    },
    {
      "entropy": 5.384429550170898,
      "epoch": 0.3919308357348703,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004990102627604162,
      "loss": 5.3576,
      "mean_token_accuracy": 0.18717946261167526,
      "num_tokens": 9359962.0,
      "step": 4080
    },
    {
      "entropy": 5.442182159423828,
      "epoch": 0.39241114313160425,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004990070480424778,
      "loss": 5.3492,
      "mean_token_accuracy": 0.18788860142230987,
      "num_tokens": 9370251.0,
      "step": 4085
    },
    {
      "entropy": 5.529190635681152,
      "epoch": 0.3928914505283381,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004990038281237608,
      "loss": 5.4962,
      "mean_token_accuracy": 0.1772561714053154,
      "num_tokens": 9381848.0,
      "step": 4090
    },
    {
      "entropy": 5.45083270072937,
      "epoch": 0.39337175792507206,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004990006030043401,
      "loss": 5.4076,
      "mean_token_accuracy": 0.18683607429265975,
      "num_tokens": 9392475.0,
      "step": 4095
    },
    {
      "entropy": 5.5099693775177006,
      "epoch": 0.39385206532180594,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004989973726842906,
      "loss": 5.5358,
      "mean_token_accuracy": 0.1804587125778198,
      "num_tokens": 9404014.0,
      "step": 4100
    },
    {
      "entropy": 5.522710990905762,
      "epoch": 0.3943323727185399,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989941371636872,
      "loss": 5.4054,
      "mean_token_accuracy": 0.18026885092258454,
      "num_tokens": 9415590.0,
      "step": 4105
    },
    {
      "entropy": 5.573627662658692,
      "epoch": 0.39481268011527376,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004989908964426051,
      "loss": 5.5684,
      "mean_token_accuracy": 0.17600324302911757,
      "num_tokens": 9427486.0,
      "step": 4110
    },
    {
      "entropy": 5.538381767272949,
      "epoch": 0.3952929875120077,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004989876505211194,
      "loss": 5.4647,
      "mean_token_accuracy": 0.18080680817365646,
      "num_tokens": 9438212.0,
      "step": 4115
    },
    {
      "entropy": 5.513815832138062,
      "epoch": 0.3957732949087416,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004989843993993056,
      "loss": 5.412,
      "mean_token_accuracy": 0.1923951029777527,
      "num_tokens": 9450922.0,
      "step": 4120
    },
    {
      "entropy": 5.414572429656983,
      "epoch": 0.3962536023054755,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004989811430772392,
      "loss": 5.3807,
      "mean_token_accuracy": 0.18224074989557265,
      "num_tokens": 9461676.0,
      "step": 4125
    },
    {
      "entropy": 5.351324844360351,
      "epoch": 0.3967339097022094,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004989778815549957,
      "loss": 5.3505,
      "mean_token_accuracy": 0.19106744974851608,
      "num_tokens": 9473517.0,
      "step": 4130
    },
    {
      "entropy": 5.60519814491272,
      "epoch": 0.39721421709894333,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004989746148326508,
      "loss": 5.485,
      "mean_token_accuracy": 0.17537111341953276,
      "num_tokens": 9485660.0,
      "step": 4135
    },
    {
      "entropy": 5.52572922706604,
      "epoch": 0.3976945244956772,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004989713429102805,
      "loss": 5.4887,
      "mean_token_accuracy": 0.17903834730386733,
      "num_tokens": 9495983.0,
      "step": 4140
    },
    {
      "entropy": 5.47251706123352,
      "epoch": 0.39817483189241115,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004989680657879607,
      "loss": 5.436,
      "mean_token_accuracy": 0.17506579160690308,
      "num_tokens": 9507438.0,
      "step": 4145
    },
    {
      "entropy": 5.406937599182129,
      "epoch": 0.39865513928914503,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004989647834657675,
      "loss": 5.3427,
      "mean_token_accuracy": 0.1858995169401169,
      "num_tokens": 9517508.0,
      "step": 4150
    },
    {
      "entropy": 5.3676002502441404,
      "epoch": 0.39913544668587897,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000498961495943777,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18892575353384017,
      "num_tokens": 9527641.0,
      "step": 4155
    },
    {
      "entropy": 5.528365707397461,
      "epoch": 0.39961575408261285,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004989582032220656,
      "loss": 5.4446,
      "mean_token_accuracy": 0.18333982080221176,
      "num_tokens": 9538762.0,
      "step": 4160
    },
    {
      "entropy": 5.53466567993164,
      "epoch": 0.4000960614793468,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004989549053007096,
      "loss": 5.5004,
      "mean_token_accuracy": 0.17668719887733458,
      "num_tokens": 9550246.0,
      "step": 4165
    },
    {
      "entropy": 5.489660739898682,
      "epoch": 0.40057636887608067,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004989516021797858,
      "loss": 5.4345,
      "mean_token_accuracy": 0.18076204359531403,
      "num_tokens": 9562341.0,
      "step": 4170
    },
    {
      "entropy": 5.557612943649292,
      "epoch": 0.4010566762728146,
      "grad_norm": 1.21875,
      "learning_rate": 0.000498948293859371,
      "loss": 5.4992,
      "mean_token_accuracy": 0.1803925946354866,
      "num_tokens": 9574484.0,
      "step": 4175
    },
    {
      "entropy": 5.492554616928101,
      "epoch": 0.4015369836695485,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004989449803395415,
      "loss": 5.4545,
      "mean_token_accuracy": 0.18405203372240067,
      "num_tokens": 9585241.0,
      "step": 4180
    },
    {
      "entropy": 5.530519580841064,
      "epoch": 0.4020172910662824,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004989416616203747,
      "loss": 5.4729,
      "mean_token_accuracy": 0.1819280803203583,
      "num_tokens": 9598683.0,
      "step": 4185
    },
    {
      "entropy": 5.520482158660888,
      "epoch": 0.4024975984630163,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004989383377019476,
      "loss": 5.4792,
      "mean_token_accuracy": 0.18044654726982118,
      "num_tokens": 9610127.0,
      "step": 4190
    },
    {
      "entropy": 5.494972562789917,
      "epoch": 0.40297790585975024,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004989350085843371,
      "loss": 5.3784,
      "mean_token_accuracy": 0.18540242165327073,
      "num_tokens": 9620009.0,
      "step": 4195
    },
    {
      "entropy": 5.463833379745483,
      "epoch": 0.4034582132564842,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004989316742676207,
      "loss": 5.4727,
      "mean_token_accuracy": 0.18554787784814836,
      "num_tokens": 9632479.0,
      "step": 4200
    },
    {
      "entropy": 5.529710388183593,
      "epoch": 0.40393852065321806,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989283347518757,
      "loss": 5.5104,
      "mean_token_accuracy": 0.18498131781816482,
      "num_tokens": 9644447.0,
      "step": 4205
    },
    {
      "entropy": 5.508409833908081,
      "epoch": 0.404418828049952,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004989249900371797,
      "loss": 5.4191,
      "mean_token_accuracy": 0.175240458548069,
      "num_tokens": 9656886.0,
      "step": 4210
    },
    {
      "entropy": 5.514116859436035,
      "epoch": 0.4048991354466859,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004989216401236103,
      "loss": 5.46,
      "mean_token_accuracy": 0.18342093825340272,
      "num_tokens": 9667079.0,
      "step": 4215
    },
    {
      "entropy": 5.546218204498291,
      "epoch": 0.4053794428434198,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004989182850112455,
      "loss": 5.5314,
      "mean_token_accuracy": 0.18024542927742004,
      "num_tokens": 9678237.0,
      "step": 4220
    },
    {
      "entropy": 5.479090642929077,
      "epoch": 0.4058597502401537,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989149247001629,
      "loss": 5.3654,
      "mean_token_accuracy": 0.18879555612802507,
      "num_tokens": 9688809.0,
      "step": 4225
    },
    {
      "entropy": 5.506127262115479,
      "epoch": 0.40634005763688763,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004989115591904407,
      "loss": 5.4761,
      "mean_token_accuracy": 0.18512516021728515,
      "num_tokens": 9700460.0,
      "step": 4230
    },
    {
      "entropy": 5.435467624664307,
      "epoch": 0.4068203650336215,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989081884821569,
      "loss": 5.4232,
      "mean_token_accuracy": 0.17947154939174653,
      "num_tokens": 9711379.0,
      "step": 4235
    },
    {
      "entropy": 5.446878910064697,
      "epoch": 0.40730067243035545,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989048125753899,
      "loss": 5.3581,
      "mean_token_accuracy": 0.18849164247512817,
      "num_tokens": 9722792.0,
      "step": 4240
    },
    {
      "entropy": 5.467892980575561,
      "epoch": 0.40778097982708933,
      "grad_norm": 1.171875,
      "learning_rate": 0.000498901431470218,
      "loss": 5.4113,
      "mean_token_accuracy": 0.18827234208583832,
      "num_tokens": 9733524.0,
      "step": 4245
    },
    {
      "entropy": 5.494660711288452,
      "epoch": 0.40826128722382327,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004988980451667198,
      "loss": 5.4705,
      "mean_token_accuracy": 0.18629121035337448,
      "num_tokens": 9745430.0,
      "step": 4250
    },
    {
      "entropy": 5.445364713668823,
      "epoch": 0.40874159462055715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004988946536649737,
      "loss": 5.3789,
      "mean_token_accuracy": 0.1829912081360817,
      "num_tokens": 9756927.0,
      "step": 4255
    },
    {
      "entropy": 5.457226800918579,
      "epoch": 0.4092219020172911,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004988912569650585,
      "loss": 5.4165,
      "mean_token_accuracy": 0.1858849585056305,
      "num_tokens": 9768350.0,
      "step": 4260
    },
    {
      "entropy": 5.430975008010864,
      "epoch": 0.40970220941402496,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004988878550670533,
      "loss": 5.3572,
      "mean_token_accuracy": 0.19087884128093718,
      "num_tokens": 9779238.0,
      "step": 4265
    },
    {
      "entropy": 5.543310165405273,
      "epoch": 0.4101825168107589,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004988844479710369,
      "loss": 5.5398,
      "mean_token_accuracy": 0.17559687942266464,
      "num_tokens": 9789262.0,
      "step": 4270
    },
    {
      "entropy": 5.603669548034668,
      "epoch": 0.4106628242074928,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004988810356770884,
      "loss": 5.5125,
      "mean_token_accuracy": 0.1808350995182991,
      "num_tokens": 9800909.0,
      "step": 4275
    },
    {
      "entropy": 5.541848754882812,
      "epoch": 0.4111431316042267,
      "grad_norm": 1.359375,
      "learning_rate": 0.000498877618185287,
      "loss": 5.4592,
      "mean_token_accuracy": 0.18005858659744262,
      "num_tokens": 9812538.0,
      "step": 4280
    },
    {
      "entropy": 5.481583595275879,
      "epoch": 0.4116234390009606,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004988741954957121,
      "loss": 5.4472,
      "mean_token_accuracy": 0.18316709250211716,
      "num_tokens": 9824071.0,
      "step": 4285
    },
    {
      "entropy": 5.616250133514404,
      "epoch": 0.41210374639769454,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004988707676084432,
      "loss": 5.5597,
      "mean_token_accuracy": 0.17763041853904724,
      "num_tokens": 9836713.0,
      "step": 4290
    },
    {
      "entropy": 5.45892858505249,
      "epoch": 0.4125840537944284,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004988673345235597,
      "loss": 5.3865,
      "mean_token_accuracy": 0.18704597353935243,
      "num_tokens": 9847336.0,
      "step": 4295
    },
    {
      "entropy": 5.399388790130615,
      "epoch": 0.41306436119116235,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004988638962411416,
      "loss": 5.3885,
      "mean_token_accuracy": 0.1803923949599266,
      "num_tokens": 9858567.0,
      "step": 4300
    },
    {
      "entropy": 5.568233394622803,
      "epoch": 0.41354466858789624,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004988604527612685,
      "loss": 5.4578,
      "mean_token_accuracy": 0.17673510760068895,
      "num_tokens": 9869254.0,
      "step": 4305
    },
    {
      "entropy": 5.464548826217651,
      "epoch": 0.4140249759846302,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004988570040840205,
      "loss": 5.4004,
      "mean_token_accuracy": 0.18541614562273026,
      "num_tokens": 9879635.0,
      "step": 4310
    },
    {
      "entropy": 5.507536220550537,
      "epoch": 0.41450528338136405,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004988535502094774,
      "loss": 5.4714,
      "mean_token_accuracy": 0.18638484328985214,
      "num_tokens": 9891728.0,
      "step": 4315
    },
    {
      "entropy": 5.462026214599609,
      "epoch": 0.414985590778098,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004988500911377198,
      "loss": 5.2837,
      "mean_token_accuracy": 0.19292479157447814,
      "num_tokens": 9903140.0,
      "step": 4320
    },
    {
      "entropy": 5.489208745956421,
      "epoch": 0.41546589817483187,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004988466268688276,
      "loss": 5.4184,
      "mean_token_accuracy": 0.18310181945562362,
      "num_tokens": 9915562.0,
      "step": 4325
    },
    {
      "entropy": 5.399180269241333,
      "epoch": 0.4159462055715658,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004988431574028814,
      "loss": 5.3714,
      "mean_token_accuracy": 0.17812881022691726,
      "num_tokens": 9925720.0,
      "step": 4330
    },
    {
      "entropy": 5.475612354278565,
      "epoch": 0.4164265129682997,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004988396827399618,
      "loss": 5.4403,
      "mean_token_accuracy": 0.1799673169851303,
      "num_tokens": 9936492.0,
      "step": 4335
    },
    {
      "entropy": 5.476762056350708,
      "epoch": 0.4169068203650336,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004988362028801495,
      "loss": 5.3871,
      "mean_token_accuracy": 0.18283660858869552,
      "num_tokens": 9948658.0,
      "step": 4340
    },
    {
      "entropy": 5.3699225902557375,
      "epoch": 0.4173871277617675,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004988327178235253,
      "loss": 5.3261,
      "mean_token_accuracy": 0.1888917088508606,
      "num_tokens": 9959710.0,
      "step": 4345
    },
    {
      "entropy": 5.422818231582641,
      "epoch": 0.41786743515850144,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004988292275701699,
      "loss": 5.3119,
      "mean_token_accuracy": 0.19038798660039902,
      "num_tokens": 9970302.0,
      "step": 4350
    },
    {
      "entropy": 5.4733960151672365,
      "epoch": 0.4183477425552353,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004988257321201646,
      "loss": 5.4483,
      "mean_token_accuracy": 0.18172130286693572,
      "num_tokens": 9981776.0,
      "step": 4355
    },
    {
      "entropy": 5.54775071144104,
      "epoch": 0.41882804995196926,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004988222314735902,
      "loss": 5.4256,
      "mean_token_accuracy": 0.17825412899255752,
      "num_tokens": 9993471.0,
      "step": 4360
    },
    {
      "entropy": 5.511180019378662,
      "epoch": 0.41930835734870314,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004988187256305284,
      "loss": 5.4373,
      "mean_token_accuracy": 0.18433743417263032,
      "num_tokens": 10003818.0,
      "step": 4365
    },
    {
      "entropy": 5.520654535293579,
      "epoch": 0.4197886647454371,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004988152145910603,
      "loss": 5.5325,
      "mean_token_accuracy": 0.1697872295975685,
      "num_tokens": 10015507.0,
      "step": 4370
    },
    {
      "entropy": 5.494216966629028,
      "epoch": 0.420268972142171,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004988116983552675,
      "loss": 5.3529,
      "mean_token_accuracy": 0.18661011159420013,
      "num_tokens": 10026994.0,
      "step": 4375
    },
    {
      "entropy": 5.556194162368774,
      "epoch": 0.4207492795389049,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004988081769232317,
      "loss": 5.4865,
      "mean_token_accuracy": 0.17480666041374207,
      "num_tokens": 10039547.0,
      "step": 4380
    },
    {
      "entropy": 5.5414666652679445,
      "epoch": 0.42122958693563883,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004988046502950346,
      "loss": 5.5278,
      "mean_token_accuracy": 0.17513879239559174,
      "num_tokens": 10050568.0,
      "step": 4385
    },
    {
      "entropy": 5.501482439041138,
      "epoch": 0.4217098943323727,
      "grad_norm": 1.671875,
      "learning_rate": 0.000498801118470758,
      "loss": 5.4271,
      "mean_token_accuracy": 0.1797420412302017,
      "num_tokens": 10061780.0,
      "step": 4390
    },
    {
      "entropy": 5.416441011428833,
      "epoch": 0.42219020172910665,
      "grad_norm": 1.46875,
      "learning_rate": 0.000498797581450484,
      "loss": 5.3911,
      "mean_token_accuracy": 0.18381205797195435,
      "num_tokens": 10073101.0,
      "step": 4395
    },
    {
      "entropy": 5.427166414260864,
      "epoch": 0.42267050912584053,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004987940392342948,
      "loss": 5.3394,
      "mean_token_accuracy": 0.19149589389562607,
      "num_tokens": 10084397.0,
      "step": 4400
    },
    {
      "entropy": 5.497838401794434,
      "epoch": 0.42315081652257447,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004987904918222726,
      "loss": 5.4085,
      "mean_token_accuracy": 0.1842352643609047,
      "num_tokens": 10095635.0,
      "step": 4405
    },
    {
      "entropy": 5.48901333808899,
      "epoch": 0.42363112391930835,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004987869392144996,
      "loss": 5.4793,
      "mean_token_accuracy": 0.1816812828183174,
      "num_tokens": 10107744.0,
      "step": 4410
    },
    {
      "entropy": 5.496208906173706,
      "epoch": 0.4241114313160423,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004987833814110584,
      "loss": 5.3851,
      "mean_token_accuracy": 0.1842407152056694,
      "num_tokens": 10118552.0,
      "step": 4415
    },
    {
      "entropy": 5.532422399520874,
      "epoch": 0.42459173871277617,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004987798184120316,
      "loss": 5.5252,
      "mean_token_accuracy": 0.17265765517950057,
      "num_tokens": 10130813.0,
      "step": 4420
    },
    {
      "entropy": 5.524330091476441,
      "epoch": 0.4250720461095101,
      "grad_norm": 1.5,
      "learning_rate": 0.0004987762502175018,
      "loss": 5.547,
      "mean_token_accuracy": 0.1797323852777481,
      "num_tokens": 10142196.0,
      "step": 4425
    },
    {
      "entropy": 5.455458688735962,
      "epoch": 0.425552353506244,
      "grad_norm": 1.328125,
      "learning_rate": 0.000498772676827552,
      "loss": 5.436,
      "mean_token_accuracy": 0.18626082092523574,
      "num_tokens": 10155012.0,
      "step": 4430
    },
    {
      "entropy": 5.562879610061645,
      "epoch": 0.4260326609029779,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004987690982422652,
      "loss": 5.5231,
      "mean_token_accuracy": 0.17669414579868317,
      "num_tokens": 10166160.0,
      "step": 4435
    },
    {
      "entropy": 5.5610884666442875,
      "epoch": 0.4265129682997118,
      "grad_norm": 1.375,
      "learning_rate": 0.0004987655144617243,
      "loss": 5.3167,
      "mean_token_accuracy": 0.1907524585723877,
      "num_tokens": 10177663.0,
      "step": 4440
    },
    {
      "entropy": 5.389883518218994,
      "epoch": 0.42699327569644574,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004987619254860126,
      "loss": 5.4097,
      "mean_token_accuracy": 0.18991009294986724,
      "num_tokens": 10189127.0,
      "step": 4445
    },
    {
      "entropy": 5.544426345825196,
      "epoch": 0.4274735830931796,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004987583313152134,
      "loss": 5.5692,
      "mean_token_accuracy": 0.17470676451921463,
      "num_tokens": 10200765.0,
      "step": 4450
    },
    {
      "entropy": 5.571499347686768,
      "epoch": 0.42795389048991356,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004987547319494104,
      "loss": 5.4134,
      "mean_token_accuracy": 0.18762336522340775,
      "num_tokens": 10213047.0,
      "step": 4455
    },
    {
      "entropy": 5.486662864685059,
      "epoch": 0.42843419788664744,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004987511273886867,
      "loss": 5.3762,
      "mean_token_accuracy": 0.18954455107450485,
      "num_tokens": 10223790.0,
      "step": 4460
    },
    {
      "entropy": 5.417471694946289,
      "epoch": 0.4289145052833814,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004987475176331263,
      "loss": 5.3671,
      "mean_token_accuracy": 0.18484554588794708,
      "num_tokens": 10235077.0,
      "step": 4465
    },
    {
      "entropy": 5.510037088394165,
      "epoch": 0.42939481268011526,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004987439026828129,
      "loss": 5.4662,
      "mean_token_accuracy": 0.16899558305740356,
      "num_tokens": 10248844.0,
      "step": 4470
    },
    {
      "entropy": 5.493098974227905,
      "epoch": 0.4298751200768492,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004987402825378305,
      "loss": 5.373,
      "mean_token_accuracy": 0.19217505306005478,
      "num_tokens": 10259013.0,
      "step": 4475
    },
    {
      "entropy": 5.4027419090271,
      "epoch": 0.4303554274735831,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004987366571982631,
      "loss": 5.3912,
      "mean_token_accuracy": 0.1829615756869316,
      "num_tokens": 10271415.0,
      "step": 4480
    },
    {
      "entropy": 5.373300838470459,
      "epoch": 0.430835734870317,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004987330266641948,
      "loss": 5.4173,
      "mean_token_accuracy": 0.18090762794017792,
      "num_tokens": 10283204.0,
      "step": 4485
    },
    {
      "entropy": 5.555873060226441,
      "epoch": 0.4313160422670509,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004987293909357101,
      "loss": 5.5695,
      "mean_token_accuracy": 0.17647680789232253,
      "num_tokens": 10294111.0,
      "step": 4490
    },
    {
      "entropy": 5.61440486907959,
      "epoch": 0.43179634966378483,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004987257500128933,
      "loss": 5.4548,
      "mean_token_accuracy": 0.17417747229337693,
      "num_tokens": 10305417.0,
      "step": 4495
    },
    {
      "entropy": 5.414704751968384,
      "epoch": 0.4322766570605187,
      "grad_norm": 1.375,
      "learning_rate": 0.0004987221038958288,
      "loss": 5.3509,
      "mean_token_accuracy": 0.18823618143796922,
      "num_tokens": 10316416.0,
      "step": 4500
    },
    {
      "entropy": 5.453503274917603,
      "epoch": 0.43275696445725265,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004987184525846015,
      "loss": 5.5101,
      "mean_token_accuracy": 0.1855538010597229,
      "num_tokens": 10328251.0,
      "step": 4505
    },
    {
      "entropy": 5.528881454467774,
      "epoch": 0.4332372718539865,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004987147960792958,
      "loss": 5.4751,
      "mean_token_accuracy": 0.190151546895504,
      "num_tokens": 10340189.0,
      "step": 4510
    },
    {
      "entropy": 5.532504892349243,
      "epoch": 0.43371757925072046,
      "grad_norm": 1.375,
      "learning_rate": 0.0004987111343799971,
      "loss": 5.4182,
      "mean_token_accuracy": 0.18358281552791594,
      "num_tokens": 10351660.0,
      "step": 4515
    },
    {
      "entropy": 5.500876617431641,
      "epoch": 0.43419788664745435,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00049870746748679,
      "loss": 5.4477,
      "mean_token_accuracy": 0.18505928963422774,
      "num_tokens": 10362240.0,
      "step": 4520
    },
    {
      "entropy": 5.549033355712891,
      "epoch": 0.4346781940441883,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004987037953997598,
      "loss": 5.4865,
      "mean_token_accuracy": 0.18396148830652237,
      "num_tokens": 10373646.0,
      "step": 4525
    },
    {
      "entropy": 5.519213771820068,
      "epoch": 0.43515850144092216,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004987001181189918,
      "loss": 5.4941,
      "mean_token_accuracy": 0.1776837483048439,
      "num_tokens": 10385540.0,
      "step": 4530
    },
    {
      "entropy": 5.514508628845215,
      "epoch": 0.4356388088376561,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004986964356445713,
      "loss": 5.4957,
      "mean_token_accuracy": 0.17722374349832534,
      "num_tokens": 10396641.0,
      "step": 4535
    },
    {
      "entropy": 5.542151689529419,
      "epoch": 0.43611911623439004,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004986927479765837,
      "loss": 5.4703,
      "mean_token_accuracy": 0.17439449429512024,
      "num_tokens": 10408165.0,
      "step": 4540
    },
    {
      "entropy": 5.431505870819092,
      "epoch": 0.4365994236311239,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004986890551151148,
      "loss": 5.412,
      "mean_token_accuracy": 0.18728436678647994,
      "num_tokens": 10419646.0,
      "step": 4545
    },
    {
      "entropy": 5.512647581100464,
      "epoch": 0.43707973102785785,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004986853570602503,
      "loss": 5.4426,
      "mean_token_accuracy": 0.17937116622924804,
      "num_tokens": 10431051.0,
      "step": 4550
    },
    {
      "entropy": 5.559714508056641,
      "epoch": 0.43756003842459174,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004986816538120758,
      "loss": 5.5143,
      "mean_token_accuracy": 0.1776018351316452,
      "num_tokens": 10441833.0,
      "step": 4555
    },
    {
      "entropy": 5.498169469833374,
      "epoch": 0.43804034582132567,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004986779453706778,
      "loss": 5.4682,
      "mean_token_accuracy": 0.18185556530952454,
      "num_tokens": 10453544.0,
      "step": 4560
    },
    {
      "entropy": 5.486437511444092,
      "epoch": 0.43852065321805955,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004986742317361419,
      "loss": 5.3128,
      "mean_token_accuracy": 0.1900118961930275,
      "num_tokens": 10464140.0,
      "step": 4565
    },
    {
      "entropy": 5.3946116924285885,
      "epoch": 0.4390009606147935,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004986705129085546,
      "loss": 5.3682,
      "mean_token_accuracy": 0.18866920918226243,
      "num_tokens": 10475739.0,
      "step": 4570
    },
    {
      "entropy": 5.333141279220581,
      "epoch": 0.43948126801152737,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004986667888880021,
      "loss": 5.418,
      "mean_token_accuracy": 0.18941385596990584,
      "num_tokens": 10487451.0,
      "step": 4575
    },
    {
      "entropy": 5.512581920623779,
      "epoch": 0.4399615754082613,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004986630596745709,
      "loss": 5.4757,
      "mean_token_accuracy": 0.17348228245973588,
      "num_tokens": 10499367.0,
      "step": 4580
    },
    {
      "entropy": 5.495959854125976,
      "epoch": 0.4404418828049952,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004986593252683477,
      "loss": 5.3025,
      "mean_token_accuracy": 0.19116291105747224,
      "num_tokens": 10511192.0,
      "step": 4585
    },
    {
      "entropy": 5.356853771209717,
      "epoch": 0.4409221902017291,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004986555856694191,
      "loss": 5.3287,
      "mean_token_accuracy": 0.18143054395914077,
      "num_tokens": 10522754.0,
      "step": 4590
    },
    {
      "entropy": 5.471630573272705,
      "epoch": 0.441402497598463,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004986518408778718,
      "loss": 5.4537,
      "mean_token_accuracy": 0.18229032605886458,
      "num_tokens": 10533439.0,
      "step": 4595
    },
    {
      "entropy": 5.552867031097412,
      "epoch": 0.44188280499519694,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004986480908937929,
      "loss": 5.41,
      "mean_token_accuracy": 0.1835053011775017,
      "num_tokens": 10543654.0,
      "step": 4600
    },
    {
      "entropy": 5.414029836654663,
      "epoch": 0.4423631123919308,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004986443357172695,
      "loss": 5.3287,
      "mean_token_accuracy": 0.18729309886693954,
      "num_tokens": 10554936.0,
      "step": 4605
    },
    {
      "entropy": 5.4253771781921385,
      "epoch": 0.44284341978866476,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004986405753483887,
      "loss": 5.4085,
      "mean_token_accuracy": 0.1833935260772705,
      "num_tokens": 10566410.0,
      "step": 4610
    },
    {
      "entropy": 5.47337064743042,
      "epoch": 0.44332372718539864,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004986368097872377,
      "loss": 5.3935,
      "mean_token_accuracy": 0.19434593766927719,
      "num_tokens": 10577258.0,
      "step": 4615
    },
    {
      "entropy": 5.524574089050293,
      "epoch": 0.4438040345821326,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004986330390339042,
      "loss": 5.5073,
      "mean_token_accuracy": 0.178845639526844,
      "num_tokens": 10588783.0,
      "step": 4620
    },
    {
      "entropy": 5.466887617111206,
      "epoch": 0.44428434197886646,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004986292630884755,
      "loss": 5.3702,
      "mean_token_accuracy": 0.18453073799610137,
      "num_tokens": 10599778.0,
      "step": 4625
    },
    {
      "entropy": 5.5022495746612545,
      "epoch": 0.4447646493756004,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004986254819510393,
      "loss": 5.3953,
      "mean_token_accuracy": 0.1926836684346199,
      "num_tokens": 10611211.0,
      "step": 4630
    },
    {
      "entropy": 5.36075005531311,
      "epoch": 0.4452449567723343,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004986216956216835,
      "loss": 5.3436,
      "mean_token_accuracy": 0.1826932579278946,
      "num_tokens": 10624535.0,
      "step": 4635
    },
    {
      "entropy": 5.4547281742095945,
      "epoch": 0.4457252641690682,
      "grad_norm": 1.40625,
      "learning_rate": 0.000498617904100496,
      "loss": 5.364,
      "mean_token_accuracy": 0.1895545944571495,
      "num_tokens": 10635206.0,
      "step": 4640
    },
    {
      "entropy": 5.505380582809448,
      "epoch": 0.4462055715658021,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004986141073875646,
      "loss": 5.425,
      "mean_token_accuracy": 0.19218022376298904,
      "num_tokens": 10646993.0,
      "step": 4645
    },
    {
      "entropy": 5.473623847961425,
      "epoch": 0.44668587896253603,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004986103054829779,
      "loss": 5.3737,
      "mean_token_accuracy": 0.18942495286464692,
      "num_tokens": 10657262.0,
      "step": 4650
    },
    {
      "entropy": 5.44066014289856,
      "epoch": 0.4471661863592699,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004986064983868237,
      "loss": 5.4192,
      "mean_token_accuracy": 0.18157739341259002,
      "num_tokens": 10668962.0,
      "step": 4655
    },
    {
      "entropy": 5.504913854598999,
      "epoch": 0.44764649375600385,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004986026860991906,
      "loss": 5.408,
      "mean_token_accuracy": 0.19493381828069686,
      "num_tokens": 10680183.0,
      "step": 4660
    },
    {
      "entropy": 5.470511531829834,
      "epoch": 0.44812680115273773,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004985988686201672,
      "loss": 5.3485,
      "mean_token_accuracy": 0.19322926551103592,
      "num_tokens": 10691221.0,
      "step": 4665
    },
    {
      "entropy": 5.445162010192871,
      "epoch": 0.44860710854947167,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004985950459498419,
      "loss": 5.3777,
      "mean_token_accuracy": 0.185987688601017,
      "num_tokens": 10703462.0,
      "step": 4670
    },
    {
      "entropy": 5.462856149673462,
      "epoch": 0.44908741594620555,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004985912180883037,
      "loss": 5.3566,
      "mean_token_accuracy": 0.1904120072722435,
      "num_tokens": 10713774.0,
      "step": 4675
    },
    {
      "entropy": 5.575588703155518,
      "epoch": 0.4495677233429395,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004985873850356411,
      "loss": 5.5726,
      "mean_token_accuracy": 0.17069013193249702,
      "num_tokens": 10725109.0,
      "step": 4680
    },
    {
      "entropy": 5.450364017486573,
      "epoch": 0.45004803073967337,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004985835467919436,
      "loss": 5.3791,
      "mean_token_accuracy": 0.1900913417339325,
      "num_tokens": 10735695.0,
      "step": 4685
    },
    {
      "entropy": 5.5042625904083256,
      "epoch": 0.4505283381364073,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004985797033572999,
      "loss": 5.4632,
      "mean_token_accuracy": 0.18182757198810579,
      "num_tokens": 10747554.0,
      "step": 4690
    },
    {
      "entropy": 5.534461879730225,
      "epoch": 0.4510086455331412,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004985758547317994,
      "loss": 5.4682,
      "mean_token_accuracy": 0.18399068862199783,
      "num_tokens": 10759175.0,
      "step": 4695
    },
    {
      "entropy": 5.468274736404419,
      "epoch": 0.4514889529298751,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004985720009155315,
      "loss": 5.3448,
      "mean_token_accuracy": 0.18884512335062026,
      "num_tokens": 10770365.0,
      "step": 4700
    },
    {
      "entropy": 5.475963687896728,
      "epoch": 0.45196926032660906,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004985681419085856,
      "loss": 5.4411,
      "mean_token_accuracy": 0.178588305413723,
      "num_tokens": 10783096.0,
      "step": 4705
    },
    {
      "entropy": 5.567919683456421,
      "epoch": 0.45244956772334294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004985642777110513,
      "loss": 5.5168,
      "mean_token_accuracy": 0.17430274039506913,
      "num_tokens": 10795312.0,
      "step": 4710
    },
    {
      "entropy": 5.490743017196655,
      "epoch": 0.4529298751200769,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004985604083230183,
      "loss": 5.3523,
      "mean_token_accuracy": 0.1856426477432251,
      "num_tokens": 10806451.0,
      "step": 4715
    },
    {
      "entropy": 5.524587106704712,
      "epoch": 0.45341018251681076,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004985565337445765,
      "loss": 5.5086,
      "mean_token_accuracy": 0.18065753430128098,
      "num_tokens": 10818185.0,
      "step": 4720
    },
    {
      "entropy": 5.481802463531494,
      "epoch": 0.4538904899135447,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004985526539758158,
      "loss": 5.37,
      "mean_token_accuracy": 0.18981760442256929,
      "num_tokens": 10830203.0,
      "step": 4725
    },
    {
      "entropy": 5.455829095840454,
      "epoch": 0.4543707973102786,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004985487690168263,
      "loss": 5.4339,
      "mean_token_accuracy": 0.17300689220428467,
      "num_tokens": 10842953.0,
      "step": 4730
    },
    {
      "entropy": 5.5628132820129395,
      "epoch": 0.4548511047070125,
      "grad_norm": 1.859375,
      "learning_rate": 0.000498544878867698,
      "loss": 5.5556,
      "mean_token_accuracy": 0.17378109693527222,
      "num_tokens": 10853322.0,
      "step": 4735
    },
    {
      "entropy": 5.536968040466308,
      "epoch": 0.4553314121037464,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004985409835285215,
      "loss": 5.4497,
      "mean_token_accuracy": 0.17889120429754257,
      "num_tokens": 10865480.0,
      "step": 4740
    },
    {
      "entropy": 5.435639333724976,
      "epoch": 0.45581171950048033,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004985370829993873,
      "loss": 5.3119,
      "mean_token_accuracy": 0.19336266964673995,
      "num_tokens": 10875700.0,
      "step": 4745
    },
    {
      "entropy": 5.419657135009766,
      "epoch": 0.4562920268972142,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004985331772803857,
      "loss": 5.4154,
      "mean_token_accuracy": 0.1836688607931137,
      "num_tokens": 10887350.0,
      "step": 4750
    },
    {
      "entropy": 5.450266981124878,
      "epoch": 0.45677233429394815,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004985292663716074,
      "loss": 5.3357,
      "mean_token_accuracy": 0.18904644697904588,
      "num_tokens": 10899355.0,
      "step": 4755
    },
    {
      "entropy": 5.427554178237915,
      "epoch": 0.457252641690682,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004985253502731435,
      "loss": 5.367,
      "mean_token_accuracy": 0.18772731721401215,
      "num_tokens": 10911521.0,
      "step": 4760
    },
    {
      "entropy": 5.424086999893189,
      "epoch": 0.45773294908741596,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004985214289850845,
      "loss": 5.4139,
      "mean_token_accuracy": 0.1889353096485138,
      "num_tokens": 10923363.0,
      "step": 4765
    },
    {
      "entropy": 5.470712327957154,
      "epoch": 0.45821325648414984,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004985175025075217,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1832897409796715,
      "num_tokens": 10934965.0,
      "step": 4770
    },
    {
      "entropy": 5.458177661895752,
      "epoch": 0.4586935638808838,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004985135708405462,
      "loss": 5.4215,
      "mean_token_accuracy": 0.17735306471586226,
      "num_tokens": 10946126.0,
      "step": 4775
    },
    {
      "entropy": 5.489032363891601,
      "epoch": 0.45917387127761766,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004985096339842493,
      "loss": 5.4978,
      "mean_token_accuracy": 0.17940519899129867,
      "num_tokens": 10957667.0,
      "step": 4780
    },
    {
      "entropy": 5.597390937805176,
      "epoch": 0.4596541786743516,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004985056919387224,
      "loss": 5.5694,
      "mean_token_accuracy": 0.17519876211881638,
      "num_tokens": 10969999.0,
      "step": 4785
    },
    {
      "entropy": 5.3790318965911865,
      "epoch": 0.4601344860710855,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004985017447040569,
      "loss": 5.2591,
      "mean_token_accuracy": 0.19265517294406892,
      "num_tokens": 10980672.0,
      "step": 4790
    },
    {
      "entropy": 5.436908149719239,
      "epoch": 0.4606147934678194,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004984977922803447,
      "loss": 5.3816,
      "mean_token_accuracy": 0.18623633831739425,
      "num_tokens": 10991471.0,
      "step": 4795
    },
    {
      "entropy": 5.489195775985718,
      "epoch": 0.4610951008645533,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004984938346676772,
      "loss": 5.4822,
      "mean_token_accuracy": 0.18003522753715515,
      "num_tokens": 11001960.0,
      "step": 4800
    },
    {
      "entropy": 5.520635604858398,
      "epoch": 0.46157540826128723,
      "grad_norm": 1.25,
      "learning_rate": 0.0004984898718661468,
      "loss": 5.3385,
      "mean_token_accuracy": 0.1887583240866661,
      "num_tokens": 11013274.0,
      "step": 4805
    },
    {
      "entropy": 5.454546737670898,
      "epoch": 0.4620557156580211,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004984859038758451,
      "loss": 5.3459,
      "mean_token_accuracy": 0.195092011988163,
      "num_tokens": 11024635.0,
      "step": 4810
    },
    {
      "entropy": 5.409391307830811,
      "epoch": 0.46253602305475505,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004984819306968642,
      "loss": 5.3932,
      "mean_token_accuracy": 0.18462751805782318,
      "num_tokens": 11036317.0,
      "step": 4815
    },
    {
      "entropy": 5.492571496963501,
      "epoch": 0.46301633045148893,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004984779523292966,
      "loss": 5.3891,
      "mean_token_accuracy": 0.1878635674715042,
      "num_tokens": 11047776.0,
      "step": 4820
    },
    {
      "entropy": 5.451682281494141,
      "epoch": 0.46349663784822287,
      "grad_norm": 2.875,
      "learning_rate": 0.0004984739687732345,
      "loss": 5.4143,
      "mean_token_accuracy": 0.18734978139400482,
      "num_tokens": 11058746.0,
      "step": 4825
    },
    {
      "entropy": 5.479636383056641,
      "epoch": 0.46397694524495675,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004984699800287705,
      "loss": 5.4359,
      "mean_token_accuracy": 0.18848110288381575,
      "num_tokens": 11068892.0,
      "step": 4830
    },
    {
      "entropy": 5.451904296875,
      "epoch": 0.4644572526416907,
      "grad_norm": 1.578125,
      "learning_rate": 0.000498465986095997,
      "loss": 5.4181,
      "mean_token_accuracy": 0.18504210859537124,
      "num_tokens": 11080038.0,
      "step": 4835
    },
    {
      "entropy": 5.445950746536255,
      "epoch": 0.46493756003842457,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004984619869750069,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1845824599266052,
      "num_tokens": 11091261.0,
      "step": 4840
    },
    {
      "entropy": 5.3959002017974855,
      "epoch": 0.4654178674351585,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000498457982665893,
      "loss": 5.3326,
      "mean_token_accuracy": 0.19228337705135345,
      "num_tokens": 11103015.0,
      "step": 4845
    },
    {
      "entropy": 5.444662761688233,
      "epoch": 0.4658981748318924,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004984539731687483,
      "loss": 5.3217,
      "mean_token_accuracy": 0.19602457731962203,
      "num_tokens": 11115075.0,
      "step": 4850
    },
    {
      "entropy": 5.461604499816895,
      "epoch": 0.4663784822286263,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004984499584836659,
      "loss": 5.2958,
      "mean_token_accuracy": 0.1870917484164238,
      "num_tokens": 11125937.0,
      "step": 4855
    },
    {
      "entropy": 5.425907611846924,
      "epoch": 0.4668587896253602,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000498445938610739,
      "loss": 5.3892,
      "mean_token_accuracy": 0.1811567708849907,
      "num_tokens": 11136863.0,
      "step": 4860
    },
    {
      "entropy": 5.509153938293457,
      "epoch": 0.46733909702209414,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004984419135500608,
      "loss": 5.4526,
      "mean_token_accuracy": 0.18005056977272033,
      "num_tokens": 11149934.0,
      "step": 4865
    },
    {
      "entropy": 5.434760904312133,
      "epoch": 0.4678194044188281,
      "grad_norm": 1.25,
      "learning_rate": 0.0004984378833017249,
      "loss": 5.3928,
      "mean_token_accuracy": 0.18313904404640197,
      "num_tokens": 11162208.0,
      "step": 4870
    },
    {
      "entropy": 5.437519550323486,
      "epoch": 0.46829971181556196,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004984338478658248,
      "loss": 5.4233,
      "mean_token_accuracy": 0.1863478258252144,
      "num_tokens": 11172892.0,
      "step": 4875
    },
    {
      "entropy": 5.48405089378357,
      "epoch": 0.4687800192122959,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004984298072424542,
      "loss": 5.3743,
      "mean_token_accuracy": 0.19089971333742142,
      "num_tokens": 11184339.0,
      "step": 4880
    },
    {
      "entropy": 5.48475399017334,
      "epoch": 0.4692603266090298,
      "grad_norm": 1.4375,
      "learning_rate": 0.000498425761431707,
      "loss": 5.4657,
      "mean_token_accuracy": 0.18189417570829391,
      "num_tokens": 11195681.0,
      "step": 4885
    },
    {
      "entropy": 5.503478860855102,
      "epoch": 0.4697406340057637,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498421710433677,
      "loss": 5.4012,
      "mean_token_accuracy": 0.17995550185441972,
      "num_tokens": 11207982.0,
      "step": 4890
    },
    {
      "entropy": 5.480038022994995,
      "epoch": 0.4702209414024976,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004984176542484584,
      "loss": 5.4397,
      "mean_token_accuracy": 0.18344718217849731,
      "num_tokens": 11218901.0,
      "step": 4895
    },
    {
      "entropy": 5.442627573013306,
      "epoch": 0.47070124879923153,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004984135928761452,
      "loss": 5.4108,
      "mean_token_accuracy": 0.18672479391098024,
      "num_tokens": 11230764.0,
      "step": 4900
    },
    {
      "entropy": 5.4512451648712155,
      "epoch": 0.4711815561959654,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004984095263168317,
      "loss": 5.3338,
      "mean_token_accuracy": 0.18761596381664275,
      "num_tokens": 11241489.0,
      "step": 4905
    },
    {
      "entropy": 5.341250276565551,
      "epoch": 0.47166186359269935,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004984054545706124,
      "loss": 5.3221,
      "mean_token_accuracy": 0.19300410151481628,
      "num_tokens": 11254500.0,
      "step": 4910
    },
    {
      "entropy": 5.531213903427124,
      "epoch": 0.47214217098943323,
      "grad_norm": 1.515625,
      "learning_rate": 0.000498401377637582,
      "loss": 5.4441,
      "mean_token_accuracy": 0.18692879676818847,
      "num_tokens": 11266235.0,
      "step": 4915
    },
    {
      "entropy": 5.517435598373413,
      "epoch": 0.47262247838616717,
      "grad_norm": 1.5,
      "learning_rate": 0.000498397295517835,
      "loss": 5.4033,
      "mean_token_accuracy": 0.19072997123003005,
      "num_tokens": 11277471.0,
      "step": 4920
    },
    {
      "entropy": 5.325749921798706,
      "epoch": 0.47310278578290105,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004983932082114659,
      "loss": 5.3297,
      "mean_token_accuracy": 0.1865825429558754,
      "num_tokens": 11290501.0,
      "step": 4925
    },
    {
      "entropy": 5.427298593521118,
      "epoch": 0.473583093179635,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004983891157185699,
      "loss": 5.3611,
      "mean_token_accuracy": 0.1869247242808342,
      "num_tokens": 11301933.0,
      "step": 4930
    },
    {
      "entropy": 5.552150440216065,
      "epoch": 0.47406340057636887,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004983850180392421,
      "loss": 5.4265,
      "mean_token_accuracy": 0.18141084909439087,
      "num_tokens": 11314539.0,
      "step": 4935
    },
    {
      "entropy": 5.398589944839477,
      "epoch": 0.4745437079731028,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004983809151735775,
      "loss": 5.346,
      "mean_token_accuracy": 0.1900244951248169,
      "num_tokens": 11326641.0,
      "step": 4940
    },
    {
      "entropy": 5.445167350769043,
      "epoch": 0.4750240153698367,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004983768071216713,
      "loss": 5.4338,
      "mean_token_accuracy": 0.1810634598135948,
      "num_tokens": 11337214.0,
      "step": 4945
    },
    {
      "entropy": 5.495223093032837,
      "epoch": 0.4755043227665706,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004983726938836189,
      "loss": 5.3901,
      "mean_token_accuracy": 0.18667189180850982,
      "num_tokens": 11349540.0,
      "step": 4950
    },
    {
      "entropy": 5.492793083190918,
      "epoch": 0.4759846301633045,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004983685754595159,
      "loss": 5.3723,
      "mean_token_accuracy": 0.18532039523124694,
      "num_tokens": 11359817.0,
      "step": 4955
    },
    {
      "entropy": 5.462947273254395,
      "epoch": 0.47646493756003844,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004983644518494578,
      "loss": 5.4482,
      "mean_token_accuracy": 0.1787272647023201,
      "num_tokens": 11371120.0,
      "step": 4960
    },
    {
      "entropy": 5.430369329452515,
      "epoch": 0.4769452449567723,
      "grad_norm": 1.5,
      "learning_rate": 0.0004983603230535403,
      "loss": 5.4297,
      "mean_token_accuracy": 0.17996030002832414,
      "num_tokens": 11382071.0,
      "step": 4965
    },
    {
      "entropy": 5.5358741760253904,
      "epoch": 0.47742555235350626,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004983561890718594,
      "loss": 5.4384,
      "mean_token_accuracy": 0.1795857608318329,
      "num_tokens": 11392938.0,
      "step": 4970
    },
    {
      "entropy": 5.482577991485596,
      "epoch": 0.47790585975024014,
      "grad_norm": 1.328125,
      "learning_rate": 0.000498352049904511,
      "loss": 5.4471,
      "mean_token_accuracy": 0.18274903744459153,
      "num_tokens": 11405416.0,
      "step": 4975
    },
    {
      "entropy": 5.488064527511597,
      "epoch": 0.4783861671469741,
      "grad_norm": 1.375,
      "learning_rate": 0.0004983479055515914,
      "loss": 5.4351,
      "mean_token_accuracy": 0.18943743258714676,
      "num_tokens": 11417008.0,
      "step": 4980
    },
    {
      "entropy": 5.544620180130005,
      "epoch": 0.47886647454370795,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004983437560131964,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1895648792386055,
      "num_tokens": 11428154.0,
      "step": 4985
    },
    {
      "entropy": 5.487364006042481,
      "epoch": 0.4793467819404419,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004983396012894228,
      "loss": 5.4052,
      "mean_token_accuracy": 0.18201917707920073,
      "num_tokens": 11438559.0,
      "step": 4990
    },
    {
      "entropy": 5.533839893341065,
      "epoch": 0.47982708933717577,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004983354413803666,
      "loss": 5.478,
      "mean_token_accuracy": 0.18239017128944396,
      "num_tokens": 11449501.0,
      "step": 4995
    },
    {
      "entropy": 5.43055911064148,
      "epoch": 0.4803073967339097,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004983312762861248,
      "loss": 5.4659,
      "mean_token_accuracy": 0.17930015325546264,
      "num_tokens": 11461841.0,
      "step": 5000
    },
    {
      "entropy": 5.444394779205322,
      "epoch": 0.4807877041306436,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004983271060067939,
      "loss": 5.3722,
      "mean_token_accuracy": 0.18473062217235564,
      "num_tokens": 11472876.0,
      "step": 5005
    },
    {
      "entropy": 5.47320032119751,
      "epoch": 0.4812680115273775,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004983229305424707,
      "loss": 5.4354,
      "mean_token_accuracy": 0.18353282213211058,
      "num_tokens": 11484541.0,
      "step": 5010
    },
    {
      "entropy": 5.386233186721801,
      "epoch": 0.4817483189241114,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004983187498932522,
      "loss": 5.2905,
      "mean_token_accuracy": 0.18736893236637114,
      "num_tokens": 11496071.0,
      "step": 5015
    },
    {
      "entropy": 5.441421842575073,
      "epoch": 0.48222862632084534,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004983145640592354,
      "loss": 5.3636,
      "mean_token_accuracy": 0.1883857563138008,
      "num_tokens": 11507574.0,
      "step": 5020
    },
    {
      "entropy": 5.440950775146485,
      "epoch": 0.4827089337175792,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004983103730405176,
      "loss": 5.4095,
      "mean_token_accuracy": 0.18637361377477646,
      "num_tokens": 11519738.0,
      "step": 5025
    },
    {
      "entropy": 5.465356111526489,
      "epoch": 0.48318924111431316,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498306176837196,
      "loss": 5.4131,
      "mean_token_accuracy": 0.1857826292514801,
      "num_tokens": 11530669.0,
      "step": 5030
    },
    {
      "entropy": 5.364138412475586,
      "epoch": 0.48366954851104704,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004983019754493681,
      "loss": 5.3045,
      "mean_token_accuracy": 0.1915925696492195,
      "num_tokens": 11541680.0,
      "step": 5035
    },
    {
      "entropy": 5.434883451461792,
      "epoch": 0.484149855907781,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004982977688771314,
      "loss": 5.3595,
      "mean_token_accuracy": 0.18698243647813798,
      "num_tokens": 11553204.0,
      "step": 5040
    },
    {
      "entropy": 5.494551324844361,
      "epoch": 0.4846301633045149,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004982935571205835,
      "loss": 5.4069,
      "mean_token_accuracy": 0.17969363927841187,
      "num_tokens": 11564841.0,
      "step": 5045
    },
    {
      "entropy": 5.432051181793213,
      "epoch": 0.4851104707012488,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004982893401798223,
      "loss": 5.4092,
      "mean_token_accuracy": 0.18372393250465394,
      "num_tokens": 11576723.0,
      "step": 5050
    },
    {
      "entropy": 5.507033395767212,
      "epoch": 0.48559077809798273,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004982851180549456,
      "loss": 5.3741,
      "mean_token_accuracy": 0.18664456009864808,
      "num_tokens": 11587727.0,
      "step": 5055
    },
    {
      "entropy": 5.383077955245971,
      "epoch": 0.4860710854947166,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004982808907460515,
      "loss": 5.363,
      "mean_token_accuracy": 0.1817546159029007,
      "num_tokens": 11598508.0,
      "step": 5060
    },
    {
      "entropy": 5.344477891921997,
      "epoch": 0.48655139289145055,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004982766582532382,
      "loss": 5.2982,
      "mean_token_accuracy": 0.19086076617240905,
      "num_tokens": 11609684.0,
      "step": 5065
    },
    {
      "entropy": 5.481044912338257,
      "epoch": 0.48703170028818443,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004982724205766038,
      "loss": 5.4351,
      "mean_token_accuracy": 0.17675031274557113,
      "num_tokens": 11621642.0,
      "step": 5070
    },
    {
      "entropy": 5.466136455535889,
      "epoch": 0.48751200768491837,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004982681777162468,
      "loss": 5.3546,
      "mean_token_accuracy": 0.18459942936897278,
      "num_tokens": 11632778.0,
      "step": 5075
    },
    {
      "entropy": 5.385703611373901,
      "epoch": 0.48799231508165225,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004982639296722657,
      "loss": 5.3424,
      "mean_token_accuracy": 0.19703588932752608,
      "num_tokens": 11643581.0,
      "step": 5080
    },
    {
      "entropy": 5.552964448928833,
      "epoch": 0.4884726224783862,
      "grad_norm": 1.5,
      "learning_rate": 0.0004982596764447591,
      "loss": 5.5227,
      "mean_token_accuracy": 0.17867382615804672,
      "num_tokens": 11656126.0,
      "step": 5085
    },
    {
      "entropy": 5.5898202419281,
      "epoch": 0.48895292987512007,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004982554180338258,
      "loss": 5.4303,
      "mean_token_accuracy": 0.1850968047976494,
      "num_tokens": 11667765.0,
      "step": 5090
    },
    {
      "entropy": 5.511855411529541,
      "epoch": 0.489433237271854,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004982511544395646,
      "loss": 5.4424,
      "mean_token_accuracy": 0.17825420945882797,
      "num_tokens": 11679268.0,
      "step": 5095
    },
    {
      "entropy": 5.415318298339844,
      "epoch": 0.4899135446685879,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004982468856620745,
      "loss": 5.3523,
      "mean_token_accuracy": 0.1884788230061531,
      "num_tokens": 11690049.0,
      "step": 5100
    },
    {
      "entropy": 5.434323644638061,
      "epoch": 0.4903938520653218,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004982426117014545,
      "loss": 5.3592,
      "mean_token_accuracy": 0.1916387513279915,
      "num_tokens": 11701782.0,
      "step": 5105
    },
    {
      "entropy": 5.389872455596924,
      "epoch": 0.4908741594620557,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004982383325578041,
      "loss": 5.373,
      "mean_token_accuracy": 0.1935378208756447,
      "num_tokens": 11712488.0,
      "step": 5110
    },
    {
      "entropy": 5.478123521804809,
      "epoch": 0.49135446685878964,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004982340482312226,
      "loss": 5.4016,
      "mean_token_accuracy": 0.18805212676525115,
      "num_tokens": 11724036.0,
      "step": 5115
    },
    {
      "entropy": 5.4282684326171875,
      "epoch": 0.4918347742555235,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004982297587218092,
      "loss": 5.3817,
      "mean_token_accuracy": 0.18427670001983643,
      "num_tokens": 11736461.0,
      "step": 5120
    },
    {
      "entropy": 5.36871337890625,
      "epoch": 0.49231508165225746,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004982254640296637,
      "loss": 5.3203,
      "mean_token_accuracy": 0.18877253979444503,
      "num_tokens": 11747549.0,
      "step": 5125
    },
    {
      "entropy": 5.520420217514038,
      "epoch": 0.49279538904899134,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004982211641548857,
      "loss": 5.4773,
      "mean_token_accuracy": 0.1765229657292366,
      "num_tokens": 11759012.0,
      "step": 5130
    },
    {
      "entropy": 5.555425024032592,
      "epoch": 0.4932756964457253,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004982168590975752,
      "loss": 5.4098,
      "mean_token_accuracy": 0.18458688855171204,
      "num_tokens": 11770535.0,
      "step": 5135
    },
    {
      "entropy": 5.475128078460694,
      "epoch": 0.49375600384245916,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004982125488578321,
      "loss": 5.4265,
      "mean_token_accuracy": 0.18305824548006058,
      "num_tokens": 11783708.0,
      "step": 5140
    },
    {
      "entropy": 5.444312715530396,
      "epoch": 0.4942363112391931,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004982082334357563,
      "loss": 5.3932,
      "mean_token_accuracy": 0.18637588024139404,
      "num_tokens": 11794565.0,
      "step": 5145
    },
    {
      "entropy": 5.5029500961303714,
      "epoch": 0.494716618635927,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004982039128314481,
      "loss": 5.3979,
      "mean_token_accuracy": 0.18649012595415115,
      "num_tokens": 11805631.0,
      "step": 5150
    },
    {
      "entropy": 5.515172100067138,
      "epoch": 0.4951969260326609,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004981995870450079,
      "loss": 5.4343,
      "mean_token_accuracy": 0.18024242222309111,
      "num_tokens": 11817041.0,
      "step": 5155
    },
    {
      "entropy": 5.471421098709106,
      "epoch": 0.4956772334293948,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004981952560765361,
      "loss": 5.3739,
      "mean_token_accuracy": 0.1874469444155693,
      "num_tokens": 11829090.0,
      "step": 5160
    },
    {
      "entropy": 5.560026025772094,
      "epoch": 0.49615754082612873,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004981909199261331,
      "loss": 5.4969,
      "mean_token_accuracy": 0.17861385643482208,
      "num_tokens": 11839980.0,
      "step": 5165
    },
    {
      "entropy": 5.473791790008545,
      "epoch": 0.4966378482228626,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004981865785938998,
      "loss": 5.3998,
      "mean_token_accuracy": 0.18403689414262772,
      "num_tokens": 11851217.0,
      "step": 5170
    },
    {
      "entropy": 5.435330820083618,
      "epoch": 0.49711815561959655,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004981822320799367,
      "loss": 5.4089,
      "mean_token_accuracy": 0.1848284050822258,
      "num_tokens": 11862506.0,
      "step": 5175
    },
    {
      "entropy": 5.511408042907715,
      "epoch": 0.49759846301633043,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004981778803843449,
      "loss": 5.4551,
      "mean_token_accuracy": 0.17914380580186845,
      "num_tokens": 11874844.0,
      "step": 5180
    },
    {
      "entropy": 5.420107173919678,
      "epoch": 0.49807877041306436,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004981735235072256,
      "loss": 5.3388,
      "mean_token_accuracy": 0.19396358132362365,
      "num_tokens": 11885231.0,
      "step": 5185
    },
    {
      "entropy": 5.479499053955078,
      "epoch": 0.49855907780979825,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004981691614486796,
      "loss": 5.4099,
      "mean_token_accuracy": 0.18174998164176942,
      "num_tokens": 11897222.0,
      "step": 5190
    },
    {
      "entropy": 5.432892322540283,
      "epoch": 0.4990393852065322,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004981647942088084,
      "loss": 5.3658,
      "mean_token_accuracy": 0.18804378658533097,
      "num_tokens": 11909493.0,
      "step": 5195
    },
    {
      "entropy": 5.518645811080932,
      "epoch": 0.49951969260326606,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004981604217877135,
      "loss": 5.3594,
      "mean_token_accuracy": 0.18577739745378494,
      "num_tokens": 11920833.0,
      "step": 5200
    },
    {
      "entropy": 5.396304130554199,
      "epoch": 0.5,
      "grad_norm": 1.65625,
      "learning_rate": 0.000498156044185496,
      "loss": 5.2715,
      "mean_token_accuracy": 0.19440183341503142,
      "num_tokens": 11932910.0,
      "step": 5205
    },
    {
      "entropy": 5.468098115921021,
      "epoch": 0.5004803073967339,
      "grad_norm": 1.25,
      "learning_rate": 0.0004981516614022579,
      "loss": 5.3945,
      "mean_token_accuracy": 0.19133666008710862,
      "num_tokens": 11944536.0,
      "step": 5210
    },
    {
      "entropy": 5.420602512359619,
      "epoch": 0.5009606147934679,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004981472734381008,
      "loss": 5.394,
      "mean_token_accuracy": 0.1924171507358551,
      "num_tokens": 11955724.0,
      "step": 5215
    },
    {
      "entropy": 5.436355924606323,
      "epoch": 0.5014409221902018,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004981428802931267,
      "loss": 5.3528,
      "mean_token_accuracy": 0.18886628448963166,
      "num_tokens": 11965921.0,
      "step": 5220
    },
    {
      "entropy": 5.398175191879273,
      "epoch": 0.5019212295869356,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004981384819674375,
      "loss": 5.3237,
      "mean_token_accuracy": 0.19049944132566451,
      "num_tokens": 11976356.0,
      "step": 5225
    },
    {
      "entropy": 5.450118923187256,
      "epoch": 0.5024015369836695,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004981340784611354,
      "loss": 5.382,
      "mean_token_accuracy": 0.18158023357391356,
      "num_tokens": 11987959.0,
      "step": 5230
    },
    {
      "entropy": 5.474130916595459,
      "epoch": 0.5028818443804035,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004981296697743224,
      "loss": 5.3966,
      "mean_token_accuracy": 0.18466383665800096,
      "num_tokens": 12000089.0,
      "step": 5235
    },
    {
      "entropy": 5.477195882797242,
      "epoch": 0.5033621517771374,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004981252559071012,
      "loss": 5.4641,
      "mean_token_accuracy": 0.18619239926338196,
      "num_tokens": 12011286.0,
      "step": 5240
    },
    {
      "entropy": 5.401328229904175,
      "epoch": 0.5038424591738713,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004981208368595739,
      "loss": 5.336,
      "mean_token_accuracy": 0.18776115477085115,
      "num_tokens": 12022809.0,
      "step": 5245
    },
    {
      "entropy": 5.441013193130493,
      "epoch": 0.5043227665706052,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004981164126318435,
      "loss": 5.3274,
      "mean_token_accuracy": 0.1900719553232193,
      "num_tokens": 12034311.0,
      "step": 5250
    },
    {
      "entropy": 5.412341165542602,
      "epoch": 0.5048030739673391,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004981119832240124,
      "loss": 5.3303,
      "mean_token_accuracy": 0.19273549765348436,
      "num_tokens": 12046405.0,
      "step": 5255
    },
    {
      "entropy": 5.427926826477051,
      "epoch": 0.505283381364073,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004981075486361837,
      "loss": 5.3797,
      "mean_token_accuracy": 0.18678055107593536,
      "num_tokens": 12059022.0,
      "step": 5260
    },
    {
      "entropy": 5.438950443267823,
      "epoch": 0.5057636887608069,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004981031088684601,
      "loss": 5.394,
      "mean_token_accuracy": 0.1849857747554779,
      "num_tokens": 12070666.0,
      "step": 5265
    },
    {
      "entropy": 5.455779361724853,
      "epoch": 0.5062439961575408,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004980986639209448,
      "loss": 5.3854,
      "mean_token_accuracy": 0.18551258444786073,
      "num_tokens": 12082702.0,
      "step": 5270
    },
    {
      "entropy": 5.457168388366699,
      "epoch": 0.5067243035542748,
      "grad_norm": 1.3125,
      "learning_rate": 0.000498094213793741,
      "loss": 5.4005,
      "mean_token_accuracy": 0.18254365026950836,
      "num_tokens": 12094034.0,
      "step": 5275
    },
    {
      "entropy": 5.502580165863037,
      "epoch": 0.5072046109510087,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498089758486952,
      "loss": 5.3797,
      "mean_token_accuracy": 0.17486527115106582,
      "num_tokens": 12104597.0,
      "step": 5280
    },
    {
      "entropy": 5.479168128967285,
      "epoch": 0.5076849183477425,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004980852980006812,
      "loss": 5.3881,
      "mean_token_accuracy": 0.18257687538862227,
      "num_tokens": 12116577.0,
      "step": 5285
    },
    {
      "entropy": 5.413305711746216,
      "epoch": 0.5081652257444764,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004980808323350323,
      "loss": 5.3267,
      "mean_token_accuracy": 0.18752931505441667,
      "num_tokens": 12128005.0,
      "step": 5290
    },
    {
      "entropy": 5.475668621063233,
      "epoch": 0.5086455331412104,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004980763614901089,
      "loss": 5.4374,
      "mean_token_accuracy": 0.18278273642063142,
      "num_tokens": 12138591.0,
      "step": 5295
    },
    {
      "entropy": 5.431951951980591,
      "epoch": 0.5091258405379443,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004980718854660146,
      "loss": 5.4224,
      "mean_token_accuracy": 0.18703065663576127,
      "num_tokens": 12150010.0,
      "step": 5300
    },
    {
      "entropy": 5.493814706802368,
      "epoch": 0.5096061479346782,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004980674042628537,
      "loss": 5.3917,
      "mean_token_accuracy": 0.17938946783542634,
      "num_tokens": 12161268.0,
      "step": 5305
    },
    {
      "entropy": 5.451682090759277,
      "epoch": 0.5100864553314121,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00049806291788073,
      "loss": 5.345,
      "mean_token_accuracy": 0.1900179147720337,
      "num_tokens": 12172647.0,
      "step": 5310
    },
    {
      "entropy": 5.397039413452148,
      "epoch": 0.510566762728146,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004980584263197477,
      "loss": 5.4469,
      "mean_token_accuracy": 0.1861713707447052,
      "num_tokens": 12183521.0,
      "step": 5315
    },
    {
      "entropy": 5.595460653305054,
      "epoch": 0.5110470701248799,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004980539295800111,
      "loss": 5.4538,
      "mean_token_accuracy": 0.177712544798851,
      "num_tokens": 12193616.0,
      "step": 5320
    },
    {
      "entropy": 5.457161283493042,
      "epoch": 0.5115273775216138,
      "grad_norm": 1.375,
      "learning_rate": 0.0004980494276616246,
      "loss": 5.3989,
      "mean_token_accuracy": 0.18704742044210435,
      "num_tokens": 12204817.0,
      "step": 5325
    },
    {
      "entropy": 5.471468687057495,
      "epoch": 0.5120076849183477,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004980449205646926,
      "loss": 5.4626,
      "mean_token_accuracy": 0.17779629975557326,
      "num_tokens": 12217281.0,
      "step": 5330
    },
    {
      "entropy": 5.4687206745147705,
      "epoch": 0.5124879923150817,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00049804040828932,
      "loss": 5.3687,
      "mean_token_accuracy": 0.1845517635345459,
      "num_tokens": 12228705.0,
      "step": 5335
    },
    {
      "entropy": 5.50502576828003,
      "epoch": 0.5129682997118156,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004980358908356113,
      "loss": 5.4579,
      "mean_token_accuracy": 0.17887165397405624,
      "num_tokens": 12240010.0,
      "step": 5340
    },
    {
      "entropy": 5.492001867294311,
      "epoch": 0.5134486071085494,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004980313682036717,
      "loss": 5.5088,
      "mean_token_accuracy": 0.17649380415678023,
      "num_tokens": 12251384.0,
      "step": 5345
    },
    {
      "entropy": 5.450707054138183,
      "epoch": 0.5139289145052833,
      "grad_norm": 1.5,
      "learning_rate": 0.0004980268403936058,
      "loss": 5.3534,
      "mean_token_accuracy": 0.18193223625421523,
      "num_tokens": 12262548.0,
      "step": 5350
    },
    {
      "entropy": 5.434090042114258,
      "epoch": 0.5144092219020173,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004980223074055189,
      "loss": 5.3139,
      "mean_token_accuracy": 0.19134428352117538,
      "num_tokens": 12274022.0,
      "step": 5355
    },
    {
      "entropy": 5.3991340637207035,
      "epoch": 0.5148895292987512,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004980177692395164,
      "loss": 5.2867,
      "mean_token_accuracy": 0.18951193541288375,
      "num_tokens": 12285000.0,
      "step": 5360
    },
    {
      "entropy": 5.338136863708496,
      "epoch": 0.5153698366954851,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004980132258957035,
      "loss": 5.3477,
      "mean_token_accuracy": 0.19007130116224288,
      "num_tokens": 12297472.0,
      "step": 5365
    },
    {
      "entropy": 5.503325796127319,
      "epoch": 0.515850144092219,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004980086773741856,
      "loss": 5.3939,
      "mean_token_accuracy": 0.18749501258134843,
      "num_tokens": 12309310.0,
      "step": 5370
    },
    {
      "entropy": 5.525248384475708,
      "epoch": 0.516330451488953,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004980041236750685,
      "loss": 5.4636,
      "mean_token_accuracy": 0.18202975690364837,
      "num_tokens": 12321897.0,
      "step": 5375
    },
    {
      "entropy": 5.373438596725464,
      "epoch": 0.5168107588856868,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004979995647984577,
      "loss": 5.2986,
      "mean_token_accuracy": 0.19803487062454223,
      "num_tokens": 12333484.0,
      "step": 5380
    },
    {
      "entropy": 5.407309484481812,
      "epoch": 0.5172910662824207,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004979950007444593,
      "loss": 5.3829,
      "mean_token_accuracy": 0.18875620067119597,
      "num_tokens": 12345154.0,
      "step": 5385
    },
    {
      "entropy": 5.531070184707642,
      "epoch": 0.5177713736791547,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004979904315131792,
      "loss": 5.3985,
      "mean_token_accuracy": 0.18682202696800232,
      "num_tokens": 12354934.0,
      "step": 5390
    },
    {
      "entropy": 5.580181455612182,
      "epoch": 0.5182516810758886,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004979858571047233,
      "loss": 5.5701,
      "mean_token_accuracy": 0.17747551798820496,
      "num_tokens": 12367729.0,
      "step": 5395
    },
    {
      "entropy": 5.537011623382568,
      "epoch": 0.5187319884726225,
      "grad_norm": 1.625,
      "learning_rate": 0.0004979812775191979,
      "loss": 5.3717,
      "mean_token_accuracy": 0.18911508321762086,
      "num_tokens": 12380366.0,
      "step": 5400
    },
    {
      "entropy": 5.4212806224823,
      "epoch": 0.5192122958693564,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004979766927567094,
      "loss": 5.3991,
      "mean_token_accuracy": 0.185391803085804,
      "num_tokens": 12392338.0,
      "step": 5405
    },
    {
      "entropy": 5.411877775192261,
      "epoch": 0.5196926032660903,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004979721028173643,
      "loss": 5.3571,
      "mean_token_accuracy": 0.19052527099847794,
      "num_tokens": 12403280.0,
      "step": 5410
    },
    {
      "entropy": 5.428349351882934,
      "epoch": 0.5201729106628242,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000497967507701269,
      "loss": 5.339,
      "mean_token_accuracy": 0.19180969446897506,
      "num_tokens": 12414279.0,
      "step": 5415
    },
    {
      "entropy": 5.369792318344116,
      "epoch": 0.5206532180595581,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004979629074085303,
      "loss": 5.2574,
      "mean_token_accuracy": 0.19326327443122865,
      "num_tokens": 12425486.0,
      "step": 5420
    },
    {
      "entropy": 5.429877424240113,
      "epoch": 0.521133525456292,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004979583019392548,
      "loss": 5.3845,
      "mean_token_accuracy": 0.18277973234653472,
      "num_tokens": 12437295.0,
      "step": 5425
    },
    {
      "entropy": 5.488057661056518,
      "epoch": 0.521613832853026,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004979536912935497,
      "loss": 5.3409,
      "mean_token_accuracy": 0.19044878631830214,
      "num_tokens": 12448684.0,
      "step": 5430
    },
    {
      "entropy": 5.416075134277344,
      "epoch": 0.5220941402497599,
      "grad_norm": 1.265625,
      "learning_rate": 0.000497949075471522,
      "loss": 5.2904,
      "mean_token_accuracy": 0.1959683492779732,
      "num_tokens": 12461209.0,
      "step": 5435
    },
    {
      "entropy": 5.439440536499023,
      "epoch": 0.5225744476464937,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004979444544732786,
      "loss": 5.3735,
      "mean_token_accuracy": 0.18593170940876008,
      "num_tokens": 12472863.0,
      "step": 5440
    },
    {
      "entropy": 5.487802410125733,
      "epoch": 0.5230547550432276,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000497939828298927,
      "loss": 5.3999,
      "mean_token_accuracy": 0.18451804667711258,
      "num_tokens": 12483510.0,
      "step": 5445
    },
    {
      "entropy": 5.566813945770264,
      "epoch": 0.5235350624399616,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004979351969485747,
      "loss": 5.4393,
      "mean_token_accuracy": 0.1832005724310875,
      "num_tokens": 12495532.0,
      "step": 5450
    },
    {
      "entropy": 5.328968048095703,
      "epoch": 0.5240153698366955,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004979305604223291,
      "loss": 5.3321,
      "mean_token_accuracy": 0.18730416595935823,
      "num_tokens": 12508413.0,
      "step": 5455
    },
    {
      "entropy": 5.472768878936767,
      "epoch": 0.5244956772334294,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004979259187202978,
      "loss": 5.4258,
      "mean_token_accuracy": 0.18066055029630662,
      "num_tokens": 12520705.0,
      "step": 5460
    },
    {
      "entropy": 5.515762042999268,
      "epoch": 0.5249759846301633,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004979212718425887,
      "loss": 5.3754,
      "mean_token_accuracy": 0.1898324728012085,
      "num_tokens": 12531560.0,
      "step": 5465
    },
    {
      "entropy": 5.4374120235443115,
      "epoch": 0.5254562920268973,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004979166197893096,
      "loss": 5.4019,
      "mean_token_accuracy": 0.1806342676281929,
      "num_tokens": 12542539.0,
      "step": 5470
    },
    {
      "entropy": 5.494349431991577,
      "epoch": 0.5259365994236311,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004979119625605683,
      "loss": 5.4264,
      "mean_token_accuracy": 0.17977063953876496,
      "num_tokens": 12555228.0,
      "step": 5475
    },
    {
      "entropy": 5.515709638595581,
      "epoch": 0.526416906820365,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004979073001564734,
      "loss": 5.3516,
      "mean_token_accuracy": 0.19102157950401305,
      "num_tokens": 12565715.0,
      "step": 5480
    },
    {
      "entropy": 5.459074687957764,
      "epoch": 0.5268972142170989,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004979026325771328,
      "loss": 5.2994,
      "mean_token_accuracy": 0.18871578127145766,
      "num_tokens": 12576896.0,
      "step": 5485
    },
    {
      "entropy": 5.411230421066284,
      "epoch": 0.5273775216138329,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004978979598226549,
      "loss": 5.3653,
      "mean_token_accuracy": 0.1876650035381317,
      "num_tokens": 12588117.0,
      "step": 5490
    },
    {
      "entropy": 5.381819152832032,
      "epoch": 0.5278578290105668,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004978932818931483,
      "loss": 5.3793,
      "mean_token_accuracy": 0.18598207086324692,
      "num_tokens": 12599987.0,
      "step": 5495
    },
    {
      "entropy": 5.484870481491089,
      "epoch": 0.5283381364073007,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004978885987887216,
      "loss": 5.457,
      "mean_token_accuracy": 0.18079727739095688,
      "num_tokens": 12611453.0,
      "step": 5500
    },
    {
      "entropy": 5.549768304824829,
      "epoch": 0.5288184438040345,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004978839105094833,
      "loss": 5.3891,
      "mean_token_accuracy": 0.1803484871983528,
      "num_tokens": 12623068.0,
      "step": 5505
    },
    {
      "entropy": 5.487051725387573,
      "epoch": 0.5292987512007685,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004978792170555426,
      "loss": 5.4396,
      "mean_token_accuracy": 0.18667725324630738,
      "num_tokens": 12634952.0,
      "step": 5510
    },
    {
      "entropy": 5.45725793838501,
      "epoch": 0.5297790585975024,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004978745184270083,
      "loss": 5.3807,
      "mean_token_accuracy": 0.18108554929494858,
      "num_tokens": 12647027.0,
      "step": 5515
    },
    {
      "entropy": 5.370086240768432,
      "epoch": 0.5302593659942363,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004978698146239893,
      "loss": 5.2388,
      "mean_token_accuracy": 0.19929148703813554,
      "num_tokens": 12657249.0,
      "step": 5520
    },
    {
      "entropy": 5.405987405776978,
      "epoch": 0.5307396733909702,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004978651056465952,
      "loss": 5.2768,
      "mean_token_accuracy": 0.19359083771705626,
      "num_tokens": 12666714.0,
      "step": 5525
    },
    {
      "entropy": 5.4099356174469,
      "epoch": 0.5312199807877042,
      "grad_norm": 1.3125,
      "learning_rate": 0.000497860391494935,
      "loss": 5.3664,
      "mean_token_accuracy": 0.18576207011938095,
      "num_tokens": 12678627.0,
      "step": 5530
    },
    {
      "entropy": 5.4683794498443605,
      "epoch": 0.531700288184438,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004978556721691183,
      "loss": 5.4143,
      "mean_token_accuracy": 0.18748129159212112,
      "num_tokens": 12691009.0,
      "step": 5535
    },
    {
      "entropy": 5.4890233993530275,
      "epoch": 0.5321805955811719,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004978509476692547,
      "loss": 5.3765,
      "mean_token_accuracy": 0.1840150013566017,
      "num_tokens": 12702503.0,
      "step": 5540
    },
    {
      "entropy": 5.5058454990386965,
      "epoch": 0.5326609029779059,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004978462179954538,
      "loss": 5.4243,
      "mean_token_accuracy": 0.1805101990699768,
      "num_tokens": 12714018.0,
      "step": 5545
    },
    {
      "entropy": 5.499725580215454,
      "epoch": 0.5331412103746398,
      "grad_norm": 1.625,
      "learning_rate": 0.0004978414831478253,
      "loss": 5.5094,
      "mean_token_accuracy": 0.18239614218473435,
      "num_tokens": 12726726.0,
      "step": 5550
    },
    {
      "entropy": 5.421793651580811,
      "epoch": 0.5336215177713737,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004978367431264794,
      "loss": 5.3691,
      "mean_token_accuracy": 0.18756531327962875,
      "num_tokens": 12737790.0,
      "step": 5555
    },
    {
      "entropy": 5.4131616115570065,
      "epoch": 0.5341018251681076,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004978319979315261,
      "loss": 5.255,
      "mean_token_accuracy": 0.19221504777669907,
      "num_tokens": 12749473.0,
      "step": 5560
    },
    {
      "entropy": 5.403030204772949,
      "epoch": 0.5345821325648416,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004978272475630752,
      "loss": 5.4012,
      "mean_token_accuracy": 0.18725448250770568,
      "num_tokens": 12762220.0,
      "step": 5565
    },
    {
      "entropy": 5.306041097640991,
      "epoch": 0.5350624399615754,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004978224920212374,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19876040071249007,
      "num_tokens": 12772788.0,
      "step": 5570
    },
    {
      "entropy": 5.314635419845581,
      "epoch": 0.5355427473583093,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004978177313061232,
      "loss": 5.2981,
      "mean_token_accuracy": 0.18902058899402618,
      "num_tokens": 12784552.0,
      "step": 5575
    },
    {
      "entropy": 5.4929139614105225,
      "epoch": 0.5360230547550432,
      "grad_norm": 1.5,
      "learning_rate": 0.0004978129654178426,
      "loss": 5.3497,
      "mean_token_accuracy": 0.19190627783536912,
      "num_tokens": 12797618.0,
      "step": 5580
    },
    {
      "entropy": 5.396257305145264,
      "epoch": 0.5365033621517772,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004978081943565067,
      "loss": 5.2794,
      "mean_token_accuracy": 0.18863389194011687,
      "num_tokens": 12807789.0,
      "step": 5585
    },
    {
      "entropy": 5.415760183334351,
      "epoch": 0.5369836695485111,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004978034181222261,
      "loss": 5.4194,
      "mean_token_accuracy": 0.18857093751430512,
      "num_tokens": 12819020.0,
      "step": 5590
    },
    {
      "entropy": 5.4485376358032225,
      "epoch": 0.537463976945245,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004977986367151119,
      "loss": 5.3736,
      "mean_token_accuracy": 0.18792831897735596,
      "num_tokens": 12829527.0,
      "step": 5595
    },
    {
      "entropy": 5.476880073547363,
      "epoch": 0.5379442843419788,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004977938501352747,
      "loss": 5.3655,
      "mean_token_accuracy": 0.18210554718971253,
      "num_tokens": 12840821.0,
      "step": 5600
    },
    {
      "entropy": 5.4507091522216795,
      "epoch": 0.5384245917387128,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004977890583828259,
      "loss": 5.3661,
      "mean_token_accuracy": 0.1881468653678894,
      "num_tokens": 12853314.0,
      "step": 5605
    },
    {
      "entropy": 5.414139366149902,
      "epoch": 0.5389048991354467,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004977842614578768,
      "loss": 5.349,
      "mean_token_accuracy": 0.18417798578739167,
      "num_tokens": 12863507.0,
      "step": 5610
    },
    {
      "entropy": 5.472694253921508,
      "epoch": 0.5393852065321806,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004977794593605386,
      "loss": 5.3854,
      "mean_token_accuracy": 0.18529417961835862,
      "num_tokens": 12874142.0,
      "step": 5615
    },
    {
      "entropy": 5.480929613113403,
      "epoch": 0.5398655139289145,
      "grad_norm": 2.5625,
      "learning_rate": 0.000497774652090923,
      "loss": 5.3651,
      "mean_token_accuracy": 0.19421031028032304,
      "num_tokens": 12884548.0,
      "step": 5620
    },
    {
      "entropy": 5.420640277862549,
      "epoch": 0.5403458213256485,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004977698396491414,
      "loss": 5.3841,
      "mean_token_accuracy": 0.1893293485045433,
      "num_tokens": 12895562.0,
      "step": 5625
    },
    {
      "entropy": 5.396530723571777,
      "epoch": 0.5408261287223823,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004977650220353055,
      "loss": 5.367,
      "mean_token_accuracy": 0.195559298992157,
      "num_tokens": 12905308.0,
      "step": 5630
    },
    {
      "entropy": 5.4512217998504635,
      "epoch": 0.5413064361191162,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004977601992495274,
      "loss": 5.3962,
      "mean_token_accuracy": 0.18294627964496613,
      "num_tokens": 12916731.0,
      "step": 5635
    },
    {
      "entropy": 5.576397562026978,
      "epoch": 0.5417867435158501,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004977553712919189,
      "loss": 5.4496,
      "mean_token_accuracy": 0.18038235157728194,
      "num_tokens": 12927609.0,
      "step": 5640
    },
    {
      "entropy": 5.500153350830078,
      "epoch": 0.5422670509125841,
      "grad_norm": 1.375,
      "learning_rate": 0.0004977505381625921,
      "loss": 5.3695,
      "mean_token_accuracy": 0.1856964871287346,
      "num_tokens": 12939445.0,
      "step": 5645
    },
    {
      "entropy": 5.358836555480957,
      "epoch": 0.542747358309318,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004977456998616593,
      "loss": 5.3636,
      "mean_token_accuracy": 0.18769825100898743,
      "num_tokens": 12951638.0,
      "step": 5650
    },
    {
      "entropy": 5.491495943069458,
      "epoch": 0.5432276657060519,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004977408563892327,
      "loss": 5.3537,
      "mean_token_accuracy": 0.1883303478360176,
      "num_tokens": 12963157.0,
      "step": 5655
    },
    {
      "entropy": 5.459182405471802,
      "epoch": 0.5437079731027857,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004977360077454249,
      "loss": 5.3232,
      "mean_token_accuracy": 0.1848008930683136,
      "num_tokens": 12974187.0,
      "step": 5660
    },
    {
      "entropy": 5.447499179840088,
      "epoch": 0.5441882804995197,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004977311539303483,
      "loss": 5.4181,
      "mean_token_accuracy": 0.1852130800485611,
      "num_tokens": 12986612.0,
      "step": 5665
    },
    {
      "entropy": 5.509026002883911,
      "epoch": 0.5446685878962536,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004977262949441158,
      "loss": 5.4191,
      "mean_token_accuracy": 0.1818256512284279,
      "num_tokens": 12998343.0,
      "step": 5670
    },
    {
      "entropy": 5.432409858703613,
      "epoch": 0.5451488952929875,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004977214307868399,
      "loss": 5.3914,
      "mean_token_accuracy": 0.18320675790309907,
      "num_tokens": 13009272.0,
      "step": 5675
    },
    {
      "entropy": 5.472721290588379,
      "epoch": 0.5456292026897214,
      "grad_norm": 1.578125,
      "learning_rate": 0.000497716561458634,
      "loss": 5.3744,
      "mean_token_accuracy": 0.19270115345716476,
      "num_tokens": 13019961.0,
      "step": 5680
    },
    {
      "entropy": 5.409308815002442,
      "epoch": 0.5461095100864554,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004977116869596107,
      "loss": 5.3229,
      "mean_token_accuracy": 0.19604558497667313,
      "num_tokens": 13030827.0,
      "step": 5685
    },
    {
      "entropy": 5.545911836624145,
      "epoch": 0.5465898174831892,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004977068072898834,
      "loss": 5.5615,
      "mean_token_accuracy": 0.174258317053318,
      "num_tokens": 13043702.0,
      "step": 5690
    },
    {
      "entropy": 5.371285057067871,
      "epoch": 0.5470701248799231,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004977019224495652,
      "loss": 5.2011,
      "mean_token_accuracy": 0.20260500013828278,
      "num_tokens": 13054474.0,
      "step": 5695
    },
    {
      "entropy": 5.458423662185669,
      "epoch": 0.547550432276657,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004976970324387698,
      "loss": 5.3565,
      "mean_token_accuracy": 0.181874817609787,
      "num_tokens": 13064920.0,
      "step": 5700
    },
    {
      "entropy": 5.489721298217773,
      "epoch": 0.548030739673391,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004976921372576104,
      "loss": 5.4588,
      "mean_token_accuracy": 0.18271468281745912,
      "num_tokens": 13077214.0,
      "step": 5705
    },
    {
      "entropy": 5.496641778945923,
      "epoch": 0.5485110470701249,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004976872369062011,
      "loss": 5.3489,
      "mean_token_accuracy": 0.18365099430084228,
      "num_tokens": 13089472.0,
      "step": 5710
    },
    {
      "entropy": 5.43571252822876,
      "epoch": 0.5489913544668588,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004976823313846552,
      "loss": 5.4294,
      "mean_token_accuracy": 0.19171821922063828,
      "num_tokens": 13100369.0,
      "step": 5715
    },
    {
      "entropy": 5.434876585006714,
      "epoch": 0.5494716618635928,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004976774206930869,
      "loss": 5.3821,
      "mean_token_accuracy": 0.17715881168842315,
      "num_tokens": 13110813.0,
      "step": 5720
    },
    {
      "entropy": 5.465077877044678,
      "epoch": 0.5499519692603266,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004976725048316101,
      "loss": 5.3899,
      "mean_token_accuracy": 0.19080623835325242,
      "num_tokens": 13121946.0,
      "step": 5725
    },
    {
      "entropy": 5.418812561035156,
      "epoch": 0.5504322766570605,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004976675838003388,
      "loss": 5.3317,
      "mean_token_accuracy": 0.19176877588033675,
      "num_tokens": 13133286.0,
      "step": 5730
    },
    {
      "entropy": 5.352372169494629,
      "epoch": 0.5509125840537944,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004976626575993877,
      "loss": 5.3079,
      "mean_token_accuracy": 0.19047755151987075,
      "num_tokens": 13145023.0,
      "step": 5735
    },
    {
      "entropy": 5.429314851760864,
      "epoch": 0.5513928914505284,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004976577262288705,
      "loss": 5.2567,
      "mean_token_accuracy": 0.19412717819213868,
      "num_tokens": 13156379.0,
      "step": 5740
    },
    {
      "entropy": 5.361028480529785,
      "epoch": 0.5518731988472623,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004976527896889023,
      "loss": 5.3022,
      "mean_token_accuracy": 0.19664226770401,
      "num_tokens": 13166313.0,
      "step": 5745
    },
    {
      "entropy": 5.473565244674683,
      "epoch": 0.5523535062439962,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004976478479795974,
      "loss": 5.381,
      "mean_token_accuracy": 0.18206481039524078,
      "num_tokens": 13177968.0,
      "step": 5750
    },
    {
      "entropy": 5.48589334487915,
      "epoch": 0.55283381364073,
      "grad_norm": 1.375,
      "learning_rate": 0.0004976429011010706,
      "loss": 5.3969,
      "mean_token_accuracy": 0.1862631544470787,
      "num_tokens": 13189464.0,
      "step": 5755
    },
    {
      "entropy": 5.4495502471923825,
      "epoch": 0.553314121037464,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004976379490534366,
      "loss": 5.4133,
      "mean_token_accuracy": 0.17927081435918807,
      "num_tokens": 13201716.0,
      "step": 5760
    },
    {
      "entropy": 5.356699466705322,
      "epoch": 0.5537944284341979,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004976329918368107,
      "loss": 5.2335,
      "mean_token_accuracy": 0.18967038840055467,
      "num_tokens": 13212976.0,
      "step": 5765
    },
    {
      "entropy": 5.342360019683838,
      "epoch": 0.5542747358309318,
      "grad_norm": 1.25,
      "learning_rate": 0.0004976280294513079,
      "loss": 5.2808,
      "mean_token_accuracy": 0.20202525407075883,
      "num_tokens": 13225265.0,
      "step": 5770
    },
    {
      "entropy": 5.567855405807495,
      "epoch": 0.5547550432276657,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004976230618970431,
      "loss": 5.5457,
      "mean_token_accuracy": 0.17104607522487641,
      "num_tokens": 13236443.0,
      "step": 5775
    },
    {
      "entropy": 5.447632598876953,
      "epoch": 0.5552353506243997,
      "grad_norm": 1.421875,
      "learning_rate": 0.000497618089174132,
      "loss": 5.4225,
      "mean_token_accuracy": 0.18674527406692504,
      "num_tokens": 13248681.0,
      "step": 5780
    },
    {
      "entropy": 5.45471830368042,
      "epoch": 0.5557156580211335,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004976131112826898,
      "loss": 5.3835,
      "mean_token_accuracy": 0.1897915154695511,
      "num_tokens": 13258719.0,
      "step": 5785
    },
    {
      "entropy": 5.4247579097747805,
      "epoch": 0.5561959654178674,
      "grad_norm": 1.625,
      "learning_rate": 0.0004976081282228323,
      "loss": 5.3666,
      "mean_token_accuracy": 0.18326718360185623,
      "num_tokens": 13270544.0,
      "step": 5790
    },
    {
      "entropy": 5.5005025386810305,
      "epoch": 0.5566762728146013,
      "grad_norm": 1.375,
      "learning_rate": 0.000497603139994675,
      "loss": 5.2919,
      "mean_token_accuracy": 0.19663695842027665,
      "num_tokens": 13280759.0,
      "step": 5795
    },
    {
      "entropy": 5.351595401763916,
      "epoch": 0.5571565802113353,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004975981465983338,
      "loss": 5.4043,
      "mean_token_accuracy": 0.19050797075033188,
      "num_tokens": 13292496.0,
      "step": 5800
    },
    {
      "entropy": 5.462756538391114,
      "epoch": 0.5576368876080692,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004975931480339246,
      "loss": 5.4,
      "mean_token_accuracy": 0.18254634588956833,
      "num_tokens": 13303003.0,
      "step": 5805
    },
    {
      "entropy": 5.450956583023071,
      "epoch": 0.5581171950048031,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004975881443015635,
      "loss": 5.3763,
      "mean_token_accuracy": 0.186819227039814,
      "num_tokens": 13315036.0,
      "step": 5810
    },
    {
      "entropy": 5.450983762741089,
      "epoch": 0.5585975024015369,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004975831354013667,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18532153069972992,
      "num_tokens": 13327529.0,
      "step": 5815
    },
    {
      "entropy": 5.454041481018066,
      "epoch": 0.5590778097982709,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004975781213334503,
      "loss": 5.3351,
      "mean_token_accuracy": 0.18906956613063813,
      "num_tokens": 13339494.0,
      "step": 5820
    },
    {
      "entropy": 5.430515480041504,
      "epoch": 0.5595581171950048,
      "grad_norm": 1.5,
      "learning_rate": 0.0004975731020979309,
      "loss": 5.3511,
      "mean_token_accuracy": 0.18208261877298354,
      "num_tokens": 13351184.0,
      "step": 5825
    },
    {
      "entropy": 5.407172727584839,
      "epoch": 0.5600384245917387,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004975680776949249,
      "loss": 5.2762,
      "mean_token_accuracy": 0.19066329896450043,
      "num_tokens": 13362402.0,
      "step": 5830
    },
    {
      "entropy": 5.398277616500854,
      "epoch": 0.5605187319884726,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004975630481245492,
      "loss": 5.4501,
      "mean_token_accuracy": 0.18694706112146378,
      "num_tokens": 13375257.0,
      "step": 5835
    },
    {
      "entropy": 5.475879573822022,
      "epoch": 0.5609990393852066,
      "grad_norm": 1.5,
      "learning_rate": 0.0004975580133869202,
      "loss": 5.3839,
      "mean_token_accuracy": 0.18546998947858812,
      "num_tokens": 13385917.0,
      "step": 5840
    },
    {
      "entropy": 5.519327878952026,
      "epoch": 0.5614793467819404,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004975529734821552,
      "loss": 5.3978,
      "mean_token_accuracy": 0.18553060442209243,
      "num_tokens": 13397962.0,
      "step": 5845
    },
    {
      "entropy": 5.369125461578369,
      "epoch": 0.5619596541786743,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004975479284103708,
      "loss": 5.267,
      "mean_token_accuracy": 0.1946108713746071,
      "num_tokens": 13408654.0,
      "step": 5850
    },
    {
      "entropy": 5.441431856155395,
      "epoch": 0.5624399615754082,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004975428781716845,
      "loss": 5.3916,
      "mean_token_accuracy": 0.1846861332654953,
      "num_tokens": 13419790.0,
      "step": 5855
    },
    {
      "entropy": 5.440395641326904,
      "epoch": 0.5629202689721422,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004975378227662134,
      "loss": 5.3078,
      "mean_token_accuracy": 0.19199451059103012,
      "num_tokens": 13430305.0,
      "step": 5860
    },
    {
      "entropy": 5.40962929725647,
      "epoch": 0.5634005763688761,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004975327621940746,
      "loss": 5.41,
      "mean_token_accuracy": 0.1890603318810463,
      "num_tokens": 13441730.0,
      "step": 5865
    },
    {
      "entropy": 5.505602598190308,
      "epoch": 0.56388088376561,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004975276964553861,
      "loss": 5.4094,
      "mean_token_accuracy": 0.18663217574357988,
      "num_tokens": 13452003.0,
      "step": 5870
    },
    {
      "entropy": 5.5089545249938965,
      "epoch": 0.5643611911623438,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004975226255502651,
      "loss": 5.2911,
      "mean_token_accuracy": 0.1950973689556122,
      "num_tokens": 13463953.0,
      "step": 5875
    },
    {
      "entropy": 5.448185634613037,
      "epoch": 0.5648414985590778,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004975175494788297,
      "loss": 5.4191,
      "mean_token_accuracy": 0.18593832105398178,
      "num_tokens": 13473820.0,
      "step": 5880
    },
    {
      "entropy": 5.460717630386353,
      "epoch": 0.5653218059558117,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004975124682411974,
      "loss": 5.433,
      "mean_token_accuracy": 0.1860885813832283,
      "num_tokens": 13484181.0,
      "step": 5885
    },
    {
      "entropy": 5.511277198791504,
      "epoch": 0.5658021133525456,
      "grad_norm": 1.5,
      "learning_rate": 0.0004975073818374863,
      "loss": 5.3831,
      "mean_token_accuracy": 0.1844969794154167,
      "num_tokens": 13494480.0,
      "step": 5890
    },
    {
      "entropy": 5.416599559783935,
      "epoch": 0.5662824207492796,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004975022902678145,
      "loss": 5.329,
      "mean_token_accuracy": 0.19147448241710663,
      "num_tokens": 13506026.0,
      "step": 5895
    },
    {
      "entropy": 5.520257091522216,
      "epoch": 0.5667627281460135,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004974971935323003,
      "loss": 5.4194,
      "mean_token_accuracy": 0.182801653444767,
      "num_tokens": 13516849.0,
      "step": 5900
    },
    {
      "entropy": 5.456636953353882,
      "epoch": 0.5672430355427474,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004974920916310619,
      "loss": 5.3037,
      "mean_token_accuracy": 0.19325821250677108,
      "num_tokens": 13527386.0,
      "step": 5905
    },
    {
      "entropy": 5.3854088306427,
      "epoch": 0.5677233429394812,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004974869845642178,
      "loss": 5.4553,
      "mean_token_accuracy": 0.18449880331754684,
      "num_tokens": 13540238.0,
      "step": 5910
    },
    {
      "entropy": 5.428995084762573,
      "epoch": 0.5682036503362152,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004974818723318866,
      "loss": 5.3097,
      "mean_token_accuracy": 0.194813135266304,
      "num_tokens": 13551808.0,
      "step": 5915
    },
    {
      "entropy": 5.3957782745361325,
      "epoch": 0.5686839577329491,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004974767549341868,
      "loss": 5.3476,
      "mean_token_accuracy": 0.1883161708712578,
      "num_tokens": 13563795.0,
      "step": 5920
    },
    {
      "entropy": 5.443729400634766,
      "epoch": 0.569164265129683,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004974716323712376,
      "loss": 5.318,
      "mean_token_accuracy": 0.1895475134253502,
      "num_tokens": 13575168.0,
      "step": 5925
    },
    {
      "entropy": 5.507234525680542,
      "epoch": 0.5696445725264169,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004974665046431576,
      "loss": 5.4302,
      "mean_token_accuracy": 0.18575030863285064,
      "num_tokens": 13587013.0,
      "step": 5930
    },
    {
      "entropy": 5.396302652359009,
      "epoch": 0.5701248799231509,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004974613717500659,
      "loss": 5.2742,
      "mean_token_accuracy": 0.18803299069404603,
      "num_tokens": 13598138.0,
      "step": 5935
    },
    {
      "entropy": 5.4670515060424805,
      "epoch": 0.5706051873198847,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004974562336920818,
      "loss": 5.4013,
      "mean_token_accuracy": 0.18367205709218978,
      "num_tokens": 13609174.0,
      "step": 5940
    },
    {
      "entropy": 5.428857803344727,
      "epoch": 0.5710854947166186,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004974510904693245,
      "loss": 5.298,
      "mean_token_accuracy": 0.1903289094567299,
      "num_tokens": 13619724.0,
      "step": 5945
    },
    {
      "entropy": 5.399194240570068,
      "epoch": 0.5715658021133525,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004974459420819134,
      "loss": 5.3471,
      "mean_token_accuracy": 0.18765629678964615,
      "num_tokens": 13631114.0,
      "step": 5950
    },
    {
      "entropy": 5.362923431396484,
      "epoch": 0.5720461095100865,
      "grad_norm": 1.765625,
      "learning_rate": 0.000497440788529968,
      "loss": 5.2797,
      "mean_token_accuracy": 0.19567984342575073,
      "num_tokens": 13643035.0,
      "step": 5955
    },
    {
      "entropy": 5.476286602020264,
      "epoch": 0.5725264169068204,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004974356298136081,
      "loss": 5.3611,
      "mean_token_accuracy": 0.18529228121042252,
      "num_tokens": 13656603.0,
      "step": 5960
    },
    {
      "entropy": 5.387987184524536,
      "epoch": 0.5730067243035543,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004974304659329533,
      "loss": 5.3815,
      "mean_token_accuracy": 0.18811299204826354,
      "num_tokens": 13668420.0,
      "step": 5965
    },
    {
      "entropy": 5.471386289596557,
      "epoch": 0.5734870317002881,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004974252968881236,
      "loss": 5.4005,
      "mean_token_accuracy": 0.18334517478942872,
      "num_tokens": 13680521.0,
      "step": 5970
    },
    {
      "entropy": 5.407588338851928,
      "epoch": 0.5739673390970221,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000497420122679239,
      "loss": 5.2836,
      "mean_token_accuracy": 0.19422682821750642,
      "num_tokens": 13692503.0,
      "step": 5975
    },
    {
      "entropy": 5.319512367248535,
      "epoch": 0.574447646493756,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004974149433064196,
      "loss": 5.1647,
      "mean_token_accuracy": 0.2031061515212059,
      "num_tokens": 13703318.0,
      "step": 5980
    },
    {
      "entropy": 5.403626155853272,
      "epoch": 0.5749279538904899,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004974097587697856,
      "loss": 5.3535,
      "mean_token_accuracy": 0.18990399837493896,
      "num_tokens": 13713957.0,
      "step": 5985
    },
    {
      "entropy": 5.453413772583008,
      "epoch": 0.5754082612872238,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004974045690694575,
      "loss": 5.4536,
      "mean_token_accuracy": 0.18412325978279115,
      "num_tokens": 13725300.0,
      "step": 5990
    },
    {
      "entropy": 5.494600439071656,
      "epoch": 0.5758885686839578,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004973993742055557,
      "loss": 5.4065,
      "mean_token_accuracy": 0.18396205753087996,
      "num_tokens": 13735463.0,
      "step": 5995
    },
    {
      "entropy": 5.513051128387451,
      "epoch": 0.5763688760806917,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004973941741782007,
      "loss": 5.4965,
      "mean_token_accuracy": 0.1788262516260147,
      "num_tokens": 13749062.0,
      "step": 6000
    },
    {
      "epoch": 0.5763688760806917,
      "eval_entropy": 5.391382442573871,
      "eval_loss": 5.3939995765686035,
      "eval_mean_token_accuracy": 0.1923779231173303,
      "eval_num_tokens": 13749062.0,
      "eval_runtime": 26.6581,
      "eval_samples_per_second": 1230.958,
      "eval_steps_per_second": 153.874,
      "step": 6000
    },
    {
      "entropy": 5.594244194030762,
      "epoch": 0.5768491834774255,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004973889689875135,
      "loss": 5.4567,
      "mean_token_accuracy": 0.17287490218877793,
      "num_tokens": 13761766.0,
      "step": 6005
    },
    {
      "entropy": 5.467732667922974,
      "epoch": 0.5773294908741594,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004973837586336147,
      "loss": 5.4181,
      "mean_token_accuracy": 0.1873711794614792,
      "num_tokens": 13774702.0,
      "step": 6010
    },
    {
      "entropy": 5.388287687301636,
      "epoch": 0.5778097982708934,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004973785431166254,
      "loss": 5.3238,
      "mean_token_accuracy": 0.1841629922389984,
      "num_tokens": 13786215.0,
      "step": 6015
    },
    {
      "entropy": 5.379306221008301,
      "epoch": 0.5782901056676273,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004973733224366666,
      "loss": 5.3517,
      "mean_token_accuracy": 0.18715571910142897,
      "num_tokens": 13797697.0,
      "step": 6020
    },
    {
      "entropy": 5.505963897705078,
      "epoch": 0.5787704130643612,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004973680965938597,
      "loss": 5.4259,
      "mean_token_accuracy": 0.18360659927129747,
      "num_tokens": 13810085.0,
      "step": 6025
    },
    {
      "entropy": 5.485981369018555,
      "epoch": 0.579250720461095,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004973628655883258,
      "loss": 5.3127,
      "mean_token_accuracy": 0.1925201028585434,
      "num_tokens": 13822054.0,
      "step": 6030
    },
    {
      "entropy": 5.295996761322021,
      "epoch": 0.579731027857829,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004973576294201865,
      "loss": 5.292,
      "mean_token_accuracy": 0.19221212714910507,
      "num_tokens": 13834342.0,
      "step": 6035
    },
    {
      "entropy": 5.438166999816895,
      "epoch": 0.5802113352545629,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004973523880895633,
      "loss": 5.3843,
      "mean_token_accuracy": 0.1854028046131134,
      "num_tokens": 13846598.0,
      "step": 6040
    },
    {
      "entropy": 5.444848537445068,
      "epoch": 0.5806916426512968,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004973471415965779,
      "loss": 5.3138,
      "mean_token_accuracy": 0.188867250084877,
      "num_tokens": 13857515.0,
      "step": 6045
    },
    {
      "entropy": 5.325572919845581,
      "epoch": 0.5811719500480308,
      "grad_norm": 1.265625,
      "learning_rate": 0.000497341889941352,
      "loss": 5.2379,
      "mean_token_accuracy": 0.1973300725221634,
      "num_tokens": 13868333.0,
      "step": 6050
    },
    {
      "entropy": 5.4464198589324955,
      "epoch": 0.5816522574447647,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004973366331240078,
      "loss": 5.4,
      "mean_token_accuracy": 0.1817735403776169,
      "num_tokens": 13879048.0,
      "step": 6055
    },
    {
      "entropy": 5.47771372795105,
      "epoch": 0.5821325648414986,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000497331371144667,
      "loss": 5.4594,
      "mean_token_accuracy": 0.17890119254589082,
      "num_tokens": 13891358.0,
      "step": 6060
    },
    {
      "entropy": 5.602402830123902,
      "epoch": 0.5826128722382324,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004973261040034521,
      "loss": 5.4929,
      "mean_token_accuracy": 0.17942275404930114,
      "num_tokens": 13902680.0,
      "step": 6065
    },
    {
      "entropy": 5.513533735275269,
      "epoch": 0.5830931796349664,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004973208317004852,
      "loss": 5.3803,
      "mean_token_accuracy": 0.18414207845926284,
      "num_tokens": 13913863.0,
      "step": 6070
    },
    {
      "entropy": 5.387319421768188,
      "epoch": 0.5835734870317003,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004973155542358889,
      "loss": 5.3468,
      "mean_token_accuracy": 0.189805369079113,
      "num_tokens": 13924917.0,
      "step": 6075
    },
    {
      "entropy": 5.282955503463745,
      "epoch": 0.5840537944284342,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004973102716097853,
      "loss": 5.1527,
      "mean_token_accuracy": 0.19826523214578629,
      "num_tokens": 13936022.0,
      "step": 6080
    },
    {
      "entropy": 5.456071472167968,
      "epoch": 0.5845341018251681,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004973049838222973,
      "loss": 5.3778,
      "mean_token_accuracy": 0.18554320186376572,
      "num_tokens": 13947594.0,
      "step": 6085
    },
    {
      "entropy": 5.481543970108032,
      "epoch": 0.5850144092219021,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004972996908735479,
      "loss": 5.3637,
      "mean_token_accuracy": 0.18329488188028337,
      "num_tokens": 13960698.0,
      "step": 6090
    },
    {
      "entropy": 5.424262666702271,
      "epoch": 0.585494716618636,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004972943927636597,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18691962212324142,
      "num_tokens": 13971004.0,
      "step": 6095
    },
    {
      "entropy": 5.402873086929321,
      "epoch": 0.5859750240153698,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004972890894927558,
      "loss": 5.36,
      "mean_token_accuracy": 0.1917417049407959,
      "num_tokens": 13982500.0,
      "step": 6100
    },
    {
      "entropy": 5.524716901779175,
      "epoch": 0.5864553314121037,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004972837810609592,
      "loss": 5.5061,
      "mean_token_accuracy": 0.18116320669651031,
      "num_tokens": 13994129.0,
      "step": 6105
    },
    {
      "entropy": 5.4113037109375,
      "epoch": 0.5869356388088377,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004972784674683933,
      "loss": 5.2851,
      "mean_token_accuracy": 0.19135573953390123,
      "num_tokens": 14005581.0,
      "step": 6110
    },
    {
      "entropy": 5.425225019454956,
      "epoch": 0.5874159462055716,
      "grad_norm": 1.625,
      "learning_rate": 0.0004972731487151815,
      "loss": 5.3684,
      "mean_token_accuracy": 0.18579945117235183,
      "num_tokens": 14015866.0,
      "step": 6115
    },
    {
      "entropy": 5.383860778808594,
      "epoch": 0.5878962536023055,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004972678248014471,
      "loss": 5.358,
      "mean_token_accuracy": 0.1943156510591507,
      "num_tokens": 14027002.0,
      "step": 6120
    },
    {
      "entropy": 5.3782281398773195,
      "epoch": 0.5883765609990393,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004972624957273139,
      "loss": 5.3718,
      "mean_token_accuracy": 0.18841624408960342,
      "num_tokens": 14038962.0,
      "step": 6125
    },
    {
      "entropy": 5.466955089569092,
      "epoch": 0.5888568683957733,
      "grad_norm": 1.375,
      "learning_rate": 0.0004972571614929055,
      "loss": 5.3941,
      "mean_token_accuracy": 0.1893245682120323,
      "num_tokens": 14049956.0,
      "step": 6130
    },
    {
      "entropy": 5.4190325260162355,
      "epoch": 0.5893371757925072,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004972518220983457,
      "loss": 5.3931,
      "mean_token_accuracy": 0.1856802999973297,
      "num_tokens": 14062413.0,
      "step": 6135
    },
    {
      "entropy": 5.473791646957397,
      "epoch": 0.5898174831892411,
      "grad_norm": 1.625,
      "learning_rate": 0.0004972464775437586,
      "loss": 5.3723,
      "mean_token_accuracy": 0.1883968785405159,
      "num_tokens": 14075166.0,
      "step": 6140
    },
    {
      "entropy": 5.400057363510132,
      "epoch": 0.590297790585975,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004972411278292683,
      "loss": 5.2924,
      "mean_token_accuracy": 0.19134089350700378,
      "num_tokens": 14085208.0,
      "step": 6145
    },
    {
      "entropy": 5.528112077713013,
      "epoch": 0.590778097982709,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004972357729549988,
      "loss": 5.4578,
      "mean_token_accuracy": 0.17733772099018097,
      "num_tokens": 14096688.0,
      "step": 6150
    },
    {
      "entropy": 5.303466796875,
      "epoch": 0.5912584053794429,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004972304129210746,
      "loss": 5.2691,
      "mean_token_accuracy": 0.1955884262919426,
      "num_tokens": 14108737.0,
      "step": 6155
    },
    {
      "entropy": 5.388396978378296,
      "epoch": 0.5917387127761767,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004972250477276202,
      "loss": 5.3312,
      "mean_token_accuracy": 0.19089388847351074,
      "num_tokens": 14120121.0,
      "step": 6160
    },
    {
      "entropy": 5.414858722686768,
      "epoch": 0.5922190201729106,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004972196773747599,
      "loss": 5.3894,
      "mean_token_accuracy": 0.1914263740181923,
      "num_tokens": 14132157.0,
      "step": 6165
    },
    {
      "entropy": 5.421557950973511,
      "epoch": 0.5926993275696446,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004972143018626186,
      "loss": 5.2992,
      "mean_token_accuracy": 0.18856575340032578,
      "num_tokens": 14143092.0,
      "step": 6170
    },
    {
      "entropy": 5.392038869857788,
      "epoch": 0.5931796349663785,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004972089211913211,
      "loss": 5.3087,
      "mean_token_accuracy": 0.18959310948848723,
      "num_tokens": 14153973.0,
      "step": 6175
    },
    {
      "entropy": 5.4936912059783936,
      "epoch": 0.5936599423631124,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004972035353609923,
      "loss": 5.3806,
      "mean_token_accuracy": 0.17793257385492325,
      "num_tokens": 14164919.0,
      "step": 6180
    },
    {
      "entropy": 5.423027610778808,
      "epoch": 0.5941402497598463,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004971981443717572,
      "loss": 5.3794,
      "mean_token_accuracy": 0.19024920612573623,
      "num_tokens": 14176803.0,
      "step": 6185
    },
    {
      "entropy": 5.44634518623352,
      "epoch": 0.5946205571565802,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004971927482237409,
      "loss": 5.2971,
      "mean_token_accuracy": 0.18827152550220488,
      "num_tokens": 14190087.0,
      "step": 6190
    },
    {
      "entropy": 5.450093936920166,
      "epoch": 0.5951008645533141,
      "grad_norm": 1.375,
      "learning_rate": 0.0004971873469170689,
      "loss": 5.5168,
      "mean_token_accuracy": 0.17948530316352845,
      "num_tokens": 14202242.0,
      "step": 6195
    },
    {
      "entropy": 5.474076890945435,
      "epoch": 0.595581171950048,
      "grad_norm": 1.375,
      "learning_rate": 0.0004971819404518664,
      "loss": 5.4311,
      "mean_token_accuracy": 0.18267126381397247,
      "num_tokens": 14214020.0,
      "step": 6200
    },
    {
      "entropy": 5.4121746063232425,
      "epoch": 0.5960614793467819,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000497176528828259,
      "loss": 5.3131,
      "mean_token_accuracy": 0.19289236813783645,
      "num_tokens": 14225455.0,
      "step": 6205
    },
    {
      "entropy": 5.338340520858765,
      "epoch": 0.5965417867435159,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004971711120463722,
      "loss": 5.2849,
      "mean_token_accuracy": 0.18932408541440965,
      "num_tokens": 14235955.0,
      "step": 6210
    },
    {
      "entropy": 5.38538179397583,
      "epoch": 0.5970220941402498,
      "grad_norm": 1.421875,
      "learning_rate": 0.000497165690106332,
      "loss": 5.3668,
      "mean_token_accuracy": 0.182011616230011,
      "num_tokens": 14246107.0,
      "step": 6215
    },
    {
      "entropy": 5.437984466552734,
      "epoch": 0.5975024015369836,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004971602630082642,
      "loss": 5.3354,
      "mean_token_accuracy": 0.18752168864011765,
      "num_tokens": 14257042.0,
      "step": 6220
    },
    {
      "entropy": 5.5255238056182865,
      "epoch": 0.5979827089337176,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004971548307522947,
      "loss": 5.5186,
      "mean_token_accuracy": 0.181718672811985,
      "num_tokens": 14268670.0,
      "step": 6225
    },
    {
      "entropy": 5.423706769943237,
      "epoch": 0.5984630163304515,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004971493933385498,
      "loss": 5.2774,
      "mean_token_accuracy": 0.19225098341703414,
      "num_tokens": 14279538.0,
      "step": 6230
    },
    {
      "entropy": 5.444534683227539,
      "epoch": 0.5989433237271854,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004971439507671556,
      "loss": 5.3827,
      "mean_token_accuracy": 0.18014575242996217,
      "num_tokens": 14291928.0,
      "step": 6235
    },
    {
      "entropy": 5.425173711776734,
      "epoch": 0.5994236311239193,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004971385030382384,
      "loss": 5.3101,
      "mean_token_accuracy": 0.19303133934736252,
      "num_tokens": 14302958.0,
      "step": 6240
    },
    {
      "entropy": 5.406954002380371,
      "epoch": 0.5999039385206533,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004971330501519248,
      "loss": 5.3559,
      "mean_token_accuracy": 0.18892101049423218,
      "num_tokens": 14315585.0,
      "step": 6245
    },
    {
      "entropy": 5.401479005813599,
      "epoch": 0.6003842459173871,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004971275921083414,
      "loss": 5.433,
      "mean_token_accuracy": 0.18227880150079728,
      "num_tokens": 14327228.0,
      "step": 6250
    },
    {
      "entropy": 5.523636722564698,
      "epoch": 0.600864553314121,
      "grad_norm": 1.5,
      "learning_rate": 0.000497122128907615,
      "loss": 5.4333,
      "mean_token_accuracy": 0.1836875319480896,
      "num_tokens": 14338079.0,
      "step": 6255
    },
    {
      "entropy": 5.4651741027832035,
      "epoch": 0.6013448607108549,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004971166605498722,
      "loss": 5.3075,
      "mean_token_accuracy": 0.19537494331598282,
      "num_tokens": 14350220.0,
      "step": 6260
    },
    {
      "entropy": 5.468128061294555,
      "epoch": 0.6018251681075889,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004971111870352401,
      "loss": 5.3995,
      "mean_token_accuracy": 0.1816209778189659,
      "num_tokens": 14360761.0,
      "step": 6265
    },
    {
      "entropy": 5.448753786087036,
      "epoch": 0.6023054755043228,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004971057083638458,
      "loss": 5.4001,
      "mean_token_accuracy": 0.1839236721396446,
      "num_tokens": 14372056.0,
      "step": 6270
    },
    {
      "entropy": 5.515234518051147,
      "epoch": 0.6027857829010567,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004971002245358166,
      "loss": 5.3751,
      "mean_token_accuracy": 0.18750105053186417,
      "num_tokens": 14382601.0,
      "step": 6275
    },
    {
      "entropy": 5.590676403045654,
      "epoch": 0.6032660902977905,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004970947355512795,
      "loss": 5.4816,
      "mean_token_accuracy": 0.17895192205905913,
      "num_tokens": 14393749.0,
      "step": 6280
    },
    {
      "entropy": 5.477749538421631,
      "epoch": 0.6037463976945245,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004970892414103622,
      "loss": 5.3723,
      "mean_token_accuracy": 0.18472715020179747,
      "num_tokens": 14405001.0,
      "step": 6285
    },
    {
      "entropy": 5.281026363372803,
      "epoch": 0.6042267050912584,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004970837421131921,
      "loss": 5.2256,
      "mean_token_accuracy": 0.1901577666401863,
      "num_tokens": 14415875.0,
      "step": 6290
    },
    {
      "entropy": 5.327046680450439,
      "epoch": 0.6047070124879923,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004970782376598972,
      "loss": 5.2818,
      "mean_token_accuracy": 0.18864146918058394,
      "num_tokens": 14426896.0,
      "step": 6295
    },
    {
      "entropy": 5.454883766174317,
      "epoch": 0.6051873198847262,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004970727280506048,
      "loss": 5.3631,
      "mean_token_accuracy": 0.1811344027519226,
      "num_tokens": 14438712.0,
      "step": 6300
    },
    {
      "entropy": 5.4120667457580565,
      "epoch": 0.6056676272814602,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004970672132854431,
      "loss": 5.3062,
      "mean_token_accuracy": 0.1871188297867775,
      "num_tokens": 14450952.0,
      "step": 6305
    },
    {
      "entropy": 5.4106958389282225,
      "epoch": 0.6061479346781941,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004970616933645403,
      "loss": 5.3385,
      "mean_token_accuracy": 0.19248796105384827,
      "num_tokens": 14462139.0,
      "step": 6310
    },
    {
      "entropy": 5.395157098770142,
      "epoch": 0.6066282420749279,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004970561682880242,
      "loss": 5.4081,
      "mean_token_accuracy": 0.18527979850769044,
      "num_tokens": 14472817.0,
      "step": 6315
    },
    {
      "entropy": 5.464319038391113,
      "epoch": 0.6071085494716618,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004970506380560233,
      "loss": 5.4341,
      "mean_token_accuracy": 0.17815070897340773,
      "num_tokens": 14484241.0,
      "step": 6320
    },
    {
      "entropy": 5.376730060577392,
      "epoch": 0.6075888568683958,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004970451026686659,
      "loss": 5.3067,
      "mean_token_accuracy": 0.18941431492567062,
      "num_tokens": 14496327.0,
      "step": 6325
    },
    {
      "entropy": 5.460070943832397,
      "epoch": 0.6080691642651297,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004970395621260806,
      "loss": 5.3413,
      "mean_token_accuracy": 0.18361499011516572,
      "num_tokens": 14508041.0,
      "step": 6330
    },
    {
      "entropy": 5.431254482269287,
      "epoch": 0.6085494716618636,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000497034016428396,
      "loss": 5.3344,
      "mean_token_accuracy": 0.1927487924695015,
      "num_tokens": 14519023.0,
      "step": 6335
    },
    {
      "entropy": 5.3656209945678714,
      "epoch": 0.6090297790585975,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004970284655757409,
      "loss": 5.3421,
      "mean_token_accuracy": 0.18459006696939467,
      "num_tokens": 14530783.0,
      "step": 6340
    },
    {
      "entropy": 5.400503015518188,
      "epoch": 0.6095100864553314,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004970229095682439,
      "loss": 5.2979,
      "mean_token_accuracy": 0.19315001517534255,
      "num_tokens": 14541680.0,
      "step": 6345
    },
    {
      "entropy": 5.451666212081909,
      "epoch": 0.6099903938520653,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004970173484060344,
      "loss": 5.3521,
      "mean_token_accuracy": 0.18292970210313797,
      "num_tokens": 14553792.0,
      "step": 6350
    },
    {
      "entropy": 5.436286497116089,
      "epoch": 0.6104707012487992,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004970117820892414,
      "loss": 5.3981,
      "mean_token_accuracy": 0.18426571935415267,
      "num_tokens": 14565410.0,
      "step": 6355
    },
    {
      "entropy": 5.3938507556915285,
      "epoch": 0.6109510086455331,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004970062106179939,
      "loss": 5.2429,
      "mean_token_accuracy": 0.1873038202524185,
      "num_tokens": 14576722.0,
      "step": 6360
    },
    {
      "entropy": 5.364889001846313,
      "epoch": 0.6114313160422671,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004970006339924214,
      "loss": 5.2763,
      "mean_token_accuracy": 0.19065423011779786,
      "num_tokens": 14589443.0,
      "step": 6365
    },
    {
      "entropy": 5.4637962818145756,
      "epoch": 0.611911623439001,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004969950522126534,
      "loss": 5.428,
      "mean_token_accuracy": 0.1833744838833809,
      "num_tokens": 14600731.0,
      "step": 6370
    },
    {
      "entropy": 5.415026617050171,
      "epoch": 0.6123919308357348,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004969894652788196,
      "loss": 5.2769,
      "mean_token_accuracy": 0.1923918157815933,
      "num_tokens": 14611152.0,
      "step": 6375
    },
    {
      "entropy": 5.44018177986145,
      "epoch": 0.6128722382324687,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004969838731910494,
      "loss": 5.4181,
      "mean_token_accuracy": 0.17922666668891907,
      "num_tokens": 14624526.0,
      "step": 6380
    },
    {
      "entropy": 5.415634250640869,
      "epoch": 0.6133525456292027,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004969782759494729,
      "loss": 5.3432,
      "mean_token_accuracy": 0.19230301380157472,
      "num_tokens": 14635527.0,
      "step": 6385
    },
    {
      "entropy": 5.461290121078491,
      "epoch": 0.6138328530259366,
      "grad_norm": 1.34375,
      "learning_rate": 0.00049697267355422,
      "loss": 5.394,
      "mean_token_accuracy": 0.18506784588098527,
      "num_tokens": 14647372.0,
      "step": 6390
    },
    {
      "entropy": 5.5496093273162845,
      "epoch": 0.6143131604226705,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004969670660054208,
      "loss": 5.4065,
      "mean_token_accuracy": 0.18408179730176927,
      "num_tokens": 14659216.0,
      "step": 6395
    },
    {
      "entropy": 5.427420568466187,
      "epoch": 0.6147934678194045,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004969614533032054,
      "loss": 5.3529,
      "mean_token_accuracy": 0.18339368849992752,
      "num_tokens": 14670336.0,
      "step": 6400
    },
    {
      "entropy": 5.31213116645813,
      "epoch": 0.6152737752161384,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004969558354477041,
      "loss": 5.224,
      "mean_token_accuracy": 0.1918257251381874,
      "num_tokens": 14680668.0,
      "step": 6405
    },
    {
      "entropy": 5.461543703079224,
      "epoch": 0.6157540826128722,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004969502124390474,
      "loss": 5.435,
      "mean_token_accuracy": 0.1822928488254547,
      "num_tokens": 14691521.0,
      "step": 6410
    },
    {
      "entropy": 5.446273708343506,
      "epoch": 0.6162343900096061,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004969445842773658,
      "loss": 5.3053,
      "mean_token_accuracy": 0.18367120921611785,
      "num_tokens": 14702473.0,
      "step": 6415
    },
    {
      "entropy": 5.345534086227417,
      "epoch": 0.6167146974063401,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00049693895096279,
      "loss": 5.2602,
      "mean_token_accuracy": 0.1895003318786621,
      "num_tokens": 14713057.0,
      "step": 6420
    },
    {
      "entropy": 5.477782440185547,
      "epoch": 0.617195004803074,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004969333124954508,
      "loss": 5.3754,
      "mean_token_accuracy": 0.18387351334095,
      "num_tokens": 14725203.0,
      "step": 6425
    },
    {
      "entropy": 5.376272296905517,
      "epoch": 0.6176753121998079,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004969276688754791,
      "loss": 5.3415,
      "mean_token_accuracy": 0.19398964643478395,
      "num_tokens": 14735851.0,
      "step": 6430
    },
    {
      "entropy": 5.472036266326905,
      "epoch": 0.6181556195965417,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004969220201030059,
      "loss": 5.4299,
      "mean_token_accuracy": 0.19013222604990004,
      "num_tokens": 14747876.0,
      "step": 6435
    },
    {
      "entropy": 5.41585693359375,
      "epoch": 0.6186359269932757,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004969163661781624,
      "loss": 5.3456,
      "mean_token_accuracy": 0.1899268090724945,
      "num_tokens": 14758673.0,
      "step": 6440
    },
    {
      "entropy": 5.407340240478516,
      "epoch": 0.6191162343900096,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004969107071010798,
      "loss": 5.2987,
      "mean_token_accuracy": 0.19403767287731172,
      "num_tokens": 14770505.0,
      "step": 6445
    },
    {
      "entropy": 5.421085453033447,
      "epoch": 0.6195965417867435,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004969050428718895,
      "loss": 5.3537,
      "mean_token_accuracy": 0.18687873035669328,
      "num_tokens": 14782518.0,
      "step": 6450
    },
    {
      "entropy": 5.392564964294434,
      "epoch": 0.6200768491834774,
      "grad_norm": 2.046875,
      "learning_rate": 0.000496899373490723,
      "loss": 5.296,
      "mean_token_accuracy": 0.19183335602283477,
      "num_tokens": 14794608.0,
      "step": 6455
    },
    {
      "entropy": 5.328503513336182,
      "epoch": 0.6205571565802114,
      "grad_norm": 1.390625,
      "learning_rate": 0.000496893698957712,
      "loss": 5.2157,
      "mean_token_accuracy": 0.19725629240274428,
      "num_tokens": 14804725.0,
      "step": 6460
    },
    {
      "entropy": 5.447431182861328,
      "epoch": 0.6210374639769453,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004968880192729882,
      "loss": 5.3543,
      "mean_token_accuracy": 0.1898462802171707,
      "num_tokens": 14816178.0,
      "step": 6465
    },
    {
      "entropy": 5.457472562789917,
      "epoch": 0.6215177713736791,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004968823344366835,
      "loss": 5.3114,
      "mean_token_accuracy": 0.18324750810861587,
      "num_tokens": 14827808.0,
      "step": 6470
    },
    {
      "entropy": 5.3914533138275145,
      "epoch": 0.621998078770413,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004968766444489298,
      "loss": 5.2849,
      "mean_token_accuracy": 0.1886556029319763,
      "num_tokens": 14839423.0,
      "step": 6475
    },
    {
      "entropy": 5.474667930603028,
      "epoch": 0.622478386167147,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004968709493098593,
      "loss": 5.4007,
      "mean_token_accuracy": 0.18815280497074127,
      "num_tokens": 14849678.0,
      "step": 6480
    },
    {
      "entropy": 5.340357923507691,
      "epoch": 0.6229586935638809,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004968652490196041,
      "loss": 5.2926,
      "mean_token_accuracy": 0.1961539551615715,
      "num_tokens": 14860912.0,
      "step": 6485
    },
    {
      "entropy": 5.359215307235718,
      "epoch": 0.6234390009606148,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004968595435782967,
      "loss": 5.314,
      "mean_token_accuracy": 0.18657833635807036,
      "num_tokens": 14873028.0,
      "step": 6490
    },
    {
      "entropy": 5.49301381111145,
      "epoch": 0.6239193083573487,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004968538329860695,
      "loss": 5.2977,
      "mean_token_accuracy": 0.18742749840021133,
      "num_tokens": 14883985.0,
      "step": 6495
    },
    {
      "entropy": 5.461597871780396,
      "epoch": 0.6243996157540826,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004968481172430549,
      "loss": 5.368,
      "mean_token_accuracy": 0.18283463567495345,
      "num_tokens": 14896569.0,
      "step": 6500
    },
    {
      "entropy": 5.43288402557373,
      "epoch": 0.6248799231508165,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000496842396349386,
      "loss": 5.3619,
      "mean_token_accuracy": 0.19527961015701295,
      "num_tokens": 14907874.0,
      "step": 6505
    },
    {
      "entropy": 5.420658826828003,
      "epoch": 0.6253602305475504,
      "grad_norm": 1.375,
      "learning_rate": 0.0004968366703051952,
      "loss": 5.3322,
      "mean_token_accuracy": 0.19150687158107757,
      "num_tokens": 14919501.0,
      "step": 6510
    },
    {
      "entropy": 5.408752489089966,
      "epoch": 0.6258405379442843,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004968309391106157,
      "loss": 5.3187,
      "mean_token_accuracy": 0.18907625526189803,
      "num_tokens": 14929876.0,
      "step": 6515
    },
    {
      "entropy": 5.391577672958374,
      "epoch": 0.6263208453410183,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004968252027657806,
      "loss": 5.398,
      "mean_token_accuracy": 0.18368214070796968,
      "num_tokens": 14941317.0,
      "step": 6520
    },
    {
      "entropy": 5.480224609375,
      "epoch": 0.6268011527377522,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004968194612708229,
      "loss": 5.3747,
      "mean_token_accuracy": 0.18525346368551254,
      "num_tokens": 14953038.0,
      "step": 6525
    },
    {
      "entropy": 5.379436111450195,
      "epoch": 0.627281460134486,
      "grad_norm": 3.75,
      "learning_rate": 0.0004968137146258759,
      "loss": 5.2976,
      "mean_token_accuracy": 0.19222625941038132,
      "num_tokens": 14964664.0,
      "step": 6530
    },
    {
      "entropy": 5.37131519317627,
      "epoch": 0.6277617675312199,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004968079628310732,
      "loss": 5.3014,
      "mean_token_accuracy": 0.19383252710103988,
      "num_tokens": 14975761.0,
      "step": 6535
    },
    {
      "entropy": 5.4234987735748295,
      "epoch": 0.6282420749279539,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004968022058865482,
      "loss": 5.2335,
      "mean_token_accuracy": 0.1946899637579918,
      "num_tokens": 14987670.0,
      "step": 6540
    },
    {
      "entropy": 5.346607255935669,
      "epoch": 0.6287223823246878,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004967964437924346,
      "loss": 5.1958,
      "mean_token_accuracy": 0.1991036906838417,
      "num_tokens": 14999390.0,
      "step": 6545
    },
    {
      "entropy": 5.344186592102051,
      "epoch": 0.6292026897214217,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004967906765488662,
      "loss": 5.3289,
      "mean_token_accuracy": 0.18837945014238358,
      "num_tokens": 15011691.0,
      "step": 6550
    },
    {
      "entropy": 5.536790609359741,
      "epoch": 0.6296829971181557,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004967849041559769,
      "loss": 5.4462,
      "mean_token_accuracy": 0.18327112942934037,
      "num_tokens": 15023539.0,
      "step": 6555
    },
    {
      "entropy": 5.4762969493865965,
      "epoch": 0.6301633045148896,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004967791266139006,
      "loss": 5.4066,
      "mean_token_accuracy": 0.19257629811763763,
      "num_tokens": 15035529.0,
      "step": 6560
    },
    {
      "entropy": 5.422998237609863,
      "epoch": 0.6306436119116234,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004967733439227716,
      "loss": 5.3563,
      "mean_token_accuracy": 0.1900174215435982,
      "num_tokens": 15047072.0,
      "step": 6565
    },
    {
      "entropy": 5.46677794456482,
      "epoch": 0.6311239193083573,
      "grad_norm": 2.28125,
      "learning_rate": 0.000496767556082724,
      "loss": 5.4026,
      "mean_token_accuracy": 0.18688941448926927,
      "num_tokens": 15058951.0,
      "step": 6570
    },
    {
      "entropy": 5.417145586013794,
      "epoch": 0.6316042267050913,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004967617630938924,
      "loss": 5.3879,
      "mean_token_accuracy": 0.19218687117099761,
      "num_tokens": 15070647.0,
      "step": 6575
    },
    {
      "entropy": 5.436992740631103,
      "epoch": 0.6320845341018252,
      "grad_norm": 1.40625,
      "learning_rate": 0.000496755964956411,
      "loss": 5.308,
      "mean_token_accuracy": 0.19497922956943511,
      "num_tokens": 15082324.0,
      "step": 6580
    },
    {
      "entropy": 5.488131141662597,
      "epoch": 0.6325648414985591,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004967501616704147,
      "loss": 5.4626,
      "mean_token_accuracy": 0.17747684121131896,
      "num_tokens": 15094641.0,
      "step": 6585
    },
    {
      "entropy": 5.557354927062988,
      "epoch": 0.633045148895293,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000496744353236038,
      "loss": 5.3726,
      "mean_token_accuracy": 0.18662358373403548,
      "num_tokens": 15106610.0,
      "step": 6590
    },
    {
      "entropy": 5.3808118343353275,
      "epoch": 0.633525456292027,
      "grad_norm": 1.453125,
      "learning_rate": 0.000496738539653416,
      "loss": 5.3349,
      "mean_token_accuracy": 0.1909119114279747,
      "num_tokens": 15118463.0,
      "step": 6595
    },
    {
      "entropy": 5.397026205062867,
      "epoch": 0.6340057636887608,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004967327209226835,
      "loss": 5.3311,
      "mean_token_accuracy": 0.19154786914587021,
      "num_tokens": 15130154.0,
      "step": 6600
    },
    {
      "entropy": 5.37779278755188,
      "epoch": 0.6344860710854947,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004967268970439755,
      "loss": 5.2821,
      "mean_token_accuracy": 0.19595194458961487,
      "num_tokens": 15140854.0,
      "step": 6605
    },
    {
      "entropy": 5.381403541564941,
      "epoch": 0.6349663784822286,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004967210680174274,
      "loss": 5.3239,
      "mean_token_accuracy": 0.19121498614549637,
      "num_tokens": 15151075.0,
      "step": 6610
    },
    {
      "entropy": 5.356026649475098,
      "epoch": 0.6354466858789626,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004967152338431745,
      "loss": 5.2727,
      "mean_token_accuracy": 0.1972847431898117,
      "num_tokens": 15162091.0,
      "step": 6615
    },
    {
      "entropy": 5.542326402664185,
      "epoch": 0.6359269932756965,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004967093945213523,
      "loss": 5.4259,
      "mean_token_accuracy": 0.18341906666755675,
      "num_tokens": 15175321.0,
      "step": 6620
    },
    {
      "entropy": 5.391190147399902,
      "epoch": 0.6364073006724303,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004967035500520962,
      "loss": 5.345,
      "mean_token_accuracy": 0.19236933290958405,
      "num_tokens": 15187058.0,
      "step": 6625
    },
    {
      "entropy": 5.4209099292755125,
      "epoch": 0.6368876080691642,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004966977004355421,
      "loss": 5.3279,
      "mean_token_accuracy": 0.19033313244581224,
      "num_tokens": 15198886.0,
      "step": 6630
    },
    {
      "entropy": 5.479446506500244,
      "epoch": 0.6373679154658982,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004966918456718256,
      "loss": 5.3744,
      "mean_token_accuracy": 0.18606213927268983,
      "num_tokens": 15211026.0,
      "step": 6635
    },
    {
      "entropy": 5.237106609344482,
      "epoch": 0.6378482228626321,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004966859857610828,
      "loss": 5.2724,
      "mean_token_accuracy": 0.20162095427513121,
      "num_tokens": 15222094.0,
      "step": 6640
    },
    {
      "entropy": 5.409952211380005,
      "epoch": 0.638328530259366,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004966801207034497,
      "loss": 5.3664,
      "mean_token_accuracy": 0.1915718659758568,
      "num_tokens": 15232464.0,
      "step": 6645
    },
    {
      "entropy": 5.457336807250977,
      "epoch": 0.6388088376560999,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004966742504990624,
      "loss": 5.2849,
      "mean_token_accuracy": 0.19506607204675674,
      "num_tokens": 15244564.0,
      "step": 6650
    },
    {
      "entropy": 5.454538059234619,
      "epoch": 0.6392891450528339,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004966683751480572,
      "loss": 5.43,
      "mean_token_accuracy": 0.18866757303476334,
      "num_tokens": 15255822.0,
      "step": 6655
    },
    {
      "entropy": 5.465950059890747,
      "epoch": 0.6397694524495677,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004966624946505706,
      "loss": 5.3305,
      "mean_token_accuracy": 0.18331922441720963,
      "num_tokens": 15267336.0,
      "step": 6660
    },
    {
      "entropy": 5.375289535522461,
      "epoch": 0.6402497598463016,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004966566090067391,
      "loss": 5.226,
      "mean_token_accuracy": 0.19735877364873886,
      "num_tokens": 15278073.0,
      "step": 6665
    },
    {
      "entropy": 5.408173608779907,
      "epoch": 0.6407300672430355,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004966507182166993,
      "loss": 5.4369,
      "mean_token_accuracy": 0.1848673477768898,
      "num_tokens": 15290119.0,
      "step": 6670
    },
    {
      "entropy": 5.342706203460693,
      "epoch": 0.6412103746397695,
      "grad_norm": 1.359375,
      "learning_rate": 0.000496644822280588,
      "loss": 5.2129,
      "mean_token_accuracy": 0.19946765452623366,
      "num_tokens": 15302235.0,
      "step": 6675
    },
    {
      "entropy": 5.455457592010498,
      "epoch": 0.6416906820365034,
      "grad_norm": 1.8125,
      "learning_rate": 0.000496638921198542,
      "loss": 5.4259,
      "mean_token_accuracy": 0.17861731052398683,
      "num_tokens": 15312924.0,
      "step": 6680
    },
    {
      "entropy": 5.287777090072632,
      "epoch": 0.6421709894332372,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004966330149706984,
      "loss": 5.2375,
      "mean_token_accuracy": 0.2019541308283806,
      "num_tokens": 15323453.0,
      "step": 6685
    },
    {
      "entropy": 5.39892840385437,
      "epoch": 0.6426512968299711,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004966271035971944,
      "loss": 5.2818,
      "mean_token_accuracy": 0.19235245883464813,
      "num_tokens": 15334907.0,
      "step": 6690
    },
    {
      "entropy": 5.398792362213134,
      "epoch": 0.6431316042267051,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004966211870781671,
      "loss": 5.275,
      "mean_token_accuracy": 0.1934860497713089,
      "num_tokens": 15347211.0,
      "step": 6695
    },
    {
      "entropy": 5.406028366088867,
      "epoch": 0.643611911623439,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000496615265413754,
      "loss": 5.452,
      "mean_token_accuracy": 0.18037029653787612,
      "num_tokens": 15360171.0,
      "step": 6700
    },
    {
      "entropy": 5.566893434524536,
      "epoch": 0.6440922190201729,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004966093386040923,
      "loss": 5.5043,
      "mean_token_accuracy": 0.18123398423194886,
      "num_tokens": 15372209.0,
      "step": 6705
    },
    {
      "entropy": 5.519690656661988,
      "epoch": 0.6445725264169068,
      "grad_norm": 2.28125,
      "learning_rate": 0.00049660340664932,
      "loss": 5.3894,
      "mean_token_accuracy": 0.18608277589082717,
      "num_tokens": 15383753.0,
      "step": 6710
    },
    {
      "entropy": 5.458930158615113,
      "epoch": 0.6450528338136408,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004965974695495746,
      "loss": 5.3641,
      "mean_token_accuracy": 0.18930920511484145,
      "num_tokens": 15395202.0,
      "step": 6715
    },
    {
      "entropy": 5.4900514602661135,
      "epoch": 0.6455331412103746,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004965915273049938,
      "loss": 5.341,
      "mean_token_accuracy": 0.1904049649834633,
      "num_tokens": 15407121.0,
      "step": 6720
    },
    {
      "entropy": 5.368436098098755,
      "epoch": 0.6460134486071085,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004965855799157158,
      "loss": 5.3112,
      "mean_token_accuracy": 0.191600301861763,
      "num_tokens": 15418063.0,
      "step": 6725
    },
    {
      "entropy": 5.434974575042725,
      "epoch": 0.6464937560038425,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004965796273818787,
      "loss": 5.3369,
      "mean_token_accuracy": 0.18815235495567323,
      "num_tokens": 15428830.0,
      "step": 6730
    },
    {
      "entropy": 5.413639879226684,
      "epoch": 0.6469740634005764,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004965736697036206,
      "loss": 5.2661,
      "mean_token_accuracy": 0.18711058497428895,
      "num_tokens": 15441009.0,
      "step": 6735
    },
    {
      "entropy": 5.401315784454345,
      "epoch": 0.6474543707973103,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004965677068810798,
      "loss": 5.3455,
      "mean_token_accuracy": 0.1954300120472908,
      "num_tokens": 15453008.0,
      "step": 6740
    },
    {
      "entropy": 5.469344186782837,
      "epoch": 0.6479346781940442,
      "grad_norm": 1.5,
      "learning_rate": 0.0004965617389143947,
      "loss": 5.3626,
      "mean_token_accuracy": 0.1931136205792427,
      "num_tokens": 15463907.0,
      "step": 6745
    },
    {
      "entropy": 5.430829906463623,
      "epoch": 0.6484149855907781,
      "grad_norm": 1.40625,
      "learning_rate": 0.000496555765803704,
      "loss": 5.2545,
      "mean_token_accuracy": 0.19242316335439683,
      "num_tokens": 15474706.0,
      "step": 6750
    },
    {
      "entropy": 5.408749389648437,
      "epoch": 0.648895292987512,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004965497875491462,
      "loss": 5.3276,
      "mean_token_accuracy": 0.1895291328430176,
      "num_tokens": 15486355.0,
      "step": 6755
    },
    {
      "entropy": 5.448469829559326,
      "epoch": 0.6493756003842459,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004965438041508604,
      "loss": 5.3432,
      "mean_token_accuracy": 0.18742077946662902,
      "num_tokens": 15498184.0,
      "step": 6760
    },
    {
      "entropy": 5.487996292114258,
      "epoch": 0.6498559077809798,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004965378156089851,
      "loss": 5.3932,
      "mean_token_accuracy": 0.18962435573339462,
      "num_tokens": 15509461.0,
      "step": 6765
    },
    {
      "entropy": 5.362503623962402,
      "epoch": 0.6503362151777138,
      "grad_norm": 1.375,
      "learning_rate": 0.0004965318219236597,
      "loss": 5.3264,
      "mean_token_accuracy": 0.1874221906065941,
      "num_tokens": 15520169.0,
      "step": 6770
    },
    {
      "entropy": 5.454718112945557,
      "epoch": 0.6508165225744477,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004965258230950232,
      "loss": 5.4094,
      "mean_token_accuracy": 0.18056175708770753,
      "num_tokens": 15532783.0,
      "step": 6775
    },
    {
      "entropy": 5.46618161201477,
      "epoch": 0.6512968299711815,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004965198191232148,
      "loss": 5.3124,
      "mean_token_accuracy": 0.18317433297634125,
      "num_tokens": 15543639.0,
      "step": 6780
    },
    {
      "entropy": 5.372718667984008,
      "epoch": 0.6517771373679154,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000496513810008374,
      "loss": 5.3047,
      "mean_token_accuracy": 0.1894352987408638,
      "num_tokens": 15555149.0,
      "step": 6785
    },
    {
      "entropy": 5.445323848724366,
      "epoch": 0.6522574447646494,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004965077957506403,
      "loss": 5.3549,
      "mean_token_accuracy": 0.19207099080085754,
      "num_tokens": 15566922.0,
      "step": 6790
    },
    {
      "entropy": 5.512098979949951,
      "epoch": 0.6527377521613833,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004965017763501533,
      "loss": 5.4156,
      "mean_token_accuracy": 0.1858382925391197,
      "num_tokens": 15578047.0,
      "step": 6795
    },
    {
      "entropy": 5.4678209781646725,
      "epoch": 0.6532180595581172,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004964957518070529,
      "loss": 5.4269,
      "mean_token_accuracy": 0.1829053059220314,
      "num_tokens": 15590378.0,
      "step": 6800
    },
    {
      "entropy": 5.465389442443848,
      "epoch": 0.6536983669548511,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004964897221214788,
      "loss": 5.3585,
      "mean_token_accuracy": 0.1873955100774765,
      "num_tokens": 15600866.0,
      "step": 6805
    },
    {
      "entropy": 5.442267370223999,
      "epoch": 0.654178674351585,
      "grad_norm": 1.6875,
      "learning_rate": 0.000496483687293571,
      "loss": 5.2324,
      "mean_token_accuracy": 0.18718597888946534,
      "num_tokens": 15612617.0,
      "step": 6810
    },
    {
      "entropy": 5.405890989303589,
      "epoch": 0.6546589817483189,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004964776473234696,
      "loss": 5.3329,
      "mean_token_accuracy": 0.1965491771697998,
      "num_tokens": 15624446.0,
      "step": 6815
    },
    {
      "entropy": 5.394840717315674,
      "epoch": 0.6551392891450528,
      "grad_norm": 1.640625,
      "learning_rate": 0.000496471602211315,
      "loss": 5.2981,
      "mean_token_accuracy": 0.19656362235546113,
      "num_tokens": 15635300.0,
      "step": 6820
    },
    {
      "entropy": 5.366667032241821,
      "epoch": 0.6556195965417867,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004964655519572475,
      "loss": 5.2208,
      "mean_token_accuracy": 0.19457219541072845,
      "num_tokens": 15646452.0,
      "step": 6825
    },
    {
      "entropy": 5.398584508895874,
      "epoch": 0.6560999039385207,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004964594965614072,
      "loss": 5.4091,
      "mean_token_accuracy": 0.18492621332406997,
      "num_tokens": 15658514.0,
      "step": 6830
    },
    {
      "entropy": 5.5249849319458,
      "epoch": 0.6565802113352546,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004964534360239353,
      "loss": 5.3289,
      "mean_token_accuracy": 0.1848491057753563,
      "num_tokens": 15670222.0,
      "step": 6835
    },
    {
      "entropy": 5.390076684951782,
      "epoch": 0.6570605187319885,
      "grad_norm": 2.0625,
      "learning_rate": 0.000496447370344972,
      "loss": 5.3375,
      "mean_token_accuracy": 0.18988256454467772,
      "num_tokens": 15681627.0,
      "step": 6840
    },
    {
      "entropy": 5.392456865310669,
      "epoch": 0.6575408261287223,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004964412995246584,
      "loss": 5.291,
      "mean_token_accuracy": 0.18991410583257676,
      "num_tokens": 15692632.0,
      "step": 6845
    },
    {
      "entropy": 5.395306730270386,
      "epoch": 0.6580211335254563,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004964352235631354,
      "loss": 5.3196,
      "mean_token_accuracy": 0.19022546857595443,
      "num_tokens": 15704736.0,
      "step": 6850
    },
    {
      "entropy": 5.437859106063843,
      "epoch": 0.6585014409221902,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000496429142460544,
      "loss": 5.4344,
      "mean_token_accuracy": 0.18680487126111983,
      "num_tokens": 15717057.0,
      "step": 6855
    },
    {
      "entropy": 5.495915222167969,
      "epoch": 0.6589817483189241,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004964230562170254,
      "loss": 5.4921,
      "mean_token_accuracy": 0.1785072550177574,
      "num_tokens": 15728577.0,
      "step": 6860
    },
    {
      "entropy": 5.459878587722779,
      "epoch": 0.659462055715658,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004964169648327209,
      "loss": 5.3183,
      "mean_token_accuracy": 0.1872129037976265,
      "num_tokens": 15739714.0,
      "step": 6865
    },
    {
      "entropy": 5.437222337722778,
      "epoch": 0.659942363112392,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000496410868307772,
      "loss": 5.226,
      "mean_token_accuracy": 0.19761881232261658,
      "num_tokens": 15751339.0,
      "step": 6870
    },
    {
      "entropy": 5.520535659790039,
      "epoch": 0.6604226705091258,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004964047666423203,
      "loss": 5.5121,
      "mean_token_accuracy": 0.18299062699079513,
      "num_tokens": 15763177.0,
      "step": 6875
    },
    {
      "entropy": 5.459117794036866,
      "epoch": 0.6609029779058597,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004963986598365072,
      "loss": 5.3926,
      "mean_token_accuracy": 0.18871883153915406,
      "num_tokens": 15774691.0,
      "step": 6880
    },
    {
      "entropy": 5.517058086395264,
      "epoch": 0.6613832853025937,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004963925478904746,
      "loss": 5.3804,
      "mean_token_accuracy": 0.18887075483798982,
      "num_tokens": 15785561.0,
      "step": 6885
    },
    {
      "entropy": 5.466109323501587,
      "epoch": 0.6618635926993276,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004963864308043645,
      "loss": 5.3984,
      "mean_token_accuracy": 0.18537548631429673,
      "num_tokens": 15795687.0,
      "step": 6890
    },
    {
      "entropy": 5.3591382026672365,
      "epoch": 0.6623439000960615,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004963803085783189,
      "loss": 5.3574,
      "mean_token_accuracy": 0.1918278232216835,
      "num_tokens": 15807191.0,
      "step": 6895
    },
    {
      "entropy": 5.452403974533081,
      "epoch": 0.6628242074927954,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004963741812124799,
      "loss": 5.3325,
      "mean_token_accuracy": 0.18980597704648972,
      "num_tokens": 15819585.0,
      "step": 6900
    },
    {
      "entropy": 5.505056524276734,
      "epoch": 0.6633045148895294,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004963680487069898,
      "loss": 5.4211,
      "mean_token_accuracy": 0.188141630589962,
      "num_tokens": 15831580.0,
      "step": 6905
    },
    {
      "entropy": 5.372838973999023,
      "epoch": 0.6637848222862632,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004963619110619908,
      "loss": 5.349,
      "mean_token_accuracy": 0.18626470267772674,
      "num_tokens": 15843848.0,
      "step": 6910
    },
    {
      "entropy": 5.509882307052612,
      "epoch": 0.6642651296829971,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004963557682776256,
      "loss": 5.4331,
      "mean_token_accuracy": 0.1808481752872467,
      "num_tokens": 15855091.0,
      "step": 6915
    },
    {
      "entropy": 5.444598722457886,
      "epoch": 0.664745437079731,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004963496203540368,
      "loss": 5.3722,
      "mean_token_accuracy": 0.19024813771247864,
      "num_tokens": 15866676.0,
      "step": 6920
    },
    {
      "entropy": 5.328142070770264,
      "epoch": 0.665225744476465,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004963434672913671,
      "loss": 5.2125,
      "mean_token_accuracy": 0.1954273462295532,
      "num_tokens": 15877901.0,
      "step": 6925
    },
    {
      "entropy": 5.398678112030029,
      "epoch": 0.6657060518731989,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004963373090897592,
      "loss": 5.3069,
      "mean_token_accuracy": 0.18265456557273865,
      "num_tokens": 15889284.0,
      "step": 6930
    },
    {
      "entropy": 5.373099327087402,
      "epoch": 0.6661863592699327,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004963311457493563,
      "loss": 5.2673,
      "mean_token_accuracy": 0.1994304671883583,
      "num_tokens": 15901218.0,
      "step": 6935
    },
    {
      "entropy": 5.30461483001709,
      "epoch": 0.6666666666666666,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004963249772703015,
      "loss": 5.2118,
      "mean_token_accuracy": 0.19947979152202605,
      "num_tokens": 15912452.0,
      "step": 6940
    },
    {
      "entropy": 5.380452871322632,
      "epoch": 0.6671469740634006,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004963188036527378,
      "loss": 5.3339,
      "mean_token_accuracy": 0.1922300785779953,
      "num_tokens": 15924353.0,
      "step": 6945
    },
    {
      "entropy": 5.5066570281982425,
      "epoch": 0.6676272814601345,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004963126248968087,
      "loss": 5.3545,
      "mean_token_accuracy": 0.19190370440483093,
      "num_tokens": 15936295.0,
      "step": 6950
    },
    {
      "entropy": 5.464797401428223,
      "epoch": 0.6681075888568684,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004963064410026577,
      "loss": 5.3367,
      "mean_token_accuracy": 0.18397328108549119,
      "num_tokens": 15948407.0,
      "step": 6955
    },
    {
      "entropy": 5.408189535140991,
      "epoch": 0.6685878962536023,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004963002519704281,
      "loss": 5.2926,
      "mean_token_accuracy": 0.19000355303287506,
      "num_tokens": 15959975.0,
      "step": 6960
    },
    {
      "entropy": 5.462561798095703,
      "epoch": 0.6690682036503363,
      "grad_norm": 1.5625,
      "learning_rate": 0.000496294057800264,
      "loss": 5.2959,
      "mean_token_accuracy": 0.19165364801883697,
      "num_tokens": 15973084.0,
      "step": 6965
    },
    {
      "entropy": 5.344120168685913,
      "epoch": 0.6695485110470701,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004962878584923089,
      "loss": 5.2058,
      "mean_token_accuracy": 0.19582221955060958,
      "num_tokens": 15984444.0,
      "step": 6970
    },
    {
      "entropy": 5.342095232009887,
      "epoch": 0.670028818443804,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004962816540467068,
      "loss": 5.3387,
      "mean_token_accuracy": 0.1910207062959671,
      "num_tokens": 15996611.0,
      "step": 6975
    },
    {
      "entropy": 5.418095445632934,
      "epoch": 0.6705091258405379,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004962754444636017,
      "loss": 5.3125,
      "mean_token_accuracy": 0.1870596617460251,
      "num_tokens": 16007144.0,
      "step": 6980
    },
    {
      "entropy": 5.389230728149414,
      "epoch": 0.6709894332372719,
      "grad_norm": 1.578125,
      "learning_rate": 0.000496269229743138,
      "loss": 5.3284,
      "mean_token_accuracy": 0.19333738535642625,
      "num_tokens": 16018878.0,
      "step": 6985
    },
    {
      "entropy": 5.478294420242309,
      "epoch": 0.6714697406340058,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004962630098854597,
      "loss": 5.346,
      "mean_token_accuracy": 0.18667296022176744,
      "num_tokens": 16031273.0,
      "step": 6990
    },
    {
      "entropy": 5.479709196090698,
      "epoch": 0.6719500480307397,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004962567848907113,
      "loss": 5.3939,
      "mean_token_accuracy": 0.18379569351673125,
      "num_tokens": 16043101.0,
      "step": 6995
    },
    {
      "entropy": 5.352395915985108,
      "epoch": 0.6724303554274735,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004962505547590374,
      "loss": 5.2679,
      "mean_token_accuracy": 0.18883143812417985,
      "num_tokens": 16054922.0,
      "step": 7000
    },
    {
      "entropy": 5.490764188766479,
      "epoch": 0.6729106628242075,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004962443194905826,
      "loss": 5.3408,
      "mean_token_accuracy": 0.1831253096461296,
      "num_tokens": 16065461.0,
      "step": 7005
    },
    {
      "entropy": 5.434230852127075,
      "epoch": 0.6733909702209414,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004962380790854916,
      "loss": 5.4206,
      "mean_token_accuracy": 0.1788187012076378,
      "num_tokens": 16076715.0,
      "step": 7010
    },
    {
      "entropy": 5.397006845474243,
      "epoch": 0.6738712776176753,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004962318335439094,
      "loss": 5.3549,
      "mean_token_accuracy": 0.19198958575725555,
      "num_tokens": 16088937.0,
      "step": 7015
    },
    {
      "entropy": 5.384761619567871,
      "epoch": 0.6743515850144092,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004962255828659809,
      "loss": 5.2667,
      "mean_token_accuracy": 0.19809892773628235,
      "num_tokens": 16100004.0,
      "step": 7020
    },
    {
      "entropy": 5.4446470737457275,
      "epoch": 0.6748318924111432,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004962193270518513,
      "loss": 5.3497,
      "mean_token_accuracy": 0.1927206665277481,
      "num_tokens": 16112435.0,
      "step": 7025
    },
    {
      "entropy": 5.46750955581665,
      "epoch": 0.675312199807877,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004962130661016659,
      "loss": 5.4471,
      "mean_token_accuracy": 0.18370037376880646,
      "num_tokens": 16124246.0,
      "step": 7030
    },
    {
      "entropy": 5.542457580566406,
      "epoch": 0.6757925072046109,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004962068000155699,
      "loss": 5.4281,
      "mean_token_accuracy": 0.18194731920957566,
      "num_tokens": 16136291.0,
      "step": 7035
    },
    {
      "entropy": 5.392105674743652,
      "epoch": 0.6762728146013448,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004962005287937088,
      "loss": 5.2617,
      "mean_token_accuracy": 0.19185828268527985,
      "num_tokens": 16148641.0,
      "step": 7040
    },
    {
      "entropy": 5.339692640304565,
      "epoch": 0.6767531219980788,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004961942524362283,
      "loss": 5.2537,
      "mean_token_accuracy": 0.20256917625665666,
      "num_tokens": 16159152.0,
      "step": 7045
    },
    {
      "entropy": 5.4464000225067135,
      "epoch": 0.6772334293948127,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004961879709432741,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18551846593618393,
      "num_tokens": 16170019.0,
      "step": 7050
    },
    {
      "entropy": 5.526552438735962,
      "epoch": 0.6777137367915466,
      "grad_norm": 1.625,
      "learning_rate": 0.000496181684314992,
      "loss": 5.3865,
      "mean_token_accuracy": 0.18400342911481857,
      "num_tokens": 16181714.0,
      "step": 7055
    },
    {
      "entropy": 5.4214578628540036,
      "epoch": 0.6781940441882806,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004961753925515279,
      "loss": 5.4157,
      "mean_token_accuracy": 0.18342450708150865,
      "num_tokens": 16192822.0,
      "step": 7060
    },
    {
      "entropy": 5.3091799259185795,
      "epoch": 0.6786743515850144,
      "grad_norm": 2.015625,
      "learning_rate": 0.000496169095653028,
      "loss": 5.2214,
      "mean_token_accuracy": 0.1941420316696167,
      "num_tokens": 16203660.0,
      "step": 7065
    },
    {
      "entropy": 5.385712766647339,
      "epoch": 0.6791546589817483,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004961627936196384,
      "loss": 5.3231,
      "mean_token_accuracy": 0.19611474722623826,
      "num_tokens": 16215495.0,
      "step": 7070
    },
    {
      "entropy": 5.422764587402344,
      "epoch": 0.6796349663784822,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004961564864515055,
      "loss": 5.2821,
      "mean_token_accuracy": 0.1954946756362915,
      "num_tokens": 16227532.0,
      "step": 7075
    },
    {
      "entropy": 5.2686676502227785,
      "epoch": 0.6801152737752162,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004961501741487757,
      "loss": 5.1323,
      "mean_token_accuracy": 0.2048894628882408,
      "num_tokens": 16238979.0,
      "step": 7080
    },
    {
      "entropy": 5.274458646774292,
      "epoch": 0.6805955811719501,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004961438567115955,
      "loss": 5.1784,
      "mean_token_accuracy": 0.19594721049070357,
      "num_tokens": 16249146.0,
      "step": 7085
    },
    {
      "entropy": 5.386977481842041,
      "epoch": 0.681075888568684,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004961375341401116,
      "loss": 5.3001,
      "mean_token_accuracy": 0.18883857429027556,
      "num_tokens": 16262269.0,
      "step": 7090
    },
    {
      "entropy": 5.397275638580322,
      "epoch": 0.6815561959654178,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004961312064344708,
      "loss": 5.2629,
      "mean_token_accuracy": 0.1970457300543785,
      "num_tokens": 16273877.0,
      "step": 7095
    },
    {
      "entropy": 5.50013575553894,
      "epoch": 0.6820365033621518,
      "grad_norm": 1.34375,
      "learning_rate": 0.00049612487359482,
      "loss": 5.5355,
      "mean_token_accuracy": 0.18614646792411804,
      "num_tokens": 16285623.0,
      "step": 7100
    },
    {
      "entropy": 5.355951929092408,
      "epoch": 0.6825168107588857,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004961185356213062,
      "loss": 5.2371,
      "mean_token_accuracy": 0.19960283786058425,
      "num_tokens": 16296242.0,
      "step": 7105
    },
    {
      "entropy": 5.445132923126221,
      "epoch": 0.6829971181556196,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004961121925140767,
      "loss": 5.4084,
      "mean_token_accuracy": 0.18793877810239792,
      "num_tokens": 16308400.0,
      "step": 7110
    },
    {
      "entropy": 5.470288705825806,
      "epoch": 0.6834774255523535,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004961058442732786,
      "loss": 5.3125,
      "mean_token_accuracy": 0.18972494900226594,
      "num_tokens": 16319201.0,
      "step": 7115
    },
    {
      "entropy": 5.4567423343658445,
      "epoch": 0.6839577329490875,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004960994908990594,
      "loss": 5.3504,
      "mean_token_accuracy": 0.18735696971416474,
      "num_tokens": 16331462.0,
      "step": 7120
    },
    {
      "entropy": 5.435454034805298,
      "epoch": 0.6844380403458213,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004960931323915665,
      "loss": 5.3607,
      "mean_token_accuracy": 0.18979617804288865,
      "num_tokens": 16343428.0,
      "step": 7125
    },
    {
      "entropy": 5.474292659759522,
      "epoch": 0.6849183477425552,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004960867687509475,
      "loss": 5.2421,
      "mean_token_accuracy": 0.19473725706338882,
      "num_tokens": 16356270.0,
      "step": 7130
    },
    {
      "entropy": 5.35652437210083,
      "epoch": 0.6853986551392891,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004960803999773504,
      "loss": 5.2839,
      "mean_token_accuracy": 0.19021278172731398,
      "num_tokens": 16367039.0,
      "step": 7135
    },
    {
      "entropy": 5.306606674194336,
      "epoch": 0.6858789625360231,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004960740260709228,
      "loss": 5.3453,
      "mean_token_accuracy": 0.19168916046619416,
      "num_tokens": 16378601.0,
      "step": 7140
    },
    {
      "entropy": 5.479401254653931,
      "epoch": 0.686359269932757,
      "grad_norm": 6.59375,
      "learning_rate": 0.0004960676470318128,
      "loss": 5.3885,
      "mean_token_accuracy": 0.1916048139333725,
      "num_tokens": 16391096.0,
      "step": 7145
    },
    {
      "entropy": 5.389495229721069,
      "epoch": 0.6868395773294909,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004960612628601683,
      "loss": 5.2165,
      "mean_token_accuracy": 0.1973610907793045,
      "num_tokens": 16401527.0,
      "step": 7150
    },
    {
      "entropy": 5.425680160522461,
      "epoch": 0.6873198847262247,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004960548735561379,
      "loss": 5.3634,
      "mean_token_accuracy": 0.17701541036367416,
      "num_tokens": 16413948.0,
      "step": 7155
    },
    {
      "entropy": 5.4044373512268065,
      "epoch": 0.6878001921229587,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004960484791198697,
      "loss": 5.2641,
      "mean_token_accuracy": 0.19330644607543945,
      "num_tokens": 16425718.0,
      "step": 7160
    },
    {
      "entropy": 5.375082635879517,
      "epoch": 0.6882804995196926,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004960420795515121,
      "loss": 5.3014,
      "mean_token_accuracy": 0.19107518941164017,
      "num_tokens": 16437606.0,
      "step": 7165
    },
    {
      "entropy": 5.441378784179688,
      "epoch": 0.6887608069164265,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004960356748512138,
      "loss": 5.3421,
      "mean_token_accuracy": 0.19059857875108718,
      "num_tokens": 16448964.0,
      "step": 7170
    },
    {
      "entropy": 5.374948358535766,
      "epoch": 0.6892411143131604,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004960292650191236,
      "loss": 5.3246,
      "mean_token_accuracy": 0.1918066069483757,
      "num_tokens": 16459974.0,
      "step": 7175
    },
    {
      "entropy": 5.420234155654907,
      "epoch": 0.6897214217098944,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004960228500553899,
      "loss": 5.2751,
      "mean_token_accuracy": 0.18620239049196244,
      "num_tokens": 16471143.0,
      "step": 7180
    },
    {
      "entropy": 5.381014966964722,
      "epoch": 0.6902017291066282,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004960164299601623,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19609259814023972,
      "num_tokens": 16482090.0,
      "step": 7185
    },
    {
      "entropy": 5.330925703048706,
      "epoch": 0.6906820365033621,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004960100047335892,
      "loss": 5.2879,
      "mean_token_accuracy": 0.19356450587511062,
      "num_tokens": 16493695.0,
      "step": 7190
    },
    {
      "entropy": 5.465975570678711,
      "epoch": 0.691162343900096,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004960035743758202,
      "loss": 5.3727,
      "mean_token_accuracy": 0.189803908765316,
      "num_tokens": 16505317.0,
      "step": 7195
    },
    {
      "entropy": 5.43049807548523,
      "epoch": 0.69164265129683,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004959971388870044,
      "loss": 5.2982,
      "mean_token_accuracy": 0.1919502481818199,
      "num_tokens": 16516719.0,
      "step": 7200
    },
    {
      "entropy": 5.382589673995971,
      "epoch": 0.6921229586935639,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004959906982672912,
      "loss": 5.2512,
      "mean_token_accuracy": 0.19547758549451827,
      "num_tokens": 16528302.0,
      "step": 7205
    },
    {
      "entropy": 5.351429414749146,
      "epoch": 0.6926032660902978,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004959842525168302,
      "loss": 5.2777,
      "mean_token_accuracy": 0.18983298689126968,
      "num_tokens": 16538471.0,
      "step": 7210
    },
    {
      "entropy": 5.440526485443115,
      "epoch": 0.6930835734870316,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004959778016357712,
      "loss": 5.2917,
      "mean_token_accuracy": 0.1949208378791809,
      "num_tokens": 16549283.0,
      "step": 7215
    },
    {
      "entropy": 5.3660050392150875,
      "epoch": 0.6935638808837656,
      "grad_norm": 2.953125,
      "learning_rate": 0.0004959713456242637,
      "loss": 5.2603,
      "mean_token_accuracy": 0.18636907190084456,
      "num_tokens": 16560895.0,
      "step": 7220
    },
    {
      "entropy": 5.37461838722229,
      "epoch": 0.6940441882804995,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004959648844824576,
      "loss": 5.3562,
      "mean_token_accuracy": 0.191277414560318,
      "num_tokens": 16571306.0,
      "step": 7225
    },
    {
      "entropy": 5.417568969726562,
      "epoch": 0.6945244956772334,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004959584182105032,
      "loss": 5.2665,
      "mean_token_accuracy": 0.19584214687347412,
      "num_tokens": 16582399.0,
      "step": 7230
    },
    {
      "entropy": 5.346726608276367,
      "epoch": 0.6950048030739674,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004959519468085504,
      "loss": 5.2406,
      "mean_token_accuracy": 0.19755541235208512,
      "num_tokens": 16593617.0,
      "step": 7235
    },
    {
      "entropy": 5.332418775558471,
      "epoch": 0.6954851104707013,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004959454702767494,
      "loss": 5.3137,
      "mean_token_accuracy": 0.19345726519823075,
      "num_tokens": 16604109.0,
      "step": 7240
    },
    {
      "entropy": 5.483483839035034,
      "epoch": 0.6959654178674352,
      "grad_norm": 1.5,
      "learning_rate": 0.0004959389886152507,
      "loss": 5.3964,
      "mean_token_accuracy": 0.18649379312992095,
      "num_tokens": 16616007.0,
      "step": 7245
    },
    {
      "entropy": 5.439164161682129,
      "epoch": 0.696445725264169,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004959325018242048,
      "loss": 5.252,
      "mean_token_accuracy": 0.19164157062768936,
      "num_tokens": 16627609.0,
      "step": 7250
    },
    {
      "entropy": 5.414792919158936,
      "epoch": 0.696926032660903,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004959260099037622,
      "loss": 5.2541,
      "mean_token_accuracy": 0.18888843655586243,
      "num_tokens": 16639842.0,
      "step": 7255
    },
    {
      "entropy": 5.451376724243164,
      "epoch": 0.6974063400576369,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004959195128540737,
      "loss": 5.4087,
      "mean_token_accuracy": 0.17943353950977325,
      "num_tokens": 16651620.0,
      "step": 7260
    },
    {
      "entropy": 5.394506216049194,
      "epoch": 0.6978866474543708,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00049591301067529,
      "loss": 5.3296,
      "mean_token_accuracy": 0.1885972186923027,
      "num_tokens": 16663637.0,
      "step": 7265
    },
    {
      "entropy": 5.396517658233643,
      "epoch": 0.6983669548511047,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004959065033675623,
      "loss": 5.3722,
      "mean_token_accuracy": 0.18626101613044738,
      "num_tokens": 16675685.0,
      "step": 7270
    },
    {
      "entropy": 5.468096828460693,
      "epoch": 0.6988472622478387,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004958999909310414,
      "loss": 5.2928,
      "mean_token_accuracy": 0.18408166319131852,
      "num_tokens": 16686268.0,
      "step": 7275
    },
    {
      "entropy": 5.377413988113403,
      "epoch": 0.6993275696445725,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004958934733658788,
      "loss": 5.341,
      "mean_token_accuracy": 0.1924181893467903,
      "num_tokens": 16698382.0,
      "step": 7280
    },
    {
      "entropy": 5.456994152069091,
      "epoch": 0.6998078770413064,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004958869506722256,
      "loss": 5.4072,
      "mean_token_accuracy": 0.18200154155492781,
      "num_tokens": 16710421.0,
      "step": 7285
    },
    {
      "entropy": 5.4129057884216305,
      "epoch": 0.7002881844380403,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004958804228502332,
      "loss": 5.2263,
      "mean_token_accuracy": 0.19510667622089387,
      "num_tokens": 16721658.0,
      "step": 7290
    },
    {
      "entropy": 5.373976945877075,
      "epoch": 0.7007684918347743,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004958738899000534,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19303322732448577,
      "num_tokens": 16734100.0,
      "step": 7295
    },
    {
      "entropy": 5.306042098999024,
      "epoch": 0.7012487992315082,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004958673518218377,
      "loss": 5.2853,
      "mean_token_accuracy": 0.19008474051952362,
      "num_tokens": 16746386.0,
      "step": 7300
    },
    {
      "entropy": 5.457220983505249,
      "epoch": 0.7017291066282421,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004958608086157379,
      "loss": 5.3414,
      "mean_token_accuracy": 0.1953557088971138,
      "num_tokens": 16756967.0,
      "step": 7305
    },
    {
      "entropy": 5.366976404190064,
      "epoch": 0.7022094140249759,
      "grad_norm": 1.484375,
      "learning_rate": 0.000495854260281906,
      "loss": 5.2251,
      "mean_token_accuracy": 0.1987711861729622,
      "num_tokens": 16767454.0,
      "step": 7310
    },
    {
      "entropy": 5.294095659255982,
      "epoch": 0.7026897214217099,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004958477068204941,
      "loss": 5.295,
      "mean_token_accuracy": 0.1918737694621086,
      "num_tokens": 16779533.0,
      "step": 7315
    },
    {
      "entropy": 5.4249396324157715,
      "epoch": 0.7031700288184438,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000495841148231654,
      "loss": 5.3351,
      "mean_token_accuracy": 0.18856797367334366,
      "num_tokens": 16792019.0,
      "step": 7320
    },
    {
      "entropy": 5.36957893371582,
      "epoch": 0.7036503362151777,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004958345845155383,
      "loss": 5.2728,
      "mean_token_accuracy": 0.193443001806736,
      "num_tokens": 16803089.0,
      "step": 7325
    },
    {
      "entropy": 5.37037410736084,
      "epoch": 0.7041306436119116,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004958280156722992,
      "loss": 5.2215,
      "mean_token_accuracy": 0.1954813063144684,
      "num_tokens": 16813651.0,
      "step": 7330
    },
    {
      "entropy": 5.35937762260437,
      "epoch": 0.7046109510086456,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004958214417020894,
      "loss": 5.2611,
      "mean_token_accuracy": 0.1992407187819481,
      "num_tokens": 16825460.0,
      "step": 7335
    },
    {
      "entropy": 5.320385265350342,
      "epoch": 0.7050912584053795,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004958148626050614,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19488627314567566,
      "num_tokens": 16836344.0,
      "step": 7340
    },
    {
      "entropy": 5.401589012145996,
      "epoch": 0.7055715658021133,
      "grad_norm": 1.59375,
      "learning_rate": 0.000495808278381368,
      "loss": 5.2885,
      "mean_token_accuracy": 0.19181105196475984,
      "num_tokens": 16848274.0,
      "step": 7345
    },
    {
      "entropy": 5.4050452709198,
      "epoch": 0.7060518731988472,
      "grad_norm": 2.21875,
      "learning_rate": 0.000495801689031162,
      "loss": 5.3215,
      "mean_token_accuracy": 0.1882026121020317,
      "num_tokens": 16860363.0,
      "step": 7350
    },
    {
      "entropy": 5.364719724655151,
      "epoch": 0.7065321805955812,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004957950945545965,
      "loss": 5.2793,
      "mean_token_accuracy": 0.19339021742343904,
      "num_tokens": 16872037.0,
      "step": 7355
    },
    {
      "entropy": 5.470697450637817,
      "epoch": 0.7070124879923151,
      "grad_norm": 1.5,
      "learning_rate": 0.0004957884949518246,
      "loss": 5.3755,
      "mean_token_accuracy": 0.1876761481165886,
      "num_tokens": 16883845.0,
      "step": 7360
    },
    {
      "entropy": 5.4732177734375,
      "epoch": 0.707492795389049,
      "grad_norm": 1.375,
      "learning_rate": 0.0004957818902229992,
      "loss": 5.3377,
      "mean_token_accuracy": 0.18597989976406099,
      "num_tokens": 16894965.0,
      "step": 7365
    },
    {
      "entropy": 5.3630016326904295,
      "epoch": 0.7079731027857828,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004957752803682741,
      "loss": 5.3226,
      "mean_token_accuracy": 0.1854296952486038,
      "num_tokens": 16907475.0,
      "step": 7370
    },
    {
      "entropy": 5.355823707580567,
      "epoch": 0.7084534101825168,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004957686653878024,
      "loss": 5.2242,
      "mean_token_accuracy": 0.19145250618457793,
      "num_tokens": 16918825.0,
      "step": 7375
    },
    {
      "entropy": 5.425852203369141,
      "epoch": 0.7089337175792507,
      "grad_norm": 2.015625,
      "learning_rate": 0.000495762045281738,
      "loss": 5.2755,
      "mean_token_accuracy": 0.19509863257408142,
      "num_tokens": 16930663.0,
      "step": 7380
    },
    {
      "entropy": 5.394859647750854,
      "epoch": 0.7094140249759846,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004957554200502344,
      "loss": 5.3009,
      "mean_token_accuracy": 0.19530547708272933,
      "num_tokens": 16942089.0,
      "step": 7385
    },
    {
      "entropy": 5.339275598526001,
      "epoch": 0.7098943323727186,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004957487896934454,
      "loss": 5.3294,
      "mean_token_accuracy": 0.19127538353204726,
      "num_tokens": 16954203.0,
      "step": 7390
    },
    {
      "entropy": 5.370598554611206,
      "epoch": 0.7103746397694525,
      "grad_norm": 1.484375,
      "learning_rate": 0.000495742154211525,
      "loss": 5.2872,
      "mean_token_accuracy": 0.19286420345306396,
      "num_tokens": 16966047.0,
      "step": 7395
    },
    {
      "entropy": 5.281306743621826,
      "epoch": 0.7108549471661864,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004957355136046272,
      "loss": 5.1765,
      "mean_token_accuracy": 0.19679884910583495,
      "num_tokens": 16977726.0,
      "step": 7400
    },
    {
      "entropy": 5.350837755203247,
      "epoch": 0.7113352545629202,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004957288678729064,
      "loss": 5.2718,
      "mean_token_accuracy": 0.19669270366430283,
      "num_tokens": 16990917.0,
      "step": 7405
    },
    {
      "entropy": 5.348010396957397,
      "epoch": 0.7118155619596542,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004957222170165166,
      "loss": 5.2776,
      "mean_token_accuracy": 0.1934112176299095,
      "num_tokens": 17001703.0,
      "step": 7410
    },
    {
      "entropy": 5.335374593734741,
      "epoch": 0.7122958693563881,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004957155610356124,
      "loss": 5.1918,
      "mean_token_accuracy": 0.19498712420463563,
      "num_tokens": 17011918.0,
      "step": 7415
    },
    {
      "entropy": 5.462284994125366,
      "epoch": 0.712776176753122,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004957088999303481,
      "loss": 5.3406,
      "mean_token_accuracy": 0.19242413192987443,
      "num_tokens": 17022653.0,
      "step": 7420
    },
    {
      "entropy": 5.329063081741333,
      "epoch": 0.7132564841498559,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004957022337008787,
      "loss": 5.211,
      "mean_token_accuracy": 0.19977978318929673,
      "num_tokens": 17034775.0,
      "step": 7425
    },
    {
      "entropy": 5.319195938110352,
      "epoch": 0.7137367915465899,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004956955623473587,
      "loss": 5.2562,
      "mean_token_accuracy": 0.18822196871042252,
      "num_tokens": 17046835.0,
      "step": 7430
    },
    {
      "entropy": 5.493042373657227,
      "epoch": 0.7142170989433237,
      "grad_norm": 1.8125,
      "learning_rate": 0.000495688885869943,
      "loss": 5.3609,
      "mean_token_accuracy": 0.190398208796978,
      "num_tokens": 17059163.0,
      "step": 7435
    },
    {
      "entropy": 5.408080959320069,
      "epoch": 0.7146974063400576,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004956822042687868,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18790239095687866,
      "num_tokens": 17070777.0,
      "step": 7440
    },
    {
      "entropy": 5.421570777893066,
      "epoch": 0.7151777137367915,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004956755175440451,
      "loss": 5.3124,
      "mean_token_accuracy": 0.19134521931409837,
      "num_tokens": 17083770.0,
      "step": 7445
    },
    {
      "entropy": 5.404145669937134,
      "epoch": 0.7156580211335255,
      "grad_norm": 1.75,
      "learning_rate": 0.0004956688256958731,
      "loss": 5.3093,
      "mean_token_accuracy": 0.19096972942352294,
      "num_tokens": 17096049.0,
      "step": 7450
    },
    {
      "entropy": 5.298589134216309,
      "epoch": 0.7161383285302594,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004956621287244262,
      "loss": 5.1809,
      "mean_token_accuracy": 0.19574082046747207,
      "num_tokens": 17107030.0,
      "step": 7455
    },
    {
      "entropy": 5.278734397888184,
      "epoch": 0.7166186359269933,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004956554266298599,
      "loss": 5.2637,
      "mean_token_accuracy": 0.19119612127542496,
      "num_tokens": 17118638.0,
      "step": 7460
    },
    {
      "entropy": 5.41753511428833,
      "epoch": 0.7170989433237271,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004956487194123298,
      "loss": 5.3379,
      "mean_token_accuracy": 0.19086818844079972,
      "num_tokens": 17129783.0,
      "step": 7465
    },
    {
      "entropy": 5.431677532196045,
      "epoch": 0.7175792507204611,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004956420070719918,
      "loss": 5.332,
      "mean_token_accuracy": 0.18988775610923767,
      "num_tokens": 17141939.0,
      "step": 7470
    },
    {
      "entropy": 5.384487867355347,
      "epoch": 0.718059558117195,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004956352896090014,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19325017035007477,
      "num_tokens": 17153471.0,
      "step": 7475
    },
    {
      "entropy": 5.444349908828736,
      "epoch": 0.7185398655139289,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004956285670235147,
      "loss": 5.3435,
      "mean_token_accuracy": 0.18837197571992875,
      "num_tokens": 17164659.0,
      "step": 7480
    },
    {
      "entropy": 5.440321207046509,
      "epoch": 0.7190201729106628,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004956218393156879,
      "loss": 5.2832,
      "mean_token_accuracy": 0.19292935729026794,
      "num_tokens": 17175749.0,
      "step": 7485
    },
    {
      "entropy": 5.392650842666626,
      "epoch": 0.7195004803073968,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004956151064856772,
      "loss": 5.3417,
      "mean_token_accuracy": 0.18586882054805756,
      "num_tokens": 17187482.0,
      "step": 7490
    },
    {
      "entropy": 5.39450421333313,
      "epoch": 0.7199807877041307,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004956083685336386,
      "loss": 5.299,
      "mean_token_accuracy": 0.19471026808023453,
      "num_tokens": 17199904.0,
      "step": 7495
    },
    {
      "entropy": 5.371392869949341,
      "epoch": 0.7204610951008645,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004956016254597289,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1884932443499565,
      "num_tokens": 17210649.0,
      "step": 7500
    },
    {
      "entropy": 5.381744050979615,
      "epoch": 0.7209414024975984,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004955948772641044,
      "loss": 5.3247,
      "mean_token_accuracy": 0.18555935770273208,
      "num_tokens": 17223069.0,
      "step": 7505
    },
    {
      "entropy": 5.397015047073364,
      "epoch": 0.7214217098943324,
      "grad_norm": 1.5625,
      "learning_rate": 0.000495588123946922,
      "loss": 5.2452,
      "mean_token_accuracy": 0.19745109379291534,
      "num_tokens": 17235144.0,
      "step": 7510
    },
    {
      "entropy": 5.370836496353149,
      "epoch": 0.7219020172910663,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004955813655083384,
      "loss": 5.2255,
      "mean_token_accuracy": 0.19031732231378556,
      "num_tokens": 17246816.0,
      "step": 7515
    },
    {
      "entropy": 5.378262948989868,
      "epoch": 0.7223823246878002,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004955746019485104,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19473235309123993,
      "num_tokens": 17259056.0,
      "step": 7520
    },
    {
      "entropy": 5.463152456283569,
      "epoch": 0.722862632084534,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000495567833267595,
      "loss": 5.3161,
      "mean_token_accuracy": 0.1908079132437706,
      "num_tokens": 17270391.0,
      "step": 7525
    },
    {
      "entropy": 5.449782848358154,
      "epoch": 0.723342939481268,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004955610594657496,
      "loss": 5.2485,
      "mean_token_accuracy": 0.19024159759283066,
      "num_tokens": 17281608.0,
      "step": 7530
    },
    {
      "entropy": 5.404439258575439,
      "epoch": 0.7238232468780019,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004955542805431313,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18304315954446793,
      "num_tokens": 17293649.0,
      "step": 7535
    },
    {
      "entropy": 5.381678247451783,
      "epoch": 0.7243035542747358,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004955474964998976,
      "loss": 5.3225,
      "mean_token_accuracy": 0.18936212956905366,
      "num_tokens": 17305548.0,
      "step": 7540
    },
    {
      "entropy": 5.446031618118286,
      "epoch": 0.7247838616714697,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004955407073362058,
      "loss": 5.2635,
      "mean_token_accuracy": 0.19551405757665635,
      "num_tokens": 17317380.0,
      "step": 7545
    },
    {
      "entropy": 5.356749391555786,
      "epoch": 0.7252641690682037,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004955339130522136,
      "loss": 5.2707,
      "mean_token_accuracy": 0.18997751474380492,
      "num_tokens": 17329263.0,
      "step": 7550
    },
    {
      "entropy": 5.453513860702515,
      "epoch": 0.7257444764649376,
      "grad_norm": 1.359375,
      "learning_rate": 0.000495527113648079,
      "loss": 5.4218,
      "mean_token_accuracy": 0.18590964376926422,
      "num_tokens": 17343268.0,
      "step": 7555
    },
    {
      "entropy": 5.436152791976928,
      "epoch": 0.7262247838616714,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004955203091239596,
      "loss": 5.247,
      "mean_token_accuracy": 0.20218702405691147,
      "num_tokens": 17353979.0,
      "step": 7560
    },
    {
      "entropy": 5.366891670227051,
      "epoch": 0.7267050912584054,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004955134994800134,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1873335897922516,
      "num_tokens": 17367587.0,
      "step": 7565
    },
    {
      "entropy": 5.4155841827392575,
      "epoch": 0.7271853986551393,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004955066847163984,
      "loss": 5.3239,
      "mean_token_accuracy": 0.18941086679697036,
      "num_tokens": 17379093.0,
      "step": 7570
    },
    {
      "entropy": 5.461610507965088,
      "epoch": 0.7276657060518732,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004954998648332731,
      "loss": 5.2979,
      "mean_token_accuracy": 0.19426215440034866,
      "num_tokens": 17391254.0,
      "step": 7575
    },
    {
      "entropy": 5.329599523544312,
      "epoch": 0.7281460134486071,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004954930398307956,
      "loss": 5.2857,
      "mean_token_accuracy": 0.1910361185669899,
      "num_tokens": 17402532.0,
      "step": 7580
    },
    {
      "entropy": 5.429715776443482,
      "epoch": 0.7286263208453411,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004954862097091245,
      "loss": 5.3117,
      "mean_token_accuracy": 0.19515106827020645,
      "num_tokens": 17413849.0,
      "step": 7585
    },
    {
      "entropy": 5.343151712417603,
      "epoch": 0.729106628242075,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004954793744684184,
      "loss": 5.2036,
      "mean_token_accuracy": 0.20000943094491958,
      "num_tokens": 17424982.0,
      "step": 7590
    },
    {
      "entropy": 5.460010099411011,
      "epoch": 0.7295869356388088,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004954725341088358,
      "loss": 5.4433,
      "mean_token_accuracy": 0.179095321893692,
      "num_tokens": 17437774.0,
      "step": 7595
    },
    {
      "entropy": 5.454359292984009,
      "epoch": 0.7300672430355427,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004954656886305356,
      "loss": 5.3107,
      "mean_token_accuracy": 0.19837107956409455,
      "num_tokens": 17448676.0,
      "step": 7600
    },
    {
      "entropy": 5.436759281158447,
      "epoch": 0.7305475504322767,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004954588380336768,
      "loss": 5.2764,
      "mean_token_accuracy": 0.194442480802536,
      "num_tokens": 17459576.0,
      "step": 7605
    },
    {
      "entropy": 5.30203104019165,
      "epoch": 0.7310278578290106,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004954519823184184,
      "loss": 5.2104,
      "mean_token_accuracy": 0.2036220461130142,
      "num_tokens": 17471220.0,
      "step": 7610
    },
    {
      "entropy": 5.369773244857788,
      "epoch": 0.7315081652257445,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004954451214849196,
      "loss": 5.3604,
      "mean_token_accuracy": 0.18353285491466523,
      "num_tokens": 17483939.0,
      "step": 7615
    },
    {
      "entropy": 5.437369537353516,
      "epoch": 0.7319884726224783,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004954382555333397,
      "loss": 5.3179,
      "mean_token_accuracy": 0.196310655772686,
      "num_tokens": 17496058.0,
      "step": 7620
    },
    {
      "entropy": 5.3552384853363035,
      "epoch": 0.7324687800192123,
      "grad_norm": 1.34375,
      "learning_rate": 0.000495431384463838,
      "loss": 5.3327,
      "mean_token_accuracy": 0.1946453332901001,
      "num_tokens": 17507750.0,
      "step": 7625
    },
    {
      "entropy": 5.347149848937988,
      "epoch": 0.7329490874159462,
      "grad_norm": 2.0,
      "learning_rate": 0.0004954245082765741,
      "loss": 5.2712,
      "mean_token_accuracy": 0.2009843945503235,
      "num_tokens": 17520208.0,
      "step": 7630
    },
    {
      "entropy": 5.397583436965943,
      "epoch": 0.7334293948126801,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004954176269717077,
      "loss": 5.2116,
      "mean_token_accuracy": 0.1947871431708336,
      "num_tokens": 17531655.0,
      "step": 7635
    },
    {
      "entropy": 5.329985284805298,
      "epoch": 0.733909702209414,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004954107405493984,
      "loss": 5.2555,
      "mean_token_accuracy": 0.19235846996307374,
      "num_tokens": 17543766.0,
      "step": 7640
    },
    {
      "entropy": 5.432396936416626,
      "epoch": 0.734390009606148,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004954038490098064,
      "loss": 5.3427,
      "mean_token_accuracy": 0.18324017971754075,
      "num_tokens": 17555710.0,
      "step": 7645
    },
    {
      "entropy": 5.334878444671631,
      "epoch": 0.7348703170028819,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004953969523530914,
      "loss": 5.2101,
      "mean_token_accuracy": 0.19792094230651855,
      "num_tokens": 17566674.0,
      "step": 7650
    },
    {
      "entropy": 5.4214776992797855,
      "epoch": 0.7353506243996157,
      "grad_norm": 1.75,
      "learning_rate": 0.0004953900505794136,
      "loss": 5.3163,
      "mean_token_accuracy": 0.19025304466485976,
      "num_tokens": 17579320.0,
      "step": 7655
    },
    {
      "entropy": 5.4200674533844,
      "epoch": 0.7358309317963496,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004953831436889332,
      "loss": 5.2579,
      "mean_token_accuracy": 0.1850394368171692,
      "num_tokens": 17590679.0,
      "step": 7660
    },
    {
      "entropy": 5.367720127105713,
      "epoch": 0.7363112391930836,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004953762316818106,
      "loss": 5.23,
      "mean_token_accuracy": 0.20518195927143096,
      "num_tokens": 17600928.0,
      "step": 7665
    },
    {
      "entropy": 5.2725715160369875,
      "epoch": 0.7367915465898175,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004953693145582064,
      "loss": 5.2,
      "mean_token_accuracy": 0.20383870005607604,
      "num_tokens": 17611906.0,
      "step": 7670
    },
    {
      "entropy": 5.323664426803589,
      "epoch": 0.7372718539865514,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000495362392318281,
      "loss": 5.3641,
      "mean_token_accuracy": 0.19108777940273286,
      "num_tokens": 17623165.0,
      "step": 7675
    },
    {
      "entropy": 5.452879619598389,
      "epoch": 0.7377521613832853,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004953554649621951,
      "loss": 5.263,
      "mean_token_accuracy": 0.19726396799087526,
      "num_tokens": 17633632.0,
      "step": 7680
    },
    {
      "entropy": 5.3472850799560545,
      "epoch": 0.7382324687800192,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004953485324901098,
      "loss": 5.2691,
      "mean_token_accuracy": 0.19999130070209503,
      "num_tokens": 17644801.0,
      "step": 7685
    },
    {
      "entropy": 5.3102232933044435,
      "epoch": 0.7387127761767531,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004953415949021857,
      "loss": 5.2822,
      "mean_token_accuracy": 0.18756644278764725,
      "num_tokens": 17656852.0,
      "step": 7690
    },
    {
      "entropy": 5.41955976486206,
      "epoch": 0.739193083573487,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004953346521985843,
      "loss": 5.2921,
      "mean_token_accuracy": 0.1916919767856598,
      "num_tokens": 17667538.0,
      "step": 7695
    },
    {
      "entropy": 5.335768604278565,
      "epoch": 0.7396733909702209,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004953277043794663,
      "loss": 5.2096,
      "mean_token_accuracy": 0.19577167481184005,
      "num_tokens": 17680049.0,
      "step": 7700
    },
    {
      "entropy": 5.317630386352539,
      "epoch": 0.7401536983669549,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004953207514449933,
      "loss": 5.2795,
      "mean_token_accuracy": 0.1916391983628273,
      "num_tokens": 17692611.0,
      "step": 7705
    },
    {
      "entropy": 5.385031509399414,
      "epoch": 0.7406340057636888,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004953137933953267,
      "loss": 5.1664,
      "mean_token_accuracy": 0.1983467757701874,
      "num_tokens": 17704145.0,
      "step": 7710
    },
    {
      "entropy": 5.448598766326905,
      "epoch": 0.7411143131604226,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000495306830230628,
      "loss": 5.2911,
      "mean_token_accuracy": 0.18565167784690856,
      "num_tokens": 17715467.0,
      "step": 7715
    },
    {
      "entropy": 5.377538251876831,
      "epoch": 0.7415946205571565,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004952998619510589,
      "loss": 5.2922,
      "mean_token_accuracy": 0.19642675668001175,
      "num_tokens": 17726709.0,
      "step": 7720
    },
    {
      "entropy": 5.3718647956848145,
      "epoch": 0.7420749279538905,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004952928885567811,
      "loss": 5.2173,
      "mean_token_accuracy": 0.20163513571023942,
      "num_tokens": 17736896.0,
      "step": 7725
    },
    {
      "entropy": 5.421212959289551,
      "epoch": 0.7425552353506244,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004952859100479566,
      "loss": 5.2949,
      "mean_token_accuracy": 0.18503678739070892,
      "num_tokens": 17747658.0,
      "step": 7730
    },
    {
      "entropy": 5.380327463150024,
      "epoch": 0.7430355427473583,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004952789264247474,
      "loss": 5.2504,
      "mean_token_accuracy": 0.1982392832636833,
      "num_tokens": 17757833.0,
      "step": 7735
    },
    {
      "entropy": 5.331891298294067,
      "epoch": 0.7435158501440923,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004952719376873156,
      "loss": 5.2377,
      "mean_token_accuracy": 0.1930388405919075,
      "num_tokens": 17769144.0,
      "step": 7740
    },
    {
      "entropy": 5.41746940612793,
      "epoch": 0.7439961575408262,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004952649438358234,
      "loss": 5.3535,
      "mean_token_accuracy": 0.18804989010095596,
      "num_tokens": 17781357.0,
      "step": 7745
    },
    {
      "entropy": 5.495258522033692,
      "epoch": 0.74447646493756,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004952579448704334,
      "loss": 5.2794,
      "mean_token_accuracy": 0.1938771441578865,
      "num_tokens": 17793640.0,
      "step": 7750
    },
    {
      "entropy": 5.374048852920533,
      "epoch": 0.7449567723342939,
      "grad_norm": 1.9765625,
      "learning_rate": 0.000495250940791308,
      "loss": 5.2589,
      "mean_token_accuracy": 0.18867946565151214,
      "num_tokens": 17804228.0,
      "step": 7755
    },
    {
      "entropy": 5.355391120910644,
      "epoch": 0.7454370797310279,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004952439315986096,
      "loss": 5.2884,
      "mean_token_accuracy": 0.19074462354183197,
      "num_tokens": 17816153.0,
      "step": 7760
    },
    {
      "entropy": 5.3792389869689945,
      "epoch": 0.7459173871277618,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004952369172925012,
      "loss": 5.3104,
      "mean_token_accuracy": 0.1932196393609047,
      "num_tokens": 17828464.0,
      "step": 7765
    },
    {
      "entropy": 5.308259391784668,
      "epoch": 0.7463976945244957,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004952298978731454,
      "loss": 5.1823,
      "mean_token_accuracy": 0.19391650408506395,
      "num_tokens": 17840126.0,
      "step": 7770
    },
    {
      "entropy": 5.339615297317505,
      "epoch": 0.7468780019212296,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004952228733407055,
      "loss": 5.2492,
      "mean_token_accuracy": 0.19899795949459076,
      "num_tokens": 17851735.0,
      "step": 7775
    },
    {
      "entropy": 5.43651967048645,
      "epoch": 0.7473583093179635,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004952158436953444,
      "loss": 5.2152,
      "mean_token_accuracy": 0.1948774352669716,
      "num_tokens": 17862897.0,
      "step": 7780
    },
    {
      "entropy": 5.3573570728302,
      "epoch": 0.7478386167146974,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004952088089372252,
      "loss": 5.2872,
      "mean_token_accuracy": 0.19403435289859772,
      "num_tokens": 17874235.0,
      "step": 7785
    },
    {
      "entropy": 5.35567569732666,
      "epoch": 0.7483189241114313,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004952017690665114,
      "loss": 5.302,
      "mean_token_accuracy": 0.19129539281129837,
      "num_tokens": 17885677.0,
      "step": 7790
    },
    {
      "entropy": 5.369703149795532,
      "epoch": 0.7487992315081652,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004951947240833664,
      "loss": 5.2056,
      "mean_token_accuracy": 0.19623667746782303,
      "num_tokens": 17896204.0,
      "step": 7795
    },
    {
      "entropy": 5.466663122177124,
      "epoch": 0.7492795389048992,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004951876739879537,
      "loss": 5.3103,
      "mean_token_accuracy": 0.19217686206102372,
      "num_tokens": 17908514.0,
      "step": 7800
    },
    {
      "entropy": 5.406633090972901,
      "epoch": 0.7497598463016331,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004951806187804371,
      "loss": 5.1962,
      "mean_token_accuracy": 0.20257627815008164,
      "num_tokens": 17918493.0,
      "step": 7805
    },
    {
      "entropy": 5.3837216854095455,
      "epoch": 0.7502401536983669,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004951735584609804,
      "loss": 5.2077,
      "mean_token_accuracy": 0.20404653251171112,
      "num_tokens": 17929016.0,
      "step": 7810
    },
    {
      "entropy": 5.27390456199646,
      "epoch": 0.7507204610951008,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004951664930297474,
      "loss": 5.266,
      "mean_token_accuracy": 0.1927507683634758,
      "num_tokens": 17940674.0,
      "step": 7815
    },
    {
      "entropy": 5.417548131942749,
      "epoch": 0.7512007684918348,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000495159422486902,
      "loss": 5.2948,
      "mean_token_accuracy": 0.19038532227277755,
      "num_tokens": 17951431.0,
      "step": 7820
    },
    {
      "entropy": 5.526469707489014,
      "epoch": 0.7516810758885687,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004951523468326088,
      "loss": 5.3227,
      "mean_token_accuracy": 0.18628415912389756,
      "num_tokens": 17961434.0,
      "step": 7825
    },
    {
      "entropy": 5.267174339294433,
      "epoch": 0.7521613832853026,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004951452660670317,
      "loss": 5.1524,
      "mean_token_accuracy": 0.20382616072893142,
      "num_tokens": 17971917.0,
      "step": 7830
    },
    {
      "entropy": 5.352742862701416,
      "epoch": 0.7526416906820365,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004951381801903352,
      "loss": 5.244,
      "mean_token_accuracy": 0.1986750155687332,
      "num_tokens": 17983754.0,
      "step": 7835
    },
    {
      "entropy": 5.343587589263916,
      "epoch": 0.7531219980787704,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004951310892026839,
      "loss": 5.2382,
      "mean_token_accuracy": 0.19234987646341323,
      "num_tokens": 17994088.0,
      "step": 7840
    },
    {
      "entropy": 5.373450660705567,
      "epoch": 0.7536023054755043,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004951239931042424,
      "loss": 5.2162,
      "mean_token_accuracy": 0.19949763566255568,
      "num_tokens": 18004847.0,
      "step": 7845
    },
    {
      "entropy": 5.313560152053833,
      "epoch": 0.7540826128722382,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004951168918951753,
      "loss": 5.2609,
      "mean_token_accuracy": 0.20220328867435455,
      "num_tokens": 18015756.0,
      "step": 7850
    },
    {
      "entropy": 5.335494136810302,
      "epoch": 0.7545629202689721,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004951097855756476,
      "loss": 5.3469,
      "mean_token_accuracy": 0.18886298090219497,
      "num_tokens": 18026417.0,
      "step": 7855
    },
    {
      "entropy": 5.381666135787964,
      "epoch": 0.7550432276657061,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004951026741458243,
      "loss": 5.2618,
      "mean_token_accuracy": 0.19316712468862535,
      "num_tokens": 18038205.0,
      "step": 7860
    },
    {
      "entropy": 5.433653688430786,
      "epoch": 0.75552353506244,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004950955576058705,
      "loss": 5.3012,
      "mean_token_accuracy": 0.1950031191110611,
      "num_tokens": 18049223.0,
      "step": 7865
    },
    {
      "entropy": 5.418736457824707,
      "epoch": 0.7560038424591738,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004950884359559513,
      "loss": 5.2724,
      "mean_token_accuracy": 0.19370467960834503,
      "num_tokens": 18061855.0,
      "step": 7870
    },
    {
      "entropy": 5.3388352394104,
      "epoch": 0.7564841498559077,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004950813091962324,
      "loss": 5.2662,
      "mean_token_accuracy": 0.19691956490278245,
      "num_tokens": 18073969.0,
      "step": 7875
    },
    {
      "entropy": 5.368044805526734,
      "epoch": 0.7569644572526417,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004950741773268788,
      "loss": 5.2018,
      "mean_token_accuracy": 0.19612170457839967,
      "num_tokens": 18085786.0,
      "step": 7880
    },
    {
      "entropy": 5.408827543258667,
      "epoch": 0.7574447646493756,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004950670403480562,
      "loss": 5.2151,
      "mean_token_accuracy": 0.1967058449983597,
      "num_tokens": 18096263.0,
      "step": 7885
    },
    {
      "entropy": 5.3796820640563965,
      "epoch": 0.7579250720461095,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004950598982599306,
      "loss": 5.3299,
      "mean_token_accuracy": 0.18323681950569154,
      "num_tokens": 18108769.0,
      "step": 7890
    },
    {
      "entropy": 5.384113502502442,
      "epoch": 0.7584053794428435,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004950527510626675,
      "loss": 5.3382,
      "mean_token_accuracy": 0.19338110387325286,
      "num_tokens": 18120255.0,
      "step": 7895
    },
    {
      "entropy": 5.396039915084839,
      "epoch": 0.7588856868395774,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004950455987564329,
      "loss": 5.2959,
      "mean_token_accuracy": 0.1955770120024681,
      "num_tokens": 18132016.0,
      "step": 7900
    },
    {
      "entropy": 5.34745602607727,
      "epoch": 0.7593659942363112,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004950384413413931,
      "loss": 5.2139,
      "mean_token_accuracy": 0.19548893719911575,
      "num_tokens": 18144101.0,
      "step": 7905
    },
    {
      "entropy": 5.4066825866699215,
      "epoch": 0.7598463016330451,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004950312788177139,
      "loss": 5.2254,
      "mean_token_accuracy": 0.19860927909612655,
      "num_tokens": 18155086.0,
      "step": 7910
    },
    {
      "entropy": 5.38053789138794,
      "epoch": 0.7603266090297791,
      "grad_norm": 3.34375,
      "learning_rate": 0.0004950241111855618,
      "loss": 5.2736,
      "mean_token_accuracy": 0.19188836514949797,
      "num_tokens": 18166902.0,
      "step": 7915
    },
    {
      "entropy": 5.398432874679566,
      "epoch": 0.760806916426513,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004950169384451031,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19742103517055512,
      "num_tokens": 18178258.0,
      "step": 7920
    },
    {
      "entropy": 5.36683759689331,
      "epoch": 0.7612872238232469,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004950097605965045,
      "loss": 5.2931,
      "mean_token_accuracy": 0.19071005284786224,
      "num_tokens": 18190580.0,
      "step": 7925
    },
    {
      "entropy": 5.421039676666259,
      "epoch": 0.7617675312199808,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004950025776399326,
      "loss": 5.3655,
      "mean_token_accuracy": 0.18787630051374435,
      "num_tokens": 18204912.0,
      "step": 7930
    },
    {
      "entropy": 5.378742361068726,
      "epoch": 0.7622478386167147,
      "grad_norm": 1.796875,
      "learning_rate": 0.000494995389575554,
      "loss": 5.2353,
      "mean_token_accuracy": 0.19148027151823044,
      "num_tokens": 18215129.0,
      "step": 7935
    },
    {
      "entropy": 5.417219114303589,
      "epoch": 0.7627281460134486,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004949881964035357,
      "loss": 5.3586,
      "mean_token_accuracy": 0.1875670626759529,
      "num_tokens": 18225640.0,
      "step": 7940
    },
    {
      "entropy": 5.377304887771606,
      "epoch": 0.7632084534101825,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004949809981240448,
      "loss": 5.2331,
      "mean_token_accuracy": 0.19860284626483918,
      "num_tokens": 18236641.0,
      "step": 7945
    },
    {
      "entropy": 5.315788221359253,
      "epoch": 0.7636887608069164,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004949737947372483,
      "loss": 5.2262,
      "mean_token_accuracy": 0.1945876345038414,
      "num_tokens": 18249244.0,
      "step": 7950
    },
    {
      "entropy": 5.4028219223022464,
      "epoch": 0.7641690682036504,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004949665862433134,
      "loss": 5.3117,
      "mean_token_accuracy": 0.1907445564866066,
      "num_tokens": 18261595.0,
      "step": 7955
    },
    {
      "entropy": 5.424557638168335,
      "epoch": 0.7646493756003843,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004949593726424077,
      "loss": 5.2758,
      "mean_token_accuracy": 0.19563941806554794,
      "num_tokens": 18272609.0,
      "step": 7960
    },
    {
      "entropy": 5.44644021987915,
      "epoch": 0.7651296829971181,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004949521539346985,
      "loss": 5.4205,
      "mean_token_accuracy": 0.19190972447395324,
      "num_tokens": 18283886.0,
      "step": 7965
    },
    {
      "entropy": 5.404463243484497,
      "epoch": 0.765609990393852,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004949449301203533,
      "loss": 5.3161,
      "mean_token_accuracy": 0.19465096443891525,
      "num_tokens": 18294794.0,
      "step": 7970
    },
    {
      "entropy": 5.495356655120849,
      "epoch": 0.766090297790586,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004949377011995399,
      "loss": 5.3151,
      "mean_token_accuracy": 0.19217901527881623,
      "num_tokens": 18306250.0,
      "step": 7975
    },
    {
      "entropy": 5.421238994598388,
      "epoch": 0.7665706051873199,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004949304671724263,
      "loss": 5.2938,
      "mean_token_accuracy": 0.19345854967832565,
      "num_tokens": 18317743.0,
      "step": 7980
    },
    {
      "entropy": 5.295167398452759,
      "epoch": 0.7670509125840538,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004949232280391802,
      "loss": 5.1327,
      "mean_token_accuracy": 0.19705990701913834,
      "num_tokens": 18328287.0,
      "step": 7985
    },
    {
      "entropy": 5.365210819244385,
      "epoch": 0.7675312199807877,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004949159837999698,
      "loss": 5.3502,
      "mean_token_accuracy": 0.18874591141939162,
      "num_tokens": 18341354.0,
      "step": 7990
    },
    {
      "entropy": 5.432317066192627,
      "epoch": 0.7680115273775217,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004949087344549633,
      "loss": 5.3434,
      "mean_token_accuracy": 0.1904577851295471,
      "num_tokens": 18353822.0,
      "step": 7995
    },
    {
      "entropy": 5.516468906402588,
      "epoch": 0.7684918347742555,
      "grad_norm": 1.5625,
      "learning_rate": 0.000494901480004329,
      "loss": 5.2823,
      "mean_token_accuracy": 0.1891055792570114,
      "num_tokens": 18365491.0,
      "step": 8000
    },
    {
      "entropy": 5.250404500961304,
      "epoch": 0.7689721421709894,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004948942204482351,
      "loss": 5.2232,
      "mean_token_accuracy": 0.19362736195325853,
      "num_tokens": 18376329.0,
      "step": 8005
    },
    {
      "entropy": 5.385899353027344,
      "epoch": 0.7694524495677233,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004948869557868506,
      "loss": 5.3128,
      "mean_token_accuracy": 0.18819482922554015,
      "num_tokens": 18387105.0,
      "step": 8010
    },
    {
      "entropy": 5.412772226333618,
      "epoch": 0.7699327569644573,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004948796860203439,
      "loss": 5.2686,
      "mean_token_accuracy": 0.19209769666194915,
      "num_tokens": 18398113.0,
      "step": 8015
    },
    {
      "entropy": 5.367074108123779,
      "epoch": 0.7704130643611912,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004948724111488838,
      "loss": 5.251,
      "mean_token_accuracy": 0.19539596289396285,
      "num_tokens": 18409315.0,
      "step": 8020
    },
    {
      "entropy": 5.292338895797729,
      "epoch": 0.770893371757925,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004948651311726391,
      "loss": 5.281,
      "mean_token_accuracy": 0.1888511836528778,
      "num_tokens": 18421110.0,
      "step": 8025
    },
    {
      "entropy": 5.429877138137817,
      "epoch": 0.7713736791546589,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004948578460917789,
      "loss": 5.3334,
      "mean_token_accuracy": 0.19012463837862015,
      "num_tokens": 18433145.0,
      "step": 8030
    },
    {
      "entropy": 5.47140908241272,
      "epoch": 0.7718539865513929,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004948505559064725,
      "loss": 5.2776,
      "mean_token_accuracy": 0.18811897039413453,
      "num_tokens": 18444689.0,
      "step": 8035
    },
    {
      "entropy": 5.337078332901001,
      "epoch": 0.7723342939481268,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004948432606168889,
      "loss": 5.2344,
      "mean_token_accuracy": 0.19592492282390594,
      "num_tokens": 18455943.0,
      "step": 8040
    },
    {
      "entropy": 5.2788911819458,
      "epoch": 0.7728146013448607,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004948359602231976,
      "loss": 5.1965,
      "mean_token_accuracy": 0.2019103229045868,
      "num_tokens": 18466763.0,
      "step": 8045
    },
    {
      "entropy": 5.392308759689331,
      "epoch": 0.7732949087415946,
      "grad_norm": 1.5,
      "learning_rate": 0.0004948286547255681,
      "loss": 5.3291,
      "mean_token_accuracy": 0.18780143260955812,
      "num_tokens": 18477647.0,
      "step": 8050
    },
    {
      "entropy": 5.381778860092163,
      "epoch": 0.7737752161383286,
      "grad_norm": 1.5625,
      "learning_rate": 0.00049482134412417,
      "loss": 5.1985,
      "mean_token_accuracy": 0.2007139027118683,
      "num_tokens": 18488465.0,
      "step": 8055
    },
    {
      "entropy": 5.4515125274658205,
      "epoch": 0.7742555235350624,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000494814028419173,
      "loss": 5.3597,
      "mean_token_accuracy": 0.19449682384729386,
      "num_tokens": 18500705.0,
      "step": 8060
    },
    {
      "entropy": 5.380729055404663,
      "epoch": 0.7747358309317963,
      "grad_norm": 1.578125,
      "learning_rate": 0.000494806707610747,
      "loss": 5.2376,
      "mean_token_accuracy": 0.19191479235887526,
      "num_tokens": 18512253.0,
      "step": 8065
    },
    {
      "entropy": 5.349826765060425,
      "epoch": 0.7752161383285303,
      "grad_norm": 1.390625,
      "learning_rate": 0.000494799381699062,
      "loss": 5.256,
      "mean_token_accuracy": 0.19469085335731506,
      "num_tokens": 18524231.0,
      "step": 8070
    },
    {
      "entropy": 5.452904129028321,
      "epoch": 0.7756964457252642,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004947920506842879,
      "loss": 5.3097,
      "mean_token_accuracy": 0.19417499154806137,
      "num_tokens": 18535824.0,
      "step": 8075
    },
    {
      "entropy": 5.4485352516174315,
      "epoch": 0.7761767531219981,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004947847145665951,
      "loss": 5.3774,
      "mean_token_accuracy": 0.19426287114620208,
      "num_tokens": 18547957.0,
      "step": 8080
    },
    {
      "entropy": 5.360448169708252,
      "epoch": 0.776657060518732,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004947773733461539,
      "loss": 5.2797,
      "mean_token_accuracy": 0.19322233349084855,
      "num_tokens": 18559297.0,
      "step": 8085
    },
    {
      "entropy": 5.406730127334595,
      "epoch": 0.777137367915466,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004947700270231347,
      "loss": 5.3415,
      "mean_token_accuracy": 0.19115198105573655,
      "num_tokens": 18570264.0,
      "step": 8090
    },
    {
      "entropy": 5.387463092803955,
      "epoch": 0.7776176753121998,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004947626755977079,
      "loss": 5.2309,
      "mean_token_accuracy": 0.19484462589025497,
      "num_tokens": 18582417.0,
      "step": 8095
    },
    {
      "entropy": 5.390276765823364,
      "epoch": 0.7780979827089337,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004947553190700444,
      "loss": 5.2764,
      "mean_token_accuracy": 0.1876293435692787,
      "num_tokens": 18593703.0,
      "step": 8100
    },
    {
      "entropy": 5.36275954246521,
      "epoch": 0.7785782901056676,
      "grad_norm": 1.546875,
      "learning_rate": 0.000494747957440315,
      "loss": 5.2321,
      "mean_token_accuracy": 0.19769291430711747,
      "num_tokens": 18605050.0,
      "step": 8105
    },
    {
      "entropy": 5.386284637451172,
      "epoch": 0.7790585975024016,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004947405907086905,
      "loss": 5.3259,
      "mean_token_accuracy": 0.18341726958751678,
      "num_tokens": 18616962.0,
      "step": 8110
    },
    {
      "entropy": 5.396425199508667,
      "epoch": 0.7795389048991355,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004947332188753419,
      "loss": 5.2686,
      "mean_token_accuracy": 0.19090291112661362,
      "num_tokens": 18628640.0,
      "step": 8115
    },
    {
      "entropy": 5.400382232666016,
      "epoch": 0.7800192122958693,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004947258419404405,
      "loss": 5.2946,
      "mean_token_accuracy": 0.19053932577371596,
      "num_tokens": 18640488.0,
      "step": 8120
    },
    {
      "entropy": 5.335408306121826,
      "epoch": 0.7804995196926032,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004947184599041576,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19515257328748703,
      "num_tokens": 18652095.0,
      "step": 8125
    },
    {
      "entropy": 5.381952476501465,
      "epoch": 0.7809798270893372,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004947110727666644,
      "loss": 5.2348,
      "mean_token_accuracy": 0.19744855910539627,
      "num_tokens": 18662767.0,
      "step": 8130
    },
    {
      "entropy": 5.389344978332519,
      "epoch": 0.7814601344860711,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004947036805281325,
      "loss": 5.3022,
      "mean_token_accuracy": 0.19188299775123596,
      "num_tokens": 18673568.0,
      "step": 8135
    },
    {
      "entropy": 5.387752056121826,
      "epoch": 0.781940441882805,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004946962831887336,
      "loss": 5.2306,
      "mean_token_accuracy": 0.20563167780637742,
      "num_tokens": 18684456.0,
      "step": 8140
    },
    {
      "entropy": 5.382151746749878,
      "epoch": 0.7824207492795389,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004946888807486393,
      "loss": 5.3096,
      "mean_token_accuracy": 0.18668509572744368,
      "num_tokens": 18695942.0,
      "step": 8145
    },
    {
      "entropy": 5.488183069229126,
      "epoch": 0.7829010566762729,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004946814732080214,
      "loss": 5.3408,
      "mean_token_accuracy": 0.19334076195955277,
      "num_tokens": 18708110.0,
      "step": 8150
    },
    {
      "entropy": 5.477972316741943,
      "epoch": 0.7833813640730067,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004946740605670523,
      "loss": 5.3308,
      "mean_token_accuracy": 0.19009837210178376,
      "num_tokens": 18718864.0,
      "step": 8155
    },
    {
      "entropy": 5.338687753677368,
      "epoch": 0.7838616714697406,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004946666428259037,
      "loss": 5.2142,
      "mean_token_accuracy": 0.19310434758663178,
      "num_tokens": 18730338.0,
      "step": 8160
    },
    {
      "entropy": 5.368350887298584,
      "epoch": 0.7843419788664745,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004946592199847478,
      "loss": 5.3917,
      "mean_token_accuracy": 0.18230630159378053,
      "num_tokens": 18743433.0,
      "step": 8165
    },
    {
      "entropy": 5.415580177307129,
      "epoch": 0.7848222862632085,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004946517920437571,
      "loss": 5.2675,
      "mean_token_accuracy": 0.19327433109283448,
      "num_tokens": 18754631.0,
      "step": 8170
    },
    {
      "entropy": 5.390597867965698,
      "epoch": 0.7853025936599424,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004946443590031041,
      "loss": 5.2598,
      "mean_token_accuracy": 0.19704635739326476,
      "num_tokens": 18766118.0,
      "step": 8175
    },
    {
      "entropy": 5.418410730361939,
      "epoch": 0.7857829010566763,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004946369208629613,
      "loss": 5.2984,
      "mean_token_accuracy": 0.1961783766746521,
      "num_tokens": 18777743.0,
      "step": 8180
    },
    {
      "entropy": 5.338965368270874,
      "epoch": 0.7862632084534101,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004946294776235013,
      "loss": 5.2457,
      "mean_token_accuracy": 0.19982785433530809,
      "num_tokens": 18789684.0,
      "step": 8185
    },
    {
      "entropy": 5.253990411758423,
      "epoch": 0.7867435158501441,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004946220292848971,
      "loss": 5.2156,
      "mean_token_accuracy": 0.19799993932247162,
      "num_tokens": 18801088.0,
      "step": 8190
    },
    {
      "entropy": 5.375244569778443,
      "epoch": 0.787223823246878,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004946145758473214,
      "loss": 5.3143,
      "mean_token_accuracy": 0.1825365826487541,
      "num_tokens": 18812478.0,
      "step": 8195
    },
    {
      "entropy": 5.544604301452637,
      "epoch": 0.7877041306436119,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004946071173109475,
      "loss": 5.3192,
      "mean_token_accuracy": 0.1940308019518852,
      "num_tokens": 18823573.0,
      "step": 8200
    },
    {
      "entropy": 5.3184661865234375,
      "epoch": 0.7881844380403458,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004945996536759484,
      "loss": 5.2085,
      "mean_token_accuracy": 0.19500154703855516,
      "num_tokens": 18833880.0,
      "step": 8205
    },
    {
      "entropy": 5.326853466033936,
      "epoch": 0.7886647454370798,
      "grad_norm": 2.71875,
      "learning_rate": 0.0004945921849424974,
      "loss": 5.1911,
      "mean_token_accuracy": 0.19733449518680574,
      "num_tokens": 18844367.0,
      "step": 8210
    },
    {
      "entropy": 5.410817670822143,
      "epoch": 0.7891450528338136,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004945847111107679,
      "loss": 5.2677,
      "mean_token_accuracy": 0.18654292076826096,
      "num_tokens": 18856309.0,
      "step": 8215
    },
    {
      "entropy": 5.408186960220337,
      "epoch": 0.7896253602305475,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004945772321809334,
      "loss": 5.2349,
      "mean_token_accuracy": 0.19013492614030839,
      "num_tokens": 18868873.0,
      "step": 8220
    },
    {
      "entropy": 5.32963228225708,
      "epoch": 0.7901056676272814,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004945697481531677,
      "loss": 5.1909,
      "mean_token_accuracy": 0.19453502595424652,
      "num_tokens": 18879953.0,
      "step": 8225
    },
    {
      "entropy": 5.444918251037597,
      "epoch": 0.7905859750240154,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004945622590276443,
      "loss": 5.3386,
      "mean_token_accuracy": 0.18775013834238052,
      "num_tokens": 18892220.0,
      "step": 8230
    },
    {
      "entropy": 5.422673940658569,
      "epoch": 0.7910662824207493,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004945547648045373,
      "loss": 5.2638,
      "mean_token_accuracy": 0.19337849467992782,
      "num_tokens": 18903261.0,
      "step": 8235
    },
    {
      "entropy": 5.369199991226196,
      "epoch": 0.7915465898174832,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004945472654840206,
      "loss": 5.2883,
      "mean_token_accuracy": 0.19510689675807952,
      "num_tokens": 18913702.0,
      "step": 8240
    },
    {
      "entropy": 5.452567434310913,
      "epoch": 0.7920268972142172,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004945397610662683,
      "loss": 5.3378,
      "mean_token_accuracy": 0.18533655554056166,
      "num_tokens": 18925024.0,
      "step": 8245
    },
    {
      "entropy": 5.383784341812134,
      "epoch": 0.792507204610951,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004945322515514547,
      "loss": 5.2403,
      "mean_token_accuracy": 0.2047622248530388,
      "num_tokens": 18935939.0,
      "step": 8250
    },
    {
      "entropy": 5.400283718109131,
      "epoch": 0.7929875120076849,
      "grad_norm": 1.5,
      "learning_rate": 0.000494524736939754,
      "loss": 5.2076,
      "mean_token_accuracy": 0.19320725947618483,
      "num_tokens": 18946091.0,
      "step": 8255
    },
    {
      "entropy": 5.474302816390991,
      "epoch": 0.7934678194044188,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004945172172313408,
      "loss": 5.3195,
      "mean_token_accuracy": 0.18396852761507035,
      "num_tokens": 18957471.0,
      "step": 8260
    },
    {
      "entropy": 5.493697547912598,
      "epoch": 0.7939481268011528,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004945096924263896,
      "loss": 5.4359,
      "mean_token_accuracy": 0.1846723437309265,
      "num_tokens": 18969462.0,
      "step": 8265
    },
    {
      "entropy": 5.42394528388977,
      "epoch": 0.7944284341978867,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004945021625250753,
      "loss": 5.2792,
      "mean_token_accuracy": 0.18909084796905518,
      "num_tokens": 18980228.0,
      "step": 8270
    },
    {
      "entropy": 5.46893630027771,
      "epoch": 0.7949087415946205,
      "grad_norm": 1.875,
      "learning_rate": 0.0004944946275275724,
      "loss": 5.4214,
      "mean_token_accuracy": 0.18460303097963332,
      "num_tokens": 18991661.0,
      "step": 8275
    },
    {
      "entropy": 5.4125611782073975,
      "epoch": 0.7953890489913544,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004944870874340561,
      "loss": 5.2921,
      "mean_token_accuracy": 0.19383617490530014,
      "num_tokens": 19003082.0,
      "step": 8280
    },
    {
      "entropy": 5.408040761947632,
      "epoch": 0.7958693563880884,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004944795422447013,
      "loss": 5.1908,
      "mean_token_accuracy": 0.19174865037202835,
      "num_tokens": 19015408.0,
      "step": 8285
    },
    {
      "entropy": 5.444769048690796,
      "epoch": 0.7963496637848223,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004944719919596835,
      "loss": 5.2547,
      "mean_token_accuracy": 0.19427991807460784,
      "num_tokens": 19027246.0,
      "step": 8290
    },
    {
      "entropy": 5.418350124359131,
      "epoch": 0.7968299711815562,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004944644365791776,
      "loss": 5.264,
      "mean_token_accuracy": 0.20031838566064836,
      "num_tokens": 19038126.0,
      "step": 8295
    },
    {
      "entropy": 5.32525634765625,
      "epoch": 0.7973102785782901,
      "grad_norm": 2.59375,
      "learning_rate": 0.000494456876103359,
      "loss": 5.2045,
      "mean_token_accuracy": 0.20074736177921296,
      "num_tokens": 19047998.0,
      "step": 8300
    },
    {
      "entropy": 5.313124275207519,
      "epoch": 0.7977905859750241,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004944493105324035,
      "loss": 5.2991,
      "mean_token_accuracy": 0.18852553218603135,
      "num_tokens": 19058969.0,
      "step": 8305
    },
    {
      "entropy": 5.351770830154419,
      "epoch": 0.7982708933717579,
      "grad_norm": 2.953125,
      "learning_rate": 0.0004944417398664866,
      "loss": 5.2187,
      "mean_token_accuracy": 0.19643116593360901,
      "num_tokens": 19070314.0,
      "step": 8310
    },
    {
      "entropy": 5.420844650268554,
      "epoch": 0.7987512007684918,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004944341641057843,
      "loss": 5.2765,
      "mean_token_accuracy": 0.19828618466854095,
      "num_tokens": 19082575.0,
      "step": 8315
    },
    {
      "entropy": 5.296345329284668,
      "epoch": 0.7992315081652257,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004944265832504721,
      "loss": 5.2411,
      "mean_token_accuracy": 0.19786602556705474,
      "num_tokens": 19095719.0,
      "step": 8320
    },
    {
      "entropy": 5.383054637908936,
      "epoch": 0.7997118155619597,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004944189973007262,
      "loss": 5.2203,
      "mean_token_accuracy": 0.19829194098711014,
      "num_tokens": 19106894.0,
      "step": 8325
    },
    {
      "entropy": 5.3365484237670895,
      "epoch": 0.8001921229586936,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004944114062567229,
      "loss": 5.241,
      "mean_token_accuracy": 0.19062076658010482,
      "num_tokens": 19118206.0,
      "step": 8330
    },
    {
      "entropy": 5.404685306549072,
      "epoch": 0.8006724303554275,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004944038101186381,
      "loss": 5.3116,
      "mean_token_accuracy": 0.18689830750226974,
      "num_tokens": 19129613.0,
      "step": 8335
    },
    {
      "entropy": 5.385619306564331,
      "epoch": 0.8011527377521613,
      "grad_norm": 2.90625,
      "learning_rate": 0.0004943962088866483,
      "loss": 5.2262,
      "mean_token_accuracy": 0.19819785803556442,
      "num_tokens": 19142214.0,
      "step": 8340
    },
    {
      "entropy": 5.403306579589843,
      "epoch": 0.8016330451488953,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004943886025609301,
      "loss": 5.3302,
      "mean_token_accuracy": 0.1897999107837677,
      "num_tokens": 19153906.0,
      "step": 8345
    },
    {
      "entropy": 5.396410608291626,
      "epoch": 0.8021133525456292,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00049438099114166,
      "loss": 5.2322,
      "mean_token_accuracy": 0.18854821920394899,
      "num_tokens": 19163677.0,
      "step": 8350
    },
    {
      "entropy": 5.400742244720459,
      "epoch": 0.8025936599423631,
      "grad_norm": 1.375,
      "learning_rate": 0.0004943733746290147,
      "loss": 5.2909,
      "mean_token_accuracy": 0.1931081086397171,
      "num_tokens": 19175240.0,
      "step": 8355
    },
    {
      "entropy": 5.3174515724182125,
      "epoch": 0.803073967339097,
      "grad_norm": 1.46875,
      "learning_rate": 0.000494365753023171,
      "loss": 5.1083,
      "mean_token_accuracy": 0.20141686052083968,
      "num_tokens": 19185529.0,
      "step": 8360
    },
    {
      "entropy": 5.325999212265015,
      "epoch": 0.803554274735831,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004943581263243059,
      "loss": 5.2215,
      "mean_token_accuracy": 0.19571770280599593,
      "num_tokens": 19197153.0,
      "step": 8365
    },
    {
      "entropy": 5.396734189987183,
      "epoch": 0.8040345821325648,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004943504945325965,
      "loss": 5.3914,
      "mean_token_accuracy": 0.18932325541973113,
      "num_tokens": 19208439.0,
      "step": 8370
    },
    {
      "entropy": 5.351518821716309,
      "epoch": 0.8045148895292987,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004943428576482198,
      "loss": 5.2015,
      "mean_token_accuracy": 0.20140155106782914,
      "num_tokens": 19218570.0,
      "step": 8375
    },
    {
      "entropy": 5.352016496658325,
      "epoch": 0.8049951969260326,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004943352156713535,
      "loss": 5.2293,
      "mean_token_accuracy": 0.19814103543758393,
      "num_tokens": 19229080.0,
      "step": 8380
    },
    {
      "entropy": 5.401249790191651,
      "epoch": 0.8054755043227666,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004943275686021747,
      "loss": 5.3164,
      "mean_token_accuracy": 0.1952598959207535,
      "num_tokens": 19239299.0,
      "step": 8385
    },
    {
      "entropy": 5.423311853408814,
      "epoch": 0.8059558117195005,
      "grad_norm": 1.265625,
      "learning_rate": 0.000494319916440861,
      "loss": 5.2486,
      "mean_token_accuracy": 0.19928672462701796,
      "num_tokens": 19250691.0,
      "step": 8390
    },
    {
      "entropy": 5.361232423782349,
      "epoch": 0.8064361191162344,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004943122591875901,
      "loss": 5.233,
      "mean_token_accuracy": 0.20244060605764388,
      "num_tokens": 19260868.0,
      "step": 8395
    },
    {
      "entropy": 5.436623811721802,
      "epoch": 0.8069164265129684,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004943045968425398,
      "loss": 5.3237,
      "mean_token_accuracy": 0.19038853496313096,
      "num_tokens": 19271924.0,
      "step": 8400
    },
    {
      "entropy": 5.3783989429473875,
      "epoch": 0.8073967339097022,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004942969294058878,
      "loss": 5.2837,
      "mean_token_accuracy": 0.1993526339530945,
      "num_tokens": 19283340.0,
      "step": 8405
    },
    {
      "entropy": 5.4472033977508545,
      "epoch": 0.8078770413064361,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004942892568778125,
      "loss": 5.333,
      "mean_token_accuracy": 0.18856506198644638,
      "num_tokens": 19294903.0,
      "step": 8410
    },
    {
      "entropy": 5.428958940505981,
      "epoch": 0.80835734870317,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004942815792584917,
      "loss": 5.2306,
      "mean_token_accuracy": 0.1962373971939087,
      "num_tokens": 19305670.0,
      "step": 8415
    },
    {
      "entropy": 5.376930856704712,
      "epoch": 0.808837656099904,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004942738965481038,
      "loss": 5.2326,
      "mean_token_accuracy": 0.19721640646457672,
      "num_tokens": 19317043.0,
      "step": 8420
    },
    {
      "entropy": 5.397046899795532,
      "epoch": 0.8093179634966379,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004942662087468272,
      "loss": 5.3032,
      "mean_token_accuracy": 0.19107830971479417,
      "num_tokens": 19329161.0,
      "step": 8425
    },
    {
      "entropy": 5.389262008666992,
      "epoch": 0.8097982708933718,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004942585158548402,
      "loss": 5.3074,
      "mean_token_accuracy": 0.19399819672107696,
      "num_tokens": 19340706.0,
      "step": 8430
    },
    {
      "entropy": 5.439792919158935,
      "epoch": 0.8102785782901056,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004942508178723214,
      "loss": 5.2748,
      "mean_token_accuracy": 0.1924287885427475,
      "num_tokens": 19352422.0,
      "step": 8435
    },
    {
      "entropy": 5.402951288223266,
      "epoch": 0.8107588856868396,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004942431147994499,
      "loss": 5.3106,
      "mean_token_accuracy": 0.1888471707701683,
      "num_tokens": 19364109.0,
      "step": 8440
    },
    {
      "entropy": 5.383220434188843,
      "epoch": 0.8112391930835735,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004942354066364042,
      "loss": 5.3112,
      "mean_token_accuracy": 0.18616687655448913,
      "num_tokens": 19377386.0,
      "step": 8445
    },
    {
      "entropy": 5.379947185516357,
      "epoch": 0.8117195004803074,
      "grad_norm": 2.75,
      "learning_rate": 0.0004942276933833634,
      "loss": 5.2217,
      "mean_token_accuracy": 0.20263410955667496,
      "num_tokens": 19388453.0,
      "step": 8450
    },
    {
      "entropy": 5.366377592086792,
      "epoch": 0.8121998078770413,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004942199750405064,
      "loss": 5.3665,
      "mean_token_accuracy": 0.18946326971054078,
      "num_tokens": 19402240.0,
      "step": 8455
    },
    {
      "entropy": 5.389968967437744,
      "epoch": 0.8126801152737753,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004942122516080127,
      "loss": 5.2675,
      "mean_token_accuracy": 0.19347688406705857,
      "num_tokens": 19414260.0,
      "step": 8460
    },
    {
      "entropy": 5.3555137634277346,
      "epoch": 0.8131604226705091,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004942045230860614,
      "loss": 5.207,
      "mean_token_accuracy": 0.1945435658097267,
      "num_tokens": 19426356.0,
      "step": 8465
    },
    {
      "entropy": 5.363530683517456,
      "epoch": 0.813640730067243,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004941967894748319,
      "loss": 5.2972,
      "mean_token_accuracy": 0.19305806905031203,
      "num_tokens": 19438881.0,
      "step": 8470
    },
    {
      "entropy": 5.392398405075073,
      "epoch": 0.8141210374639769,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004941890507745039,
      "loss": 5.2344,
      "mean_token_accuracy": 0.19344445765018464,
      "num_tokens": 19450109.0,
      "step": 8475
    },
    {
      "entropy": 5.39263424873352,
      "epoch": 0.8146013448607109,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004941813069852569,
      "loss": 5.2411,
      "mean_token_accuracy": 0.19751955568790436,
      "num_tokens": 19461155.0,
      "step": 8480
    },
    {
      "entropy": 5.456835508346558,
      "epoch": 0.8150816522574448,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004941735581072708,
      "loss": 5.4364,
      "mean_token_accuracy": 0.1894826352596283,
      "num_tokens": 19474364.0,
      "step": 8485
    },
    {
      "entropy": 5.3985669136047365,
      "epoch": 0.8155619596541787,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004941658041407255,
      "loss": 5.2159,
      "mean_token_accuracy": 0.20274985879659652,
      "num_tokens": 19485271.0,
      "step": 8490
    },
    {
      "entropy": 5.444395685195923,
      "epoch": 0.8160422670509125,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000494158045085801,
      "loss": 5.4036,
      "mean_token_accuracy": 0.18302276730537415,
      "num_tokens": 19497117.0,
      "step": 8495
    },
    {
      "entropy": 5.389369535446167,
      "epoch": 0.8165225744476465,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004941502809426776,
      "loss": 5.1711,
      "mean_token_accuracy": 0.20128502696752548,
      "num_tokens": 19507035.0,
      "step": 8500
    },
    {
      "entropy": 5.2925615310668945,
      "epoch": 0.8170028818443804,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004941425117115354,
      "loss": 5.1753,
      "mean_token_accuracy": 0.19550191462039948,
      "num_tokens": 19517791.0,
      "step": 8505
    },
    {
      "entropy": 5.3593018531799315,
      "epoch": 0.8174831892411143,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004941347373925547,
      "loss": 5.2527,
      "mean_token_accuracy": 0.19915928542613984,
      "num_tokens": 19529577.0,
      "step": 8510
    },
    {
      "entropy": 5.476502704620361,
      "epoch": 0.8179634966378482,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004941269579859161,
      "loss": 5.3547,
      "mean_token_accuracy": 0.1887993961572647,
      "num_tokens": 19542355.0,
      "step": 8515
    },
    {
      "entropy": 5.419625377655029,
      "epoch": 0.8184438040345822,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004941191734918002,
      "loss": 5.3246,
      "mean_token_accuracy": 0.18974844515323638,
      "num_tokens": 19553819.0,
      "step": 8520
    },
    {
      "entropy": 5.424053096771241,
      "epoch": 0.818924111431316,
      "grad_norm": 1.734375,
      "learning_rate": 0.000494111383910388,
      "loss": 5.3341,
      "mean_token_accuracy": 0.18588172793388366,
      "num_tokens": 19565816.0,
      "step": 8525
    },
    {
      "entropy": 5.4313558578491214,
      "epoch": 0.8194044188280499,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004941035892418597,
      "loss": 5.2489,
      "mean_token_accuracy": 0.19463450312614441,
      "num_tokens": 19577991.0,
      "step": 8530
    },
    {
      "entropy": 5.353194999694824,
      "epoch": 0.8198847262247838,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004940957894863968,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1920216679573059,
      "num_tokens": 19590225.0,
      "step": 8535
    },
    {
      "entropy": 5.378624725341797,
      "epoch": 0.8203650336215178,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004940879846441804,
      "loss": 5.2849,
      "mean_token_accuracy": 0.19345015585422515,
      "num_tokens": 19602966.0,
      "step": 8540
    },
    {
      "entropy": 5.468559217453003,
      "epoch": 0.8208453410182517,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004940801747153914,
      "loss": 5.2433,
      "mean_token_accuracy": 0.1942879244685173,
      "num_tokens": 19613148.0,
      "step": 8545
    },
    {
      "entropy": 5.267355346679688,
      "epoch": 0.8213256484149856,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004940723597002113,
      "loss": 5.1297,
      "mean_token_accuracy": 0.20505749732255935,
      "num_tokens": 19624361.0,
      "step": 8550
    },
    {
      "entropy": 5.304432725906372,
      "epoch": 0.8218059558117194,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004940645395988216,
      "loss": 5.2346,
      "mean_token_accuracy": 0.19914115071296692,
      "num_tokens": 19635915.0,
      "step": 8555
    },
    {
      "entropy": 5.390110921859741,
      "epoch": 0.8222862632084534,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004940567144114036,
      "loss": 5.2868,
      "mean_token_accuracy": 0.18822350800037385,
      "num_tokens": 19646701.0,
      "step": 8560
    },
    {
      "entropy": 5.456581497192383,
      "epoch": 0.8227665706051873,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004940488841381393,
      "loss": 5.3502,
      "mean_token_accuracy": 0.18938471674919127,
      "num_tokens": 19657804.0,
      "step": 8565
    },
    {
      "entropy": 5.45032377243042,
      "epoch": 0.8232468780019212,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004940410487792103,
      "loss": 5.3673,
      "mean_token_accuracy": 0.1936648204922676,
      "num_tokens": 19670202.0,
      "step": 8570
    },
    {
      "entropy": 5.258342742919922,
      "epoch": 0.8237271853986552,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004940332083347986,
      "loss": 5.1382,
      "mean_token_accuracy": 0.20448355078697206,
      "num_tokens": 19681362.0,
      "step": 8575
    },
    {
      "entropy": 5.31924614906311,
      "epoch": 0.8242074927953891,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004940253628050861,
      "loss": 5.3096,
      "mean_token_accuracy": 0.19031556695699692,
      "num_tokens": 19693429.0,
      "step": 8580
    },
    {
      "entropy": 5.453513336181641,
      "epoch": 0.824687800192123,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004940175121902552,
      "loss": 5.3218,
      "mean_token_accuracy": 0.1932925894856453,
      "num_tokens": 19704667.0,
      "step": 8585
    },
    {
      "entropy": 5.419979238510132,
      "epoch": 0.8251681075888568,
      "grad_norm": 1.8125,
      "learning_rate": 0.000494009656490488,
      "loss": 5.2817,
      "mean_token_accuracy": 0.19172892719507217,
      "num_tokens": 19716002.0,
      "step": 8590
    },
    {
      "entropy": 5.370490264892578,
      "epoch": 0.8256484149855908,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004940017957059668,
      "loss": 5.2847,
      "mean_token_accuracy": 0.1934389740228653,
      "num_tokens": 19726626.0,
      "step": 8595
    },
    {
      "entropy": 5.364405536651612,
      "epoch": 0.8261287223823247,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004939939298368742,
      "loss": 5.2575,
      "mean_token_accuracy": 0.19498982280492783,
      "num_tokens": 19738250.0,
      "step": 8600
    },
    {
      "entropy": 5.427344703674317,
      "epoch": 0.8266090297790586,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004939860588833929,
      "loss": 5.3453,
      "mean_token_accuracy": 0.19303781539201736,
      "num_tokens": 19750119.0,
      "step": 8605
    },
    {
      "entropy": 5.425992107391357,
      "epoch": 0.8270893371757925,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004939781828457055,
      "loss": 5.3114,
      "mean_token_accuracy": 0.1902000978589058,
      "num_tokens": 19761709.0,
      "step": 8610
    },
    {
      "entropy": 5.406402587890625,
      "epoch": 0.8275696445725265,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000493970301723995,
      "loss": 5.2929,
      "mean_token_accuracy": 0.1903778627514839,
      "num_tokens": 19773562.0,
      "step": 8615
    },
    {
      "entropy": 5.515379476547241,
      "epoch": 0.8280499519692603,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004939624155184443,
      "loss": 5.3204,
      "mean_token_accuracy": 0.187325219810009,
      "num_tokens": 19785327.0,
      "step": 8620
    },
    {
      "entropy": 5.4205245018005375,
      "epoch": 0.8285302593659942,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004939545242292365,
      "loss": 5.3267,
      "mean_token_accuracy": 0.1893976628780365,
      "num_tokens": 19796760.0,
      "step": 8625
    },
    {
      "entropy": 5.360515213012695,
      "epoch": 0.8290105667627281,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004939466278565547,
      "loss": 5.2477,
      "mean_token_accuracy": 0.19780270755290985,
      "num_tokens": 19807271.0,
      "step": 8630
    },
    {
      "entropy": 5.357036304473877,
      "epoch": 0.8294908741594621,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004939387264005825,
      "loss": 5.2748,
      "mean_token_accuracy": 0.1989939898252487,
      "num_tokens": 19818098.0,
      "step": 8635
    },
    {
      "entropy": 5.3540637493133545,
      "epoch": 0.829971181556196,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004939308198615031,
      "loss": 5.2591,
      "mean_token_accuracy": 0.19724859446287155,
      "num_tokens": 19829482.0,
      "step": 8640
    },
    {
      "entropy": 5.438948297500611,
      "epoch": 0.8304514889529299,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004939229082395001,
      "loss": 5.3148,
      "mean_token_accuracy": 0.19536018073558808,
      "num_tokens": 19840883.0,
      "step": 8645
    },
    {
      "entropy": 5.320067071914673,
      "epoch": 0.8309317963496637,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004939149915347573,
      "loss": 5.191,
      "mean_token_accuracy": 0.19495573341846467,
      "num_tokens": 19852619.0,
      "step": 8650
    },
    {
      "entropy": 5.270103025436401,
      "epoch": 0.8314121037463977,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004939070697474585,
      "loss": 5.1612,
      "mean_token_accuracy": 0.199944931268692,
      "num_tokens": 19863316.0,
      "step": 8655
    },
    {
      "entropy": 5.403096675872803,
      "epoch": 0.8318924111431316,
      "grad_norm": 1.75,
      "learning_rate": 0.0004938991428777875,
      "loss": 5.3361,
      "mean_token_accuracy": 0.19362263530492782,
      "num_tokens": 19873965.0,
      "step": 8660
    },
    {
      "entropy": 5.395878410339355,
      "epoch": 0.8323727185398655,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004938912109259284,
      "loss": 5.2606,
      "mean_token_accuracy": 0.194809927046299,
      "num_tokens": 19884685.0,
      "step": 8665
    },
    {
      "entropy": 5.356817007064819,
      "epoch": 0.8328530259365994,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004938832738920654,
      "loss": 5.2183,
      "mean_token_accuracy": 0.1967201828956604,
      "num_tokens": 19897594.0,
      "step": 8670
    },
    {
      "entropy": 5.388972520828247,
      "epoch": 0.8333333333333334,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004938753317763826,
      "loss": 5.312,
      "mean_token_accuracy": 0.19195151031017305,
      "num_tokens": 19908093.0,
      "step": 8675
    },
    {
      "entropy": 5.451654005050659,
      "epoch": 0.8338136407300673,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004938673845790646,
      "loss": 5.3454,
      "mean_token_accuracy": 0.18652617633342744,
      "num_tokens": 19920502.0,
      "step": 8680
    },
    {
      "entropy": 5.36719446182251,
      "epoch": 0.8342939481268011,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004938594323002957,
      "loss": 5.1712,
      "mean_token_accuracy": 0.20161306709051133,
      "num_tokens": 19931437.0,
      "step": 8685
    },
    {
      "entropy": 5.3172218799591064,
      "epoch": 0.834774255523535,
      "grad_norm": 1.5,
      "learning_rate": 0.0004938514749402609,
      "loss": 5.2507,
      "mean_token_accuracy": 0.19911665171384813,
      "num_tokens": 19942592.0,
      "step": 8690
    },
    {
      "entropy": 5.421259260177612,
      "epoch": 0.835254562920269,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004938435124991447,
      "loss": 5.4225,
      "mean_token_accuracy": 0.1849104180932045,
      "num_tokens": 19953802.0,
      "step": 8695
    },
    {
      "entropy": 5.400475645065308,
      "epoch": 0.8357348703170029,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004938355449771318,
      "loss": 5.2509,
      "mean_token_accuracy": 0.19384113550186158,
      "num_tokens": 19964686.0,
      "step": 8700
    },
    {
      "entropy": 5.373258304595947,
      "epoch": 0.8362151777137368,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004938275723744075,
      "loss": 5.3036,
      "mean_token_accuracy": 0.19449401795864105,
      "num_tokens": 19975603.0,
      "step": 8705
    },
    {
      "entropy": 5.442459774017334,
      "epoch": 0.8366954851104706,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004938195946911567,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18805384486913682,
      "num_tokens": 19986205.0,
      "step": 8710
    },
    {
      "entropy": 5.398763799667359,
      "epoch": 0.8371757925072046,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004938116119275645,
      "loss": 5.1853,
      "mean_token_accuracy": 0.20078344643115997,
      "num_tokens": 19997323.0,
      "step": 8715
    },
    {
      "entropy": 5.3614967346191404,
      "epoch": 0.8376560999039385,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004938036240838166,
      "loss": 5.2386,
      "mean_token_accuracy": 0.2019427612423897,
      "num_tokens": 20007334.0,
      "step": 8720
    },
    {
      "entropy": 5.4310449123382565,
      "epoch": 0.8381364073006724,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004937956311600983,
      "loss": 5.2352,
      "mean_token_accuracy": 0.1901556819677353,
      "num_tokens": 20017476.0,
      "step": 8725
    },
    {
      "entropy": 5.285951900482178,
      "epoch": 0.8386167146974063,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004937876331565951,
      "loss": 5.2168,
      "mean_token_accuracy": 0.19222691059112548,
      "num_tokens": 20029323.0,
      "step": 8730
    },
    {
      "entropy": 5.318617486953736,
      "epoch": 0.8390970220941403,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004937796300734926,
      "loss": 5.2203,
      "mean_token_accuracy": 0.18935052901506425,
      "num_tokens": 20040317.0,
      "step": 8735
    },
    {
      "entropy": 5.3973370552062985,
      "epoch": 0.8395773294908742,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004937716219109769,
      "loss": 5.272,
      "mean_token_accuracy": 0.19618357717990875,
      "num_tokens": 20050495.0,
      "step": 8740
    },
    {
      "entropy": 5.461271047592163,
      "epoch": 0.840057636887608,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004937636086692336,
      "loss": 5.3319,
      "mean_token_accuracy": 0.18598054796457292,
      "num_tokens": 20061185.0,
      "step": 8745
    },
    {
      "entropy": 5.414768648147583,
      "epoch": 0.840537944284342,
      "grad_norm": 1.3125,
      "learning_rate": 0.000493755590348449,
      "loss": 5.2867,
      "mean_token_accuracy": 0.19658587425947188,
      "num_tokens": 20073746.0,
      "step": 8750
    },
    {
      "entropy": 5.4302146434783936,
      "epoch": 0.8410182516810759,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004937475669488091,
      "loss": 5.3104,
      "mean_token_accuracy": 0.19282436221837998,
      "num_tokens": 20085667.0,
      "step": 8755
    },
    {
      "entropy": 5.411467552185059,
      "epoch": 0.8414985590778098,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004937395384705004,
      "loss": 5.1906,
      "mean_token_accuracy": 0.204672409594059,
      "num_tokens": 20096017.0,
      "step": 8760
    },
    {
      "entropy": 5.334100961685181,
      "epoch": 0.8419788664745437,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004937315049137089,
      "loss": 5.2675,
      "mean_token_accuracy": 0.19416530877351762,
      "num_tokens": 20106928.0,
      "step": 8765
    },
    {
      "entropy": 5.328470897674561,
      "epoch": 0.8424591738712777,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004937234662786216,
      "loss": 5.2577,
      "mean_token_accuracy": 0.19781202226877212,
      "num_tokens": 20118768.0,
      "step": 8770
    },
    {
      "entropy": 5.493528699874878,
      "epoch": 0.8429394812680115,
      "grad_norm": 2.875,
      "learning_rate": 0.0004937154225654246,
      "loss": 5.374,
      "mean_token_accuracy": 0.18760730177164078,
      "num_tokens": 20129009.0,
      "step": 8775
    },
    {
      "entropy": 5.406347799301147,
      "epoch": 0.8434197886647454,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004937073737743051,
      "loss": 5.2736,
      "mean_token_accuracy": 0.19011209011077881,
      "num_tokens": 20141577.0,
      "step": 8780
    },
    {
      "entropy": 5.4522148132324215,
      "epoch": 0.8439000960614793,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004936993199054499,
      "loss": 5.344,
      "mean_token_accuracy": 0.1947518765926361,
      "num_tokens": 20153636.0,
      "step": 8785
    },
    {
      "entropy": 5.335200929641724,
      "epoch": 0.8443804034582133,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004936912609590458,
      "loss": 5.2141,
      "mean_token_accuracy": 0.19758563488721848,
      "num_tokens": 20165719.0,
      "step": 8790
    },
    {
      "entropy": 5.370564079284668,
      "epoch": 0.8448607108549472,
      "grad_norm": 1.640625,
      "learning_rate": 0.00049368319693528,
      "loss": 5.302,
      "mean_token_accuracy": 0.18864322751760482,
      "num_tokens": 20178260.0,
      "step": 8795
    },
    {
      "entropy": 5.4461256980896,
      "epoch": 0.8453410182516811,
      "grad_norm": 1.625,
      "learning_rate": 0.0004936751278343397,
      "loss": 5.301,
      "mean_token_accuracy": 0.1917430117726326,
      "num_tokens": 20189667.0,
      "step": 8800
    },
    {
      "entropy": 5.4010473728179935,
      "epoch": 0.845821325648415,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004936670536564123,
      "loss": 5.2913,
      "mean_token_accuracy": 0.18697932064533235,
      "num_tokens": 20201451.0,
      "step": 8805
    },
    {
      "entropy": 5.4120279312133786,
      "epoch": 0.8463016330451489,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004936589744016853,
      "loss": 5.3183,
      "mean_token_accuracy": 0.19138943105936052,
      "num_tokens": 20213257.0,
      "step": 8810
    },
    {
      "entropy": 5.443713808059693,
      "epoch": 0.8467819404418828,
      "grad_norm": 1.515625,
      "learning_rate": 0.000493650890070346,
      "loss": 5.2953,
      "mean_token_accuracy": 0.19443745017051697,
      "num_tokens": 20225301.0,
      "step": 8815
    },
    {
      "entropy": 5.3895618438720705,
      "epoch": 0.8472622478386167,
      "grad_norm": 1.625,
      "learning_rate": 0.0004936428006625824,
      "loss": 5.2394,
      "mean_token_accuracy": 0.19551685601472854,
      "num_tokens": 20236183.0,
      "step": 8820
    },
    {
      "entropy": 5.483028268814087,
      "epoch": 0.8477425552353506,
      "grad_norm": 1.375,
      "learning_rate": 0.0004936347061785823,
      "loss": 5.3642,
      "mean_token_accuracy": 0.18863052874803543,
      "num_tokens": 20248663.0,
      "step": 8825
    },
    {
      "entropy": 5.377487087249756,
      "epoch": 0.8482228626320846,
      "grad_norm": 1.375,
      "learning_rate": 0.0004936266066185334,
      "loss": 5.2231,
      "mean_token_accuracy": 0.20531052798032762,
      "num_tokens": 20259293.0,
      "step": 8830
    },
    {
      "entropy": 5.399657535552978,
      "epoch": 0.8487031700288185,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004936185019826239,
      "loss": 5.3568,
      "mean_token_accuracy": 0.18434687703847885,
      "num_tokens": 20270601.0,
      "step": 8835
    },
    {
      "entropy": 5.428639936447143,
      "epoch": 0.8491834774255523,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004936103922710419,
      "loss": 5.3339,
      "mean_token_accuracy": 0.19664848148822783,
      "num_tokens": 20281471.0,
      "step": 8840
    },
    {
      "entropy": 5.384542798995971,
      "epoch": 0.8496637848222862,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004936022774839759,
      "loss": 5.1855,
      "mean_token_accuracy": 0.195806385576725,
      "num_tokens": 20291941.0,
      "step": 8845
    },
    {
      "entropy": 5.388844013214111,
      "epoch": 0.8501440922190202,
      "grad_norm": 1.375,
      "learning_rate": 0.0004935941576216141,
      "loss": 5.3134,
      "mean_token_accuracy": 0.18780387938022614,
      "num_tokens": 20302631.0,
      "step": 8850
    },
    {
      "entropy": 5.502013397216797,
      "epoch": 0.8506243996157541,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000493586032684145,
      "loss": 5.3699,
      "mean_token_accuracy": 0.1927383527159691,
      "num_tokens": 20314955.0,
      "step": 8855
    },
    {
      "entropy": 5.446310901641846,
      "epoch": 0.851104707012488,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004935779026717573,
      "loss": 5.307,
      "mean_token_accuracy": 0.1961928054690361,
      "num_tokens": 20326791.0,
      "step": 8860
    },
    {
      "entropy": 5.365289545059204,
      "epoch": 0.8515850144092219,
      "grad_norm": 1.5,
      "learning_rate": 0.0004935697675846396,
      "loss": 5.3124,
      "mean_token_accuracy": 0.1964481234550476,
      "num_tokens": 20338912.0,
      "step": 8865
    },
    {
      "entropy": 5.361674976348877,
      "epoch": 0.8520653218059558,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004935616274229811,
      "loss": 5.2511,
      "mean_token_accuracy": 0.19960997402668,
      "num_tokens": 20349414.0,
      "step": 8870
    },
    {
      "entropy": 5.440748023986816,
      "epoch": 0.8525456292026897,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004935534821869705,
      "loss": 5.2693,
      "mean_token_accuracy": 0.19753523170948029,
      "num_tokens": 20360346.0,
      "step": 8875
    },
    {
      "entropy": 5.38953971862793,
      "epoch": 0.8530259365994236,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004935453318767971,
      "loss": 5.2584,
      "mean_token_accuracy": 0.19536848813295365,
      "num_tokens": 20371965.0,
      "step": 8880
    },
    {
      "entropy": 5.348725652694702,
      "epoch": 0.8535062439961575,
      "grad_norm": 1.984375,
      "learning_rate": 0.00049353717649265,
      "loss": 5.2934,
      "mean_token_accuracy": 0.19363913387060167,
      "num_tokens": 20382632.0,
      "step": 8885
    },
    {
      "entropy": 5.390606117248535,
      "epoch": 0.8539865513928915,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004935290160347185,
      "loss": 5.2641,
      "mean_token_accuracy": 0.19887724071741103,
      "num_tokens": 20395796.0,
      "step": 8890
    },
    {
      "entropy": 5.440628290176392,
      "epoch": 0.8544668587896254,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004935208505031922,
      "loss": 5.3186,
      "mean_token_accuracy": 0.19276156425476074,
      "num_tokens": 20407113.0,
      "step": 8895
    },
    {
      "entropy": 5.422786712646484,
      "epoch": 0.8549471661863592,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004935126798982606,
      "loss": 5.3201,
      "mean_token_accuracy": 0.19261221587657928,
      "num_tokens": 20419012.0,
      "step": 8900
    },
    {
      "entropy": 5.43376522064209,
      "epoch": 0.8554274735830932,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004935045042201135,
      "loss": 5.2628,
      "mean_token_accuracy": 0.19217123985290527,
      "num_tokens": 20430337.0,
      "step": 8905
    },
    {
      "entropy": 5.341459608078003,
      "epoch": 0.8559077809798271,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004934963234689407,
      "loss": 5.2884,
      "mean_token_accuracy": 0.1902077838778496,
      "num_tokens": 20441614.0,
      "step": 8910
    },
    {
      "entropy": 5.3434123516082765,
      "epoch": 0.856388088376561,
      "grad_norm": 1.359375,
      "learning_rate": 0.000493488137644932,
      "loss": 5.1544,
      "mean_token_accuracy": 0.20667791813611985,
      "num_tokens": 20453170.0,
      "step": 8915
    },
    {
      "entropy": 5.338638210296631,
      "epoch": 0.8568683957732949,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004934799467482774,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1991856336593628,
      "num_tokens": 20464585.0,
      "step": 8920
    },
    {
      "entropy": 5.350225353240967,
      "epoch": 0.8573487031700289,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004934717507791673,
      "loss": 5.1129,
      "mean_token_accuracy": 0.2073020800948143,
      "num_tokens": 20473903.0,
      "step": 8925
    },
    {
      "entropy": 5.344707345962524,
      "epoch": 0.8578290105667628,
      "grad_norm": 1.75,
      "learning_rate": 0.0004934635497377919,
      "loss": 5.2815,
      "mean_token_accuracy": 0.196464966237545,
      "num_tokens": 20484778.0,
      "step": 8930
    },
    {
      "entropy": 5.346668672561646,
      "epoch": 0.8583093179634966,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004934553436243415,
      "loss": 5.3124,
      "mean_token_accuracy": 0.18907110542058944,
      "num_tokens": 20495534.0,
      "step": 8935
    },
    {
      "entropy": 5.51176905632019,
      "epoch": 0.8587896253602305,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004934471324390067,
      "loss": 5.37,
      "mean_token_accuracy": 0.18888648599386215,
      "num_tokens": 20506853.0,
      "step": 8940
    },
    {
      "entropy": 5.410428667068482,
      "epoch": 0.8592699327569645,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004934389161819783,
      "loss": 5.3015,
      "mean_token_accuracy": 0.19316813498735427,
      "num_tokens": 20519112.0,
      "step": 8945
    },
    {
      "entropy": 5.406508255004883,
      "epoch": 0.8597502401536984,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004934306948534467,
      "loss": 5.2926,
      "mean_token_accuracy": 0.1921637773513794,
      "num_tokens": 20530060.0,
      "step": 8950
    },
    {
      "entropy": 5.41430082321167,
      "epoch": 0.8602305475504323,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004934224684536031,
      "loss": 5.2309,
      "mean_token_accuracy": 0.19633077830076218,
      "num_tokens": 20541802.0,
      "step": 8955
    },
    {
      "entropy": 5.293176507949829,
      "epoch": 0.8607108549471661,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004934142369826382,
      "loss": 5.1465,
      "mean_token_accuracy": 0.19946417659521104,
      "num_tokens": 20552494.0,
      "step": 8960
    },
    {
      "entropy": 5.329283618927002,
      "epoch": 0.8611911623439001,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004934060004407434,
      "loss": 5.2976,
      "mean_token_accuracy": 0.19979164153337478,
      "num_tokens": 20564467.0,
      "step": 8965
    },
    {
      "entropy": 5.43250150680542,
      "epoch": 0.861671469740634,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004933977588281099,
      "loss": 5.3315,
      "mean_token_accuracy": 0.19024426192045213,
      "num_tokens": 20575399.0,
      "step": 8970
    },
    {
      "entropy": 5.4357610702514645,
      "epoch": 0.8621517771373679,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004933895121449288,
      "loss": 5.2959,
      "mean_token_accuracy": 0.19513940215110778,
      "num_tokens": 20586212.0,
      "step": 8975
    },
    {
      "entropy": 5.4440700054168705,
      "epoch": 0.8626320845341018,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004933812603913917,
      "loss": 5.3044,
      "mean_token_accuracy": 0.19323743581771852,
      "num_tokens": 20597020.0,
      "step": 8980
    },
    {
      "entropy": 5.403150367736816,
      "epoch": 0.8631123919308358,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004933730035676903,
      "loss": 5.2708,
      "mean_token_accuracy": 0.1938205450773239,
      "num_tokens": 20609447.0,
      "step": 8985
    },
    {
      "entropy": 5.3863848686218265,
      "epoch": 0.8635926993275697,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004933647416740161,
      "loss": 5.2318,
      "mean_token_accuracy": 0.19161730855703354,
      "num_tokens": 20621780.0,
      "step": 8990
    },
    {
      "entropy": 5.301457357406616,
      "epoch": 0.8640730067243035,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000493356474710561,
      "loss": 5.182,
      "mean_token_accuracy": 0.2016572281718254,
      "num_tokens": 20632331.0,
      "step": 8995
    },
    {
      "entropy": 5.3469079494476315,
      "epoch": 0.8645533141210374,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000493348202677517,
      "loss": 5.149,
      "mean_token_accuracy": 0.19837099313735962,
      "num_tokens": 20643357.0,
      "step": 9000
    },
    {
      "epoch": 0.8645533141210374,
      "eval_entropy": 5.214188244562507,
      "eval_loss": 5.295583248138428,
      "eval_mean_token_accuracy": 0.19974388994062198,
      "eval_num_tokens": 20643357.0,
      "eval_runtime": 26.6752,
      "eval_samples_per_second": 1230.171,
      "eval_steps_per_second": 153.776,
      "step": 9000
    },
    {
      "entropy": 5.319790649414062,
      "epoch": 0.8650336215177714,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004933399255750761,
      "loss": 5.2072,
      "mean_token_accuracy": 0.20552109479904174,
      "num_tokens": 20655212.0,
      "step": 9005
    },
    {
      "entropy": 5.292992973327637,
      "epoch": 0.8655139289145053,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004933316434034304,
      "loss": 5.0855,
      "mean_token_accuracy": 0.2076722964644432,
      "num_tokens": 20666467.0,
      "step": 9010
    },
    {
      "entropy": 5.324105739593506,
      "epoch": 0.8659942363112392,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004933233561627723,
      "loss": 5.2258,
      "mean_token_accuracy": 0.1977165386080742,
      "num_tokens": 20677709.0,
      "step": 9015
    },
    {
      "entropy": 5.374633121490478,
      "epoch": 0.866474543707973,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004933150638532942,
      "loss": 5.2552,
      "mean_token_accuracy": 0.1895079419016838,
      "num_tokens": 20688880.0,
      "step": 9020
    },
    {
      "entropy": 5.435467195510864,
      "epoch": 0.866954851104707,
      "grad_norm": 1.75,
      "learning_rate": 0.0004933067664751885,
      "loss": 5.399,
      "mean_token_accuracy": 0.18083080649375916,
      "num_tokens": 20701384.0,
      "step": 9025
    },
    {
      "entropy": 5.4591577529907225,
      "epoch": 0.8674351585014409,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000493298464028648,
      "loss": 5.336,
      "mean_token_accuracy": 0.18601617515087127,
      "num_tokens": 20713736.0,
      "step": 9030
    },
    {
      "entropy": 5.299443674087525,
      "epoch": 0.8679154658981748,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004932901565138653,
      "loss": 5.1727,
      "mean_token_accuracy": 0.19958783090114593,
      "num_tokens": 20725424.0,
      "step": 9035
    },
    {
      "entropy": 5.35780816078186,
      "epoch": 0.8683957732949087,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004932818439310334,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19774211347103118,
      "num_tokens": 20737454.0,
      "step": 9040
    },
    {
      "entropy": 5.452247571945191,
      "epoch": 0.8688760806916427,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004932735262803452,
      "loss": 5.3107,
      "mean_token_accuracy": 0.19008514881134034,
      "num_tokens": 20748882.0,
      "step": 9045
    },
    {
      "entropy": 5.372403478622436,
      "epoch": 0.8693563880883766,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004932652035619939,
      "loss": 5.2082,
      "mean_token_accuracy": 0.20026716589927673,
      "num_tokens": 20759622.0,
      "step": 9050
    },
    {
      "entropy": 5.307312536239624,
      "epoch": 0.8698366954851104,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004932568757761727,
      "loss": 5.203,
      "mean_token_accuracy": 0.19591446667909623,
      "num_tokens": 20770658.0,
      "step": 9055
    },
    {
      "entropy": 5.435375928878784,
      "epoch": 0.8703170028818443,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004932485429230748,
      "loss": 5.2762,
      "mean_token_accuracy": 0.19558858573436738,
      "num_tokens": 20782176.0,
      "step": 9060
    },
    {
      "entropy": 5.467179918289185,
      "epoch": 0.8707973102785783,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000493240205002894,
      "loss": 5.323,
      "mean_token_accuracy": 0.18266102522611619,
      "num_tokens": 20793592.0,
      "step": 9065
    },
    {
      "entropy": 5.402525949478149,
      "epoch": 0.8712776176753122,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004932318620158235,
      "loss": 5.2027,
      "mean_token_accuracy": 0.1949774757027626,
      "num_tokens": 20804226.0,
      "step": 9070
    },
    {
      "entropy": 5.361907863616944,
      "epoch": 0.8717579250720461,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004932235139620574,
      "loss": 5.3587,
      "mean_token_accuracy": 0.18637584000825883,
      "num_tokens": 20815336.0,
      "step": 9075
    },
    {
      "entropy": 5.397728490829468,
      "epoch": 0.8722382324687801,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004932151608417892,
      "loss": 5.2562,
      "mean_token_accuracy": 0.20355497896671296,
      "num_tokens": 20826936.0,
      "step": 9080
    },
    {
      "entropy": 5.411286020278931,
      "epoch": 0.872718539865514,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004932068026552127,
      "loss": 5.2945,
      "mean_token_accuracy": 0.19251975566148757,
      "num_tokens": 20838159.0,
      "step": 9085
    },
    {
      "entropy": 5.403781127929688,
      "epoch": 0.8731988472622478,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004931984394025224,
      "loss": 5.203,
      "mean_token_accuracy": 0.1905702084302902,
      "num_tokens": 20848605.0,
      "step": 9090
    },
    {
      "entropy": 5.384661340713501,
      "epoch": 0.8736791546589817,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004931900710839123,
      "loss": 5.3429,
      "mean_token_accuracy": 0.18770140856504441,
      "num_tokens": 20860494.0,
      "step": 9095
    },
    {
      "entropy": 5.419786643981934,
      "epoch": 0.8741594620557157,
      "grad_norm": 1.5,
      "learning_rate": 0.0004931816976995766,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19947386980056764,
      "num_tokens": 20871327.0,
      "step": 9100
    },
    {
      "entropy": 5.361711931228638,
      "epoch": 0.8746397694524496,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004931733192497097,
      "loss": 5.2794,
      "mean_token_accuracy": 0.1988350734114647,
      "num_tokens": 20883032.0,
      "step": 9105
    },
    {
      "entropy": 5.391395950317383,
      "epoch": 0.8751200768491835,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004931649357345062,
      "loss": 5.3062,
      "mean_token_accuracy": 0.1914994791150093,
      "num_tokens": 20894327.0,
      "step": 9110
    },
    {
      "entropy": 5.352286529541016,
      "epoch": 0.8756003842459174,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004931565471541606,
      "loss": 5.1546,
      "mean_token_accuracy": 0.20428362339735032,
      "num_tokens": 20905489.0,
      "step": 9115
    },
    {
      "entropy": 5.357045125961304,
      "epoch": 0.8760806916426513,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004931481535088679,
      "loss": 5.2534,
      "mean_token_accuracy": 0.1897742196917534,
      "num_tokens": 20917044.0,
      "step": 9120
    },
    {
      "entropy": 5.380188369750977,
      "epoch": 0.8765609990393852,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004931397547988229,
      "loss": 5.233,
      "mean_token_accuracy": 0.19599434435367585,
      "num_tokens": 20928065.0,
      "step": 9125
    },
    {
      "entropy": 5.405956602096557,
      "epoch": 0.8770413064361191,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004931313510242204,
      "loss": 5.2457,
      "mean_token_accuracy": 0.201428684592247,
      "num_tokens": 20939396.0,
      "step": 9130
    },
    {
      "entropy": 5.446123790740967,
      "epoch": 0.877521613832853,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004931229421852557,
      "loss": 5.348,
      "mean_token_accuracy": 0.191171233355999,
      "num_tokens": 20950844.0,
      "step": 9135
    },
    {
      "entropy": 5.366133832931519,
      "epoch": 0.878001921229587,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000493114528282124,
      "loss": 5.2317,
      "mean_token_accuracy": 0.19597537517547609,
      "num_tokens": 20961567.0,
      "step": 9140
    },
    {
      "entropy": 5.541252994537354,
      "epoch": 0.8784822286263209,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004931061093150206,
      "loss": 5.3942,
      "mean_token_accuracy": 0.19285230934619904,
      "num_tokens": 20972762.0,
      "step": 9145
    },
    {
      "entropy": 5.329995584487915,
      "epoch": 0.8789625360230547,
      "grad_norm": 1.625,
      "learning_rate": 0.0004930976852841409,
      "loss": 5.2204,
      "mean_token_accuracy": 0.19554354697465898,
      "num_tokens": 20983848.0,
      "step": 9150
    },
    {
      "entropy": 5.383824348449707,
      "epoch": 0.8794428434197886,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004930892561896806,
      "loss": 5.3022,
      "mean_token_accuracy": 0.19015206396579742,
      "num_tokens": 20996418.0,
      "step": 9155
    },
    {
      "entropy": 5.404303693771363,
      "epoch": 0.8799231508165226,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004930808220318354,
      "loss": 5.2831,
      "mean_token_accuracy": 0.19629869908094405,
      "num_tokens": 21008714.0,
      "step": 9160
    },
    {
      "entropy": 5.420473670959472,
      "epoch": 0.8804034582132565,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004930723828108012,
      "loss": 5.2113,
      "mean_token_accuracy": 0.1992582142353058,
      "num_tokens": 21020453.0,
      "step": 9165
    },
    {
      "entropy": 5.280115604400635,
      "epoch": 0.8808837656099904,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004930639385267736,
      "loss": 5.1907,
      "mean_token_accuracy": 0.19808709174394606,
      "num_tokens": 21032211.0,
      "step": 9170
    },
    {
      "entropy": 5.382225084304809,
      "epoch": 0.8813640730067243,
      "grad_norm": 1.828125,
      "learning_rate": 0.000493055489179949,
      "loss": 5.2815,
      "mean_token_accuracy": 0.1933873787522316,
      "num_tokens": 21042995.0,
      "step": 9175
    },
    {
      "entropy": 5.341494941711426,
      "epoch": 0.8818443804034583,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004930470347705234,
      "loss": 5.1498,
      "mean_token_accuracy": 0.2004062846302986,
      "num_tokens": 21054306.0,
      "step": 9180
    },
    {
      "entropy": 5.37236065864563,
      "epoch": 0.8823246878001921,
      "grad_norm": 1.96875,
      "learning_rate": 0.000493038575298693,
      "loss": 5.3271,
      "mean_token_accuracy": 0.1897224262356758,
      "num_tokens": 21067144.0,
      "step": 9185
    },
    {
      "entropy": 5.531492280960083,
      "epoch": 0.882804995196926,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004930301107646545,
      "loss": 5.3021,
      "mean_token_accuracy": 0.18749115318059922,
      "num_tokens": 21078885.0,
      "step": 9190
    },
    {
      "entropy": 5.5086592674255375,
      "epoch": 0.8832853025936599,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004930216411686042,
      "loss": 5.3564,
      "mean_token_accuracy": 0.18278430551290512,
      "num_tokens": 21089482.0,
      "step": 9195
    },
    {
      "entropy": 5.44271330833435,
      "epoch": 0.8837656099903939,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004930131665107387,
      "loss": 5.346,
      "mean_token_accuracy": 0.18601189106702803,
      "num_tokens": 21102046.0,
      "step": 9200
    },
    {
      "entropy": 5.362988042831421,
      "epoch": 0.8842459173871278,
      "grad_norm": 1.546875,
      "learning_rate": 0.000493004686791255,
      "loss": 5.1942,
      "mean_token_accuracy": 0.19421408623456954,
      "num_tokens": 21112827.0,
      "step": 9205
    },
    {
      "entropy": 5.400761413574219,
      "epoch": 0.8847262247838616,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004929962020103496,
      "loss": 5.297,
      "mean_token_accuracy": 0.19207022935152054,
      "num_tokens": 21123108.0,
      "step": 9210
    },
    {
      "entropy": 5.409162759780884,
      "epoch": 0.8852065321805955,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004929877121682198,
      "loss": 5.3154,
      "mean_token_accuracy": 0.19236302077770234,
      "num_tokens": 21134955.0,
      "step": 9215
    },
    {
      "entropy": 5.387013673782349,
      "epoch": 0.8856868395773295,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004929792172650627,
      "loss": 5.2002,
      "mean_token_accuracy": 0.20066718459129335,
      "num_tokens": 21146415.0,
      "step": 9220
    },
    {
      "entropy": 5.33963303565979,
      "epoch": 0.8861671469740634,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004929707173010753,
      "loss": 5.2323,
      "mean_token_accuracy": 0.20095980763435364,
      "num_tokens": 21158691.0,
      "step": 9225
    },
    {
      "entropy": 5.355684757232666,
      "epoch": 0.8866474543707973,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004929622122764552,
      "loss": 5.2572,
      "mean_token_accuracy": 0.19015011489391326,
      "num_tokens": 21170063.0,
      "step": 9230
    },
    {
      "entropy": 5.431982803344726,
      "epoch": 0.8871277617675313,
      "grad_norm": 1.5,
      "learning_rate": 0.0004929537021913997,
      "loss": 5.3289,
      "mean_token_accuracy": 0.19112800806760788,
      "num_tokens": 21182427.0,
      "step": 9235
    },
    {
      "entropy": 5.394394874572754,
      "epoch": 0.8876080691642652,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004929451870461064,
      "loss": 5.2,
      "mean_token_accuracy": 0.19691623598337174,
      "num_tokens": 21195011.0,
      "step": 9240
    },
    {
      "entropy": 5.375449895858765,
      "epoch": 0.888088376560999,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004929366668407731,
      "loss": 5.2389,
      "mean_token_accuracy": 0.1980753645300865,
      "num_tokens": 21207886.0,
      "step": 9245
    },
    {
      "entropy": 5.3231419086456295,
      "epoch": 0.8885686839577329,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004929281415755974,
      "loss": 5.203,
      "mean_token_accuracy": 0.20334324836730958,
      "num_tokens": 21219285.0,
      "step": 9250
    },
    {
      "entropy": 5.421673107147217,
      "epoch": 0.8890489913544669,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004929196112507775,
      "loss": 5.259,
      "mean_token_accuracy": 0.19196344614028932,
      "num_tokens": 21231208.0,
      "step": 9255
    },
    {
      "entropy": 5.350038051605225,
      "epoch": 0.8895292987512008,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004929110758665112,
      "loss": 5.164,
      "mean_token_accuracy": 0.19924798756837844,
      "num_tokens": 21242573.0,
      "step": 9260
    },
    {
      "entropy": 5.237231779098511,
      "epoch": 0.8900096061479347,
      "grad_norm": 1.375,
      "learning_rate": 0.0004929025354229969,
      "loss": 5.1464,
      "mean_token_accuracy": 0.20820244401693344,
      "num_tokens": 21254966.0,
      "step": 9265
    },
    {
      "entropy": 5.3665369033813475,
      "epoch": 0.8904899135446686,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004928939899204326,
      "loss": 5.2765,
      "mean_token_accuracy": 0.18944831490516661,
      "num_tokens": 21266167.0,
      "step": 9270
    },
    {
      "entropy": 5.415092134475708,
      "epoch": 0.8909702209414025,
      "grad_norm": 1.390625,
      "learning_rate": 0.000492885439359017,
      "loss": 5.279,
      "mean_token_accuracy": 0.19464683830738067,
      "num_tokens": 21277036.0,
      "step": 9275
    },
    {
      "entropy": 5.3250429153442385,
      "epoch": 0.8914505283381364,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004928768837389485,
      "loss": 5.2262,
      "mean_token_accuracy": 0.1972446471452713,
      "num_tokens": 21288887.0,
      "step": 9280
    },
    {
      "entropy": 5.3988782405853275,
      "epoch": 0.8919308357348703,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004928683230604257,
      "loss": 5.3348,
      "mean_token_accuracy": 0.1861489549279213,
      "num_tokens": 21300664.0,
      "step": 9285
    },
    {
      "entropy": 5.475399017333984,
      "epoch": 0.8924111431316042,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004928597573236474,
      "loss": 5.1948,
      "mean_token_accuracy": 0.19874988943338395,
      "num_tokens": 21311535.0,
      "step": 9290
    },
    {
      "entropy": 5.2980047225952145,
      "epoch": 0.8928914505283382,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004928511865288123,
      "loss": 5.1741,
      "mean_token_accuracy": 0.1983889237046242,
      "num_tokens": 21323248.0,
      "step": 9295
    },
    {
      "entropy": 5.315932846069336,
      "epoch": 0.8933717579250721,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004928426106761197,
      "loss": 5.2111,
      "mean_token_accuracy": 0.19994667172431946,
      "num_tokens": 21334901.0,
      "step": 9300
    },
    {
      "entropy": 5.490907478332519,
      "epoch": 0.8938520653218059,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004928340297657685,
      "loss": 5.3213,
      "mean_token_accuracy": 0.19272645860910415,
      "num_tokens": 21345902.0,
      "step": 9305
    },
    {
      "entropy": 5.45790958404541,
      "epoch": 0.8943323727185398,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004928254437979578,
      "loss": 5.3467,
      "mean_token_accuracy": 0.18789408057928086,
      "num_tokens": 21358240.0,
      "step": 9310
    },
    {
      "entropy": 5.252040672302246,
      "epoch": 0.8948126801152738,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004928168527728873,
      "loss": 5.1059,
      "mean_token_accuracy": 0.20298046618700027,
      "num_tokens": 21368209.0,
      "step": 9315
    },
    {
      "entropy": 5.282319402694702,
      "epoch": 0.8952929875120077,
      "grad_norm": 1.5,
      "learning_rate": 0.0004928082566907562,
      "loss": 5.211,
      "mean_token_accuracy": 0.1902171015739441,
      "num_tokens": 21379484.0,
      "step": 9320
    },
    {
      "entropy": 5.376897668838501,
      "epoch": 0.8957732949087416,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004927996555517642,
      "loss": 5.2194,
      "mean_token_accuracy": 0.19863976538181305,
      "num_tokens": 21391750.0,
      "step": 9325
    },
    {
      "entropy": 5.434967279434204,
      "epoch": 0.8962536023054755,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004927910493561109,
      "loss": 5.3068,
      "mean_token_accuracy": 0.19308367669582366,
      "num_tokens": 21402583.0,
      "step": 9330
    },
    {
      "entropy": 5.3867422580719,
      "epoch": 0.8967339097022095,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000492782438103996,
      "loss": 5.2186,
      "mean_token_accuracy": 0.18686265200376512,
      "num_tokens": 21412892.0,
      "step": 9335
    },
    {
      "entropy": 5.356006622314453,
      "epoch": 0.8972142170989433,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004927738217956197,
      "loss": 5.2302,
      "mean_token_accuracy": 0.19290295392274856,
      "num_tokens": 21422828.0,
      "step": 9340
    },
    {
      "entropy": 5.293871212005615,
      "epoch": 0.8976945244956772,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004927652004311819,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19278172105550767,
      "num_tokens": 21434441.0,
      "step": 9345
    },
    {
      "entropy": 5.424458646774292,
      "epoch": 0.8981748318924111,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004927565740108828,
      "loss": 5.3934,
      "mean_token_accuracy": 0.1887845903635025,
      "num_tokens": 21445945.0,
      "step": 9350
    },
    {
      "entropy": 5.456415605545044,
      "epoch": 0.8986551392891451,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004927479425349226,
      "loss": 5.2954,
      "mean_token_accuracy": 0.19610380828380586,
      "num_tokens": 21457265.0,
      "step": 9355
    },
    {
      "entropy": 5.3642435550689695,
      "epoch": 0.899135446685879,
      "grad_norm": 1.875,
      "learning_rate": 0.0004927393060035018,
      "loss": 5.2434,
      "mean_token_accuracy": 0.19245890378952027,
      "num_tokens": 21469090.0,
      "step": 9360
    },
    {
      "entropy": 5.368222188949585,
      "epoch": 0.8996157540826129,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004927306644168207,
      "loss": 5.2944,
      "mean_token_accuracy": 0.18543377965688707,
      "num_tokens": 21480280.0,
      "step": 9365
    },
    {
      "entropy": 5.34284782409668,
      "epoch": 0.9000960614793467,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004927220177750803,
      "loss": 5.2196,
      "mean_token_accuracy": 0.20020502656698227,
      "num_tokens": 21491856.0,
      "step": 9370
    },
    {
      "entropy": 5.388748836517334,
      "epoch": 0.9005763688760807,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004927133660784811,
      "loss": 5.2489,
      "mean_token_accuracy": 0.19540146142244338,
      "num_tokens": 21501814.0,
      "step": 9375
    },
    {
      "entropy": 5.383769416809082,
      "epoch": 0.9010566762728146,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004927047093272241,
      "loss": 5.2813,
      "mean_token_accuracy": 0.19255306273698808,
      "num_tokens": 21514725.0,
      "step": 9380
    },
    {
      "entropy": 5.320773458480835,
      "epoch": 0.9015369836695485,
      "grad_norm": 1.75,
      "learning_rate": 0.00049269604752151,
      "loss": 5.2328,
      "mean_token_accuracy": 0.1984529584646225,
      "num_tokens": 21526523.0,
      "step": 9385
    },
    {
      "entropy": 5.336462211608887,
      "epoch": 0.9020172910662824,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004926873806615403,
      "loss": 5.2019,
      "mean_token_accuracy": 0.19943687021732331,
      "num_tokens": 21537947.0,
      "step": 9390
    },
    {
      "entropy": 5.446700811386108,
      "epoch": 0.9024975984630164,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004926787087475158,
      "loss": 5.3111,
      "mean_token_accuracy": 0.18979995995759963,
      "num_tokens": 21550238.0,
      "step": 9395
    },
    {
      "entropy": 5.364442968368531,
      "epoch": 0.9029779058597502,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004926700317796382,
      "loss": 5.306,
      "mean_token_accuracy": 0.1879432648420334,
      "num_tokens": 21561546.0,
      "step": 9400
    },
    {
      "entropy": 5.408336925506592,
      "epoch": 0.9034582132564841,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004926613497581088,
      "loss": 5.287,
      "mean_token_accuracy": 0.19076864719390868,
      "num_tokens": 21572034.0,
      "step": 9405
    },
    {
      "entropy": 5.350472354888916,
      "epoch": 0.9039385206532181,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004926526626831292,
      "loss": 5.2499,
      "mean_token_accuracy": 0.18989201933145522,
      "num_tokens": 21583427.0,
      "step": 9410
    },
    {
      "entropy": 5.372383165359497,
      "epoch": 0.904418828049952,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004926439705549011,
      "loss": 5.2368,
      "mean_token_accuracy": 0.19854346364736558,
      "num_tokens": 21595071.0,
      "step": 9415
    },
    {
      "entropy": 5.46326642036438,
      "epoch": 0.9048991354466859,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004926352733736262,
      "loss": 5.2816,
      "mean_token_accuracy": 0.1946068823337555,
      "num_tokens": 21606007.0,
      "step": 9420
    },
    {
      "entropy": 5.383008003234863,
      "epoch": 0.9053794428434198,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004926265711395065,
      "loss": 5.2386,
      "mean_token_accuracy": 0.1993948295712471,
      "num_tokens": 21617988.0,
      "step": 9425
    },
    {
      "entropy": 5.362982988357544,
      "epoch": 0.9058597502401537,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000492617863852744,
      "loss": 5.2641,
      "mean_token_accuracy": 0.1956845372915268,
      "num_tokens": 21629832.0,
      "step": 9430
    },
    {
      "entropy": 5.386901140213013,
      "epoch": 0.9063400576368876,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004926091515135409,
      "loss": 5.2395,
      "mean_token_accuracy": 0.19297112822532653,
      "num_tokens": 21641295.0,
      "step": 9435
    },
    {
      "entropy": 5.300440263748169,
      "epoch": 0.9068203650336215,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004926004341220995,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20337535440921783,
      "num_tokens": 21652829.0,
      "step": 9440
    },
    {
      "entropy": 5.463212776184082,
      "epoch": 0.9073006724303554,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004925917116786222,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1866364523768425,
      "num_tokens": 21665032.0,
      "step": 9445
    },
    {
      "entropy": 5.269292116165161,
      "epoch": 0.9077809798270894,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004925829841833114,
      "loss": 5.1626,
      "mean_token_accuracy": 0.20581218153238295,
      "num_tokens": 21675814.0,
      "step": 9450
    },
    {
      "entropy": 5.336495971679687,
      "epoch": 0.9082612872238233,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004925742516363699,
      "loss": 5.2619,
      "mean_token_accuracy": 0.19505233317613602,
      "num_tokens": 21687581.0,
      "step": 9455
    },
    {
      "entropy": 5.445864200592041,
      "epoch": 0.9087415946205571,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004925655140380002,
      "loss": 5.2683,
      "mean_token_accuracy": 0.18777590095996857,
      "num_tokens": 21700167.0,
      "step": 9460
    },
    {
      "entropy": 5.414084529876709,
      "epoch": 0.909221902017291,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004925567713884054,
      "loss": 5.3067,
      "mean_token_accuracy": 0.19373577386140822,
      "num_tokens": 21712309.0,
      "step": 9465
    },
    {
      "entropy": 5.35190258026123,
      "epoch": 0.909702209414025,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004925480236877884,
      "loss": 5.2154,
      "mean_token_accuracy": 0.2007080465555191,
      "num_tokens": 21723391.0,
      "step": 9470
    },
    {
      "entropy": 5.35431981086731,
      "epoch": 0.9101825168107589,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004925392709363522,
      "loss": 5.165,
      "mean_token_accuracy": 0.1985024854540825,
      "num_tokens": 21735052.0,
      "step": 9475
    },
    {
      "entropy": 5.437685775756836,
      "epoch": 0.9106628242074928,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004925305131343001,
      "loss": 5.2971,
      "mean_token_accuracy": 0.18960195481777192,
      "num_tokens": 21746424.0,
      "step": 9480
    },
    {
      "entropy": 5.382255172729492,
      "epoch": 0.9111431316042267,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004925217502818355,
      "loss": 5.2658,
      "mean_token_accuracy": 0.19176243096590043,
      "num_tokens": 21756960.0,
      "step": 9485
    },
    {
      "entropy": 5.4193182468414305,
      "epoch": 0.9116234390009607,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004925129823791616,
      "loss": 5.231,
      "mean_token_accuracy": 0.1988448679447174,
      "num_tokens": 21767621.0,
      "step": 9490
    },
    {
      "entropy": 5.4150067329406735,
      "epoch": 0.9121037463976945,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004925042094264822,
      "loss": 5.2851,
      "mean_token_accuracy": 0.1916612908244133,
      "num_tokens": 21779403.0,
      "step": 9495
    },
    {
      "entropy": 5.487259197235107,
      "epoch": 0.9125840537944284,
      "grad_norm": 1.796875,
      "learning_rate": 0.000492495431424001,
      "loss": 5.3745,
      "mean_token_accuracy": 0.18581128120422363,
      "num_tokens": 21791039.0,
      "step": 9500
    },
    {
      "entropy": 5.520426940917969,
      "epoch": 0.9130643611911623,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004924866483719216,
      "loss": 5.3491,
      "mean_token_accuracy": 0.1887574389576912,
      "num_tokens": 21803731.0,
      "step": 9505
    },
    {
      "entropy": 5.437395095825195,
      "epoch": 0.9135446685878963,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004924778602704481,
      "loss": 5.2535,
      "mean_token_accuracy": 0.19384080022573472,
      "num_tokens": 21816139.0,
      "step": 9510
    },
    {
      "entropy": 5.256186199188233,
      "epoch": 0.9140249759846302,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004924690671197845,
      "loss": 5.0978,
      "mean_token_accuracy": 0.2019031748175621,
      "num_tokens": 21826861.0,
      "step": 9515
    },
    {
      "entropy": 5.321603488922119,
      "epoch": 0.914505283381364,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004924602689201348,
      "loss": 5.2703,
      "mean_token_accuracy": 0.19992187917232512,
      "num_tokens": 21838083.0,
      "step": 9520
    },
    {
      "entropy": 5.396101331710815,
      "epoch": 0.9149855907780979,
      "grad_norm": 5.53125,
      "learning_rate": 0.0004924514656717034,
      "loss": 5.2909,
      "mean_token_accuracy": 0.19334015399217605,
      "num_tokens": 21849232.0,
      "step": 9525
    },
    {
      "entropy": 5.442108631134033,
      "epoch": 0.9154658981748319,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004924426573746948,
      "loss": 5.322,
      "mean_token_accuracy": 0.18857376277446747,
      "num_tokens": 21861650.0,
      "step": 9530
    },
    {
      "entropy": 5.535091924667358,
      "epoch": 0.9159462055715658,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004924338440293131,
      "loss": 5.3619,
      "mean_token_accuracy": 0.18689346462488174,
      "num_tokens": 21873369.0,
      "step": 9535
    },
    {
      "entropy": 5.426569032669067,
      "epoch": 0.9164265129682997,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004924250256357635,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19275580942630768,
      "num_tokens": 21884563.0,
      "step": 9540
    },
    {
      "entropy": 5.407162189483643,
      "epoch": 0.9169068203650336,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004924162021942502,
      "loss": 5.2677,
      "mean_token_accuracy": 0.18890339136123657,
      "num_tokens": 21896114.0,
      "step": 9545
    },
    {
      "entropy": 5.407561922073365,
      "epoch": 0.9173871277617676,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004924073737049784,
      "loss": 5.2745,
      "mean_token_accuracy": 0.19579027444124222,
      "num_tokens": 21907386.0,
      "step": 9550
    },
    {
      "entropy": 5.447867918014526,
      "epoch": 0.9178674351585014,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004923985401681528,
      "loss": 5.3443,
      "mean_token_accuracy": 0.1974613383412361,
      "num_tokens": 21918312.0,
      "step": 9555
    },
    {
      "entropy": 5.352988958358765,
      "epoch": 0.9183477425552353,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004923897015839788,
      "loss": 5.1952,
      "mean_token_accuracy": 0.2020619332790375,
      "num_tokens": 21930101.0,
      "step": 9560
    },
    {
      "entropy": 5.369843196868897,
      "epoch": 0.9188280499519692,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004923808579526613,
      "loss": 5.3332,
      "mean_token_accuracy": 0.19904195368289948,
      "num_tokens": 21941739.0,
      "step": 9565
    },
    {
      "entropy": 5.400526666641236,
      "epoch": 0.9193083573487032,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004923720092744059,
      "loss": 5.3028,
      "mean_token_accuracy": 0.1840466007590294,
      "num_tokens": 21953342.0,
      "step": 9570
    },
    {
      "entropy": 5.472770500183105,
      "epoch": 0.9197886647454371,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004923631555494179,
      "loss": 5.3217,
      "mean_token_accuracy": 0.19470929652452468,
      "num_tokens": 21963506.0,
      "step": 9575
    },
    {
      "entropy": 5.472641944885254,
      "epoch": 0.920268972142171,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004923542967779028,
      "loss": 5.3604,
      "mean_token_accuracy": 0.187233130633831,
      "num_tokens": 21975078.0,
      "step": 9580
    },
    {
      "entropy": 5.388547897338867,
      "epoch": 0.920749279538905,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004923454329600664,
      "loss": 5.2165,
      "mean_token_accuracy": 0.1975598618388176,
      "num_tokens": 21987128.0,
      "step": 9585
    },
    {
      "entropy": 5.387301588058472,
      "epoch": 0.9212295869356388,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004923365640961143,
      "loss": 5.317,
      "mean_token_accuracy": 0.19394897818565368,
      "num_tokens": 22000106.0,
      "step": 9590
    },
    {
      "entropy": 5.363097286224365,
      "epoch": 0.9217098943323727,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004923276901862526,
      "loss": 5.2096,
      "mean_token_accuracy": 0.19964518398046494,
      "num_tokens": 22010571.0,
      "step": 9595
    },
    {
      "entropy": 5.349168586730957,
      "epoch": 0.9221902017291066,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004923188112306874,
      "loss": 5.1819,
      "mean_token_accuracy": 0.20611703097820283,
      "num_tokens": 22021399.0,
      "step": 9600
    },
    {
      "entropy": 5.340490484237671,
      "epoch": 0.9226705091258406,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004923099272296246,
      "loss": 5.1773,
      "mean_token_accuracy": 0.20286252796649934,
      "num_tokens": 22032302.0,
      "step": 9605
    },
    {
      "entropy": 5.359250450134278,
      "epoch": 0.9231508165225745,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004923010381832706,
      "loss": 5.2064,
      "mean_token_accuracy": 0.20294761061668395,
      "num_tokens": 22043762.0,
      "step": 9610
    },
    {
      "entropy": 5.421875,
      "epoch": 0.9236311239193083,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004922921440918318,
      "loss": 5.2491,
      "mean_token_accuracy": 0.20064163953065872,
      "num_tokens": 22054885.0,
      "step": 9615
    },
    {
      "entropy": 5.486204767227173,
      "epoch": 0.9241114313160422,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004922832449555144,
      "loss": 5.3283,
      "mean_token_accuracy": 0.1889108180999756,
      "num_tokens": 22066627.0,
      "step": 9620
    },
    {
      "entropy": 5.318388605117798,
      "epoch": 0.9245917387127762,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004922743407745255,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19961403012275697,
      "num_tokens": 22078062.0,
      "step": 9625
    },
    {
      "entropy": 5.325558471679687,
      "epoch": 0.9250720461095101,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004922654315490714,
      "loss": 5.1839,
      "mean_token_accuracy": 0.19793693721294403,
      "num_tokens": 22089465.0,
      "step": 9630
    },
    {
      "entropy": 5.3975550651550295,
      "epoch": 0.925552353506244,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004922565172793593,
      "loss": 5.3119,
      "mean_token_accuracy": 0.19478186666965486,
      "num_tokens": 22101688.0,
      "step": 9635
    },
    {
      "entropy": 5.38715124130249,
      "epoch": 0.9260326609029779,
      "grad_norm": 1.75,
      "learning_rate": 0.0004922475979655958,
      "loss": 5.2866,
      "mean_token_accuracy": 0.1929781600832939,
      "num_tokens": 22112912.0,
      "step": 9640
    },
    {
      "entropy": 5.426280546188354,
      "epoch": 0.9265129682997119,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004922386736079883,
      "loss": 5.3924,
      "mean_token_accuracy": 0.1836224839091301,
      "num_tokens": 22125079.0,
      "step": 9645
    },
    {
      "entropy": 5.476082515716553,
      "epoch": 0.9269932756964457,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004922297442067438,
      "loss": 5.2842,
      "mean_token_accuracy": 0.19882182478904725,
      "num_tokens": 22137146.0,
      "step": 9650
    },
    {
      "entropy": 5.335666751861572,
      "epoch": 0.9274735830931796,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004922208097620697,
      "loss": 5.229,
      "mean_token_accuracy": 0.19252820312976837,
      "num_tokens": 22148330.0,
      "step": 9655
    },
    {
      "entropy": 5.309677457809448,
      "epoch": 0.9279538904899135,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004922118702741735,
      "loss": 5.218,
      "mean_token_accuracy": 0.19758376479148865,
      "num_tokens": 22160511.0,
      "step": 9660
    },
    {
      "entropy": 5.3699281215667725,
      "epoch": 0.9284341978866475,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004922029257432625,
      "loss": 5.2199,
      "mean_token_accuracy": 0.19192028641700745,
      "num_tokens": 22171298.0,
      "step": 9665
    },
    {
      "entropy": 5.334575033187866,
      "epoch": 0.9289145052833814,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004921939761695446,
      "loss": 5.2676,
      "mean_token_accuracy": 0.19729354977607727,
      "num_tokens": 22181897.0,
      "step": 9670
    },
    {
      "entropy": 5.4648600101470945,
      "epoch": 0.9293948126801153,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004921850215532275,
      "loss": 5.3663,
      "mean_token_accuracy": 0.195632903277874,
      "num_tokens": 22194630.0,
      "step": 9675
    },
    {
      "entropy": 5.424686479568481,
      "epoch": 0.9298751200768491,
      "grad_norm": 1.5,
      "learning_rate": 0.0004921760618945192,
      "loss": 5.2853,
      "mean_token_accuracy": 0.19682292640209198,
      "num_tokens": 22206041.0,
      "step": 9680
    },
    {
      "entropy": 5.34432544708252,
      "epoch": 0.9303554274735831,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004921670971936276,
      "loss": 5.2767,
      "mean_token_accuracy": 0.1924143522977829,
      "num_tokens": 22218434.0,
      "step": 9685
    },
    {
      "entropy": 5.42786340713501,
      "epoch": 0.930835734870317,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004921581274507607,
      "loss": 5.267,
      "mean_token_accuracy": 0.18786528408527375,
      "num_tokens": 22229182.0,
      "step": 9690
    },
    {
      "entropy": 5.445178556442261,
      "epoch": 0.9313160422670509,
      "grad_norm": 1.53125,
      "learning_rate": 0.000492149152666127,
      "loss": 5.3685,
      "mean_token_accuracy": 0.18807477951049806,
      "num_tokens": 22241142.0,
      "step": 9695
    },
    {
      "entropy": 5.437403202056885,
      "epoch": 0.9317963496637848,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004921401728399348,
      "loss": 5.2522,
      "mean_token_accuracy": 0.19175815135240554,
      "num_tokens": 22252695.0,
      "step": 9700
    },
    {
      "entropy": 5.310226726531982,
      "epoch": 0.9322766570605188,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004921311879723926,
      "loss": 5.1993,
      "mean_token_accuracy": 0.2067328378558159,
      "num_tokens": 22264159.0,
      "step": 9705
    },
    {
      "entropy": 5.405073881149292,
      "epoch": 0.9327569644572526,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004921221980637088,
      "loss": 5.2651,
      "mean_token_accuracy": 0.1962537631392479,
      "num_tokens": 22276154.0,
      "step": 9710
    },
    {
      "entropy": 5.328506326675415,
      "epoch": 0.9332372718539865,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004921132031140925,
      "loss": 5.1595,
      "mean_token_accuracy": 0.19734954386949538,
      "num_tokens": 22288328.0,
      "step": 9715
    },
    {
      "entropy": 5.392610168457031,
      "epoch": 0.9337175792507204,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004921042031237521,
      "loss": 5.2491,
      "mean_token_accuracy": 0.19224864095449448,
      "num_tokens": 22300154.0,
      "step": 9720
    },
    {
      "entropy": 5.2410666942596436,
      "epoch": 0.9341978866474544,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004920951980928969,
      "loss": 5.0804,
      "mean_token_accuracy": 0.21091974675655364,
      "num_tokens": 22311176.0,
      "step": 9725
    },
    {
      "entropy": 5.370237970352173,
      "epoch": 0.9346781940441883,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004920861880217359,
      "loss": 5.2691,
      "mean_token_accuracy": 0.19298372715711593,
      "num_tokens": 22321953.0,
      "step": 9730
    },
    {
      "entropy": 5.403165578842163,
      "epoch": 0.9351585014409222,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004920771729104781,
      "loss": 5.2717,
      "mean_token_accuracy": 0.19578431993722917,
      "num_tokens": 22333152.0,
      "step": 9735
    },
    {
      "entropy": 5.381178903579712,
      "epoch": 0.9356388088376562,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004920681527593329,
      "loss": 5.2661,
      "mean_token_accuracy": 0.19532056599855424,
      "num_tokens": 22344848.0,
      "step": 9740
    },
    {
      "entropy": 5.36601710319519,
      "epoch": 0.93611911623439,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004920591275685098,
      "loss": 5.2003,
      "mean_token_accuracy": 0.19312046319246293,
      "num_tokens": 22355100.0,
      "step": 9745
    },
    {
      "entropy": 5.386777877807617,
      "epoch": 0.9365994236311239,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004920500973382184,
      "loss": 5.2171,
      "mean_token_accuracy": 0.20108671486377716,
      "num_tokens": 22366045.0,
      "step": 9750
    },
    {
      "entropy": 5.344554233551025,
      "epoch": 0.9370797310278578,
      "grad_norm": 1.375,
      "learning_rate": 0.0004920410620686682,
      "loss": 5.2223,
      "mean_token_accuracy": 0.19255712777376174,
      "num_tokens": 22378781.0,
      "step": 9755
    },
    {
      "entropy": 5.460892105102539,
      "epoch": 0.9375600384245918,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004920320217600689,
      "loss": 5.3215,
      "mean_token_accuracy": 0.19091278463602065,
      "num_tokens": 22390121.0,
      "step": 9760
    },
    {
      "entropy": 5.340709495544433,
      "epoch": 0.9380403458213257,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004920229764126306,
      "loss": 5.2055,
      "mean_token_accuracy": 0.20116310864686965,
      "num_tokens": 22402261.0,
      "step": 9765
    },
    {
      "entropy": 5.385502433776855,
      "epoch": 0.9385206532180596,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004920139260265632,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19694014340639115,
      "num_tokens": 22413939.0,
      "step": 9770
    },
    {
      "entropy": 5.395786762237549,
      "epoch": 0.9390009606147934,
      "grad_norm": 1.375,
      "learning_rate": 0.0004920048706020769,
      "loss": 5.1787,
      "mean_token_accuracy": 0.19680802375078202,
      "num_tokens": 22425439.0,
      "step": 9775
    },
    {
      "entropy": 5.419207143783569,
      "epoch": 0.9394812680115274,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004919958101393817,
      "loss": 5.3085,
      "mean_token_accuracy": 0.1908087447285652,
      "num_tokens": 22436989.0,
      "step": 9780
    },
    {
      "entropy": 5.366714382171631,
      "epoch": 0.9399615754082613,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004919867446386883,
      "loss": 5.3077,
      "mean_token_accuracy": 0.19211360365152358,
      "num_tokens": 22449181.0,
      "step": 9785
    },
    {
      "entropy": 5.289137697219848,
      "epoch": 0.9404418828049952,
      "grad_norm": 1.46875,
      "learning_rate": 0.000491977674100207,
      "loss": 5.1525,
      "mean_token_accuracy": 0.19916170090436935,
      "num_tokens": 22460592.0,
      "step": 9790
    },
    {
      "entropy": 5.2588519096374515,
      "epoch": 0.9409221902017291,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004919685985241483,
      "loss": 5.1848,
      "mean_token_accuracy": 0.20611105859279633,
      "num_tokens": 22471718.0,
      "step": 9795
    },
    {
      "entropy": 5.424605131149292,
      "epoch": 0.9414024975984631,
      "grad_norm": 1.734375,
      "learning_rate": 0.000491959517910723,
      "loss": 5.2437,
      "mean_token_accuracy": 0.20368617475032808,
      "num_tokens": 22482413.0,
      "step": 9800
    },
    {
      "entropy": 5.346623706817627,
      "epoch": 0.9418828049951969,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004919504322601421,
      "loss": 5.1903,
      "mean_token_accuracy": 0.20130807608366014,
      "num_tokens": 22494016.0,
      "step": 9805
    },
    {
      "entropy": 5.361317920684814,
      "epoch": 0.9423631123919308,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004919413415726162,
      "loss": 5.2611,
      "mean_token_accuracy": 0.1888505131006241,
      "num_tokens": 22505015.0,
      "step": 9810
    },
    {
      "entropy": 5.337130784988403,
      "epoch": 0.9428434197886647,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004919322458483566,
      "loss": 5.2996,
      "mean_token_accuracy": 0.20140142887830734,
      "num_tokens": 22516691.0,
      "step": 9815
    },
    {
      "entropy": 5.384731149673462,
      "epoch": 0.9433237271853987,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004919231450875745,
      "loss": 5.2258,
      "mean_token_accuracy": 0.19500363767147064,
      "num_tokens": 22527740.0,
      "step": 9820
    },
    {
      "entropy": 5.368558311462403,
      "epoch": 0.9438040345821326,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004919140392904809,
      "loss": 5.2141,
      "mean_token_accuracy": 0.20168010741472245,
      "num_tokens": 22537685.0,
      "step": 9825
    },
    {
      "entropy": 5.386229372024536,
      "epoch": 0.9442843419788665,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004919049284572875,
      "loss": 5.1445,
      "mean_token_accuracy": 0.20708050578832626,
      "num_tokens": 22548282.0,
      "step": 9830
    },
    {
      "entropy": 5.261020803451538,
      "epoch": 0.9447646493756003,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004918958125882058,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20510132908821105,
      "num_tokens": 22559277.0,
      "step": 9835
    },
    {
      "entropy": 5.369067430496216,
      "epoch": 0.9452449567723343,
      "grad_norm": 1.375,
      "learning_rate": 0.0004918866916834474,
      "loss": 5.2547,
      "mean_token_accuracy": 0.19554455280303956,
      "num_tokens": 22570664.0,
      "step": 9840
    },
    {
      "entropy": 5.348180294036865,
      "epoch": 0.9457252641690682,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004918775657432239,
      "loss": 5.1785,
      "mean_token_accuracy": 0.20561182349920273,
      "num_tokens": 22580410.0,
      "step": 9845
    },
    {
      "entropy": 5.273010206222534,
      "epoch": 0.9462055715658021,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004918684347677474,
      "loss": 5.248,
      "mean_token_accuracy": 0.2004478693008423,
      "num_tokens": 22591630.0,
      "step": 9850
    },
    {
      "entropy": 5.381022787094116,
      "epoch": 0.946685878962536,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004918592987572298,
      "loss": 5.2362,
      "mean_token_accuracy": 0.19584041833877563,
      "num_tokens": 22603658.0,
      "step": 9855
    },
    {
      "entropy": 5.375761318206787,
      "epoch": 0.94716618635927,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004918501577118832,
      "loss": 5.2662,
      "mean_token_accuracy": 0.18672425597906112,
      "num_tokens": 22614812.0,
      "step": 9860
    },
    {
      "entropy": 5.396744203567505,
      "epoch": 0.9476464937560038,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00049184101163192,
      "loss": 5.1974,
      "mean_token_accuracy": 0.19715383946895598,
      "num_tokens": 22626924.0,
      "step": 9865
    },
    {
      "entropy": 5.398779392242432,
      "epoch": 0.9481268011527377,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004918318605175522,
      "loss": 5.2589,
      "mean_token_accuracy": 0.1908237174153328,
      "num_tokens": 22638715.0,
      "step": 9870
    },
    {
      "entropy": 5.488548278808594,
      "epoch": 0.9486071085494716,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004918227043689924,
      "loss": 5.3862,
      "mean_token_accuracy": 0.18225012123584747,
      "num_tokens": 22650257.0,
      "step": 9875
    },
    {
      "entropy": 5.359689664840698,
      "epoch": 0.9490874159462056,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004918135431864534,
      "loss": 5.2489,
      "mean_token_accuracy": 0.19737939685583114,
      "num_tokens": 22662480.0,
      "step": 9880
    },
    {
      "entropy": 5.343315935134887,
      "epoch": 0.9495677233429395,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004918043769701478,
      "loss": 5.241,
      "mean_token_accuracy": 0.19172091782093048,
      "num_tokens": 22673927.0,
      "step": 9885
    },
    {
      "entropy": 5.501454019546509,
      "epoch": 0.9500480307396734,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004917952057202882,
      "loss": 5.2992,
      "mean_token_accuracy": 0.19981193840503692,
      "num_tokens": 22684670.0,
      "step": 9890
    },
    {
      "entropy": 5.411140727996826,
      "epoch": 0.9505283381364072,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004917860294370877,
      "loss": 5.3555,
      "mean_token_accuracy": 0.1897347390651703,
      "num_tokens": 22696571.0,
      "step": 9895
    },
    {
      "entropy": 5.279243087768554,
      "epoch": 0.9510086455331412,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004917768481207593,
      "loss": 5.2021,
      "mean_token_accuracy": 0.20495676696300508,
      "num_tokens": 22708987.0,
      "step": 9900
    },
    {
      "entropy": 5.379358243942261,
      "epoch": 0.9514889529298751,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004917676617715162,
      "loss": 5.2821,
      "mean_token_accuracy": 0.2038042888045311,
      "num_tokens": 22720769.0,
      "step": 9905
    },
    {
      "entropy": 5.376408338546753,
      "epoch": 0.951969260326609,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004917584703895717,
      "loss": 5.2254,
      "mean_token_accuracy": 0.19668340682983398,
      "num_tokens": 22732592.0,
      "step": 9910
    },
    {
      "entropy": 5.390679979324341,
      "epoch": 0.952449567723343,
      "grad_norm": 1.375,
      "learning_rate": 0.0004917492739751391,
      "loss": 5.2539,
      "mean_token_accuracy": 0.18807051479816436,
      "num_tokens": 22744440.0,
      "step": 9915
    },
    {
      "entropy": 5.369322061538696,
      "epoch": 0.9529298751200769,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000491740072528432,
      "loss": 5.2226,
      "mean_token_accuracy": 0.197923880815506,
      "num_tokens": 22756336.0,
      "step": 9920
    },
    {
      "entropy": 5.345338249206543,
      "epoch": 0.9534101825168108,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000491730866049664,
      "loss": 5.2356,
      "mean_token_accuracy": 0.19126826375722886,
      "num_tokens": 22768256.0,
      "step": 9925
    },
    {
      "entropy": 5.47787389755249,
      "epoch": 0.9538904899135446,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004917216545390489,
      "loss": 5.2874,
      "mean_token_accuracy": 0.19381358623504638,
      "num_tokens": 22779693.0,
      "step": 9930
    },
    {
      "entropy": 5.253392314910888,
      "epoch": 0.9543707973102786,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004917124379968004,
      "loss": 5.1369,
      "mean_token_accuracy": 0.1992025136947632,
      "num_tokens": 22791417.0,
      "step": 9935
    },
    {
      "entropy": 5.3891936302185055,
      "epoch": 0.9548511047070125,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004917032164231327,
      "loss": 5.2193,
      "mean_token_accuracy": 0.19839159101247789,
      "num_tokens": 22803541.0,
      "step": 9940
    },
    {
      "entropy": 5.402176570892334,
      "epoch": 0.9553314121037464,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004916939898182598,
      "loss": 5.254,
      "mean_token_accuracy": 0.19071769267320632,
      "num_tokens": 22814648.0,
      "step": 9945
    },
    {
      "entropy": 5.364959383010865,
      "epoch": 0.9558117195004803,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004916847581823958,
      "loss": 5.2369,
      "mean_token_accuracy": 0.19924654811620712,
      "num_tokens": 22825672.0,
      "step": 9950
    },
    {
      "entropy": 5.4027454376220705,
      "epoch": 0.9562920268972143,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004916755215157552,
      "loss": 5.2172,
      "mean_token_accuracy": 0.19255754649639129,
      "num_tokens": 22837663.0,
      "step": 9955
    },
    {
      "entropy": 5.255894899368286,
      "epoch": 0.9567723342939481,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004916662798185524,
      "loss": 5.1179,
      "mean_token_accuracy": 0.2012002944946289,
      "num_tokens": 22849531.0,
      "step": 9960
    },
    {
      "entropy": 5.368396997451782,
      "epoch": 0.957252641690682,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004916570330910019,
      "loss": 5.2757,
      "mean_token_accuracy": 0.1924448788166046,
      "num_tokens": 22861784.0,
      "step": 9965
    },
    {
      "entropy": 5.451567268371582,
      "epoch": 0.9577329490874159,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004916477813333185,
      "loss": 5.2802,
      "mean_token_accuracy": 0.1922279790043831,
      "num_tokens": 22873482.0,
      "step": 9970
    },
    {
      "entropy": 5.371184873580932,
      "epoch": 0.9582132564841499,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004916385245457168,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20192530155181884,
      "num_tokens": 22883917.0,
      "step": 9975
    },
    {
      "entropy": 5.35193247795105,
      "epoch": 0.9586935638808838,
      "grad_norm": 1.359375,
      "learning_rate": 0.000491629262728412,
      "loss": 5.2348,
      "mean_token_accuracy": 0.19415521621704102,
      "num_tokens": 22896167.0,
      "step": 9980
    },
    {
      "entropy": 5.357525730133057,
      "epoch": 0.9591738712776177,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004916199958816188,
      "loss": 5.2704,
      "mean_token_accuracy": 0.19849079400300979,
      "num_tokens": 22908350.0,
      "step": 9985
    },
    {
      "entropy": 5.488919401168824,
      "epoch": 0.9596541786743515,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004916107240055527,
      "loss": 5.3672,
      "mean_token_accuracy": 0.19449115991592408,
      "num_tokens": 22918913.0,
      "step": 9990
    },
    {
      "entropy": 5.292144727706909,
      "epoch": 0.9601344860710855,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004916014471004287,
      "loss": 5.1802,
      "mean_token_accuracy": 0.20597919970750808,
      "num_tokens": 22929336.0,
      "step": 9995
    },
    {
      "entropy": 5.307368993759155,
      "epoch": 0.9606147934678194,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004915921651664622,
      "loss": 5.1944,
      "mean_token_accuracy": 0.19407544881105424,
      "num_tokens": 22940406.0,
      "step": 10000
    },
    {
      "entropy": 5.348668622970581,
      "epoch": 0.9610951008645533,
      "grad_norm": 1.53125,
      "learning_rate": 0.000491582878203869,
      "loss": 5.1823,
      "mean_token_accuracy": 0.19643348902463914,
      "num_tokens": 22951518.0,
      "step": 10005
    },
    {
      "entropy": 5.2929301261901855,
      "epoch": 0.9615754082612872,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004915735862128643,
      "loss": 5.1508,
      "mean_token_accuracy": 0.20273940414190292,
      "num_tokens": 22962309.0,
      "step": 10010
    },
    {
      "entropy": 5.376764822006225,
      "epoch": 0.9620557156580212,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004915642891936641,
      "loss": 5.2115,
      "mean_token_accuracy": 0.2020348146557808,
      "num_tokens": 22973996.0,
      "step": 10015
    },
    {
      "entropy": 5.398353338241577,
      "epoch": 0.962536023054755,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004915549871464841,
      "loss": 5.3102,
      "mean_token_accuracy": 0.19329437911510466,
      "num_tokens": 22984964.0,
      "step": 10020
    },
    {
      "entropy": 5.399173831939697,
      "epoch": 0.9630163304514889,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004915456800715403,
      "loss": 5.2797,
      "mean_token_accuracy": 0.18719930648803712,
      "num_tokens": 22996058.0,
      "step": 10025
    },
    {
      "entropy": 5.333802080154419,
      "epoch": 0.9634966378482228,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000491536367969049,
      "loss": 5.098,
      "mean_token_accuracy": 0.20807814449071885,
      "num_tokens": 23006140.0,
      "step": 10030
    },
    {
      "entropy": 5.364311218261719,
      "epoch": 0.9639769452449568,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004915270508392261,
      "loss": 5.3277,
      "mean_token_accuracy": 0.18995533883571625,
      "num_tokens": 23018129.0,
      "step": 10035
    },
    {
      "entropy": 5.436270141601563,
      "epoch": 0.9644572526416907,
      "grad_norm": 1.484375,
      "learning_rate": 0.000491517728682288,
      "loss": 5.1441,
      "mean_token_accuracy": 0.20244865864515305,
      "num_tokens": 23028673.0,
      "step": 10040
    },
    {
      "entropy": 5.351939487457275,
      "epoch": 0.9649375600384246,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004915084014984512,
      "loss": 5.2029,
      "mean_token_accuracy": 0.1979590028524399,
      "num_tokens": 23040778.0,
      "step": 10045
    },
    {
      "entropy": 5.296909141540527,
      "epoch": 0.9654178674351584,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004914990692879322,
      "loss": 5.2278,
      "mean_token_accuracy": 0.2034088686108589,
      "num_tokens": 23052170.0,
      "step": 10050
    },
    {
      "entropy": 5.358656215667724,
      "epoch": 0.9658981748318924,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004914897320509478,
      "loss": 5.2095,
      "mean_token_accuracy": 0.1957473948597908,
      "num_tokens": 23063455.0,
      "step": 10055
    },
    {
      "entropy": 5.30962061882019,
      "epoch": 0.9663784822286263,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004914803897877146,
      "loss": 5.1118,
      "mean_token_accuracy": 0.20433489382266998,
      "num_tokens": 23075009.0,
      "step": 10060
    },
    {
      "entropy": 5.370506811141968,
      "epoch": 0.9668587896253602,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004914710424984495,
      "loss": 5.2234,
      "mean_token_accuracy": 0.19421422183513642,
      "num_tokens": 23085921.0,
      "step": 10065
    },
    {
      "entropy": 5.466443157196045,
      "epoch": 0.9673390970220941,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004914616901833696,
      "loss": 5.368,
      "mean_token_accuracy": 0.18565791994333267,
      "num_tokens": 23098181.0,
      "step": 10070
    },
    {
      "entropy": 5.338989639282227,
      "epoch": 0.9678194044188281,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000491452332842692,
      "loss": 5.206,
      "mean_token_accuracy": 0.1956307664513588,
      "num_tokens": 23109904.0,
      "step": 10075
    },
    {
      "entropy": 5.347525930404663,
      "epoch": 0.968299711815562,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000491442970476634,
      "loss": 5.2892,
      "mean_token_accuracy": 0.19307613521814346,
      "num_tokens": 23122356.0,
      "step": 10080
    },
    {
      "entropy": 5.367475271224976,
      "epoch": 0.9687800192122958,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004914336030854129,
      "loss": 5.1811,
      "mean_token_accuracy": 0.20187202841043472,
      "num_tokens": 23134055.0,
      "step": 10085
    },
    {
      "entropy": 5.310868787765503,
      "epoch": 0.9692603266090298,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004914242306692461,
      "loss": 5.1533,
      "mean_token_accuracy": 0.20374749451875687,
      "num_tokens": 23145388.0,
      "step": 10090
    },
    {
      "entropy": 5.36150484085083,
      "epoch": 0.9697406340057637,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004914148532283516,
      "loss": 5.2755,
      "mean_token_accuracy": 0.1874200463294983,
      "num_tokens": 23157083.0,
      "step": 10095
    },
    {
      "entropy": 5.380633878707886,
      "epoch": 0.9702209414024976,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004914054707629466,
      "loss": 5.2226,
      "mean_token_accuracy": 0.19765482544898988,
      "num_tokens": 23169154.0,
      "step": 10100
    },
    {
      "entropy": 5.318311738967895,
      "epoch": 0.9707012487992315,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004913960832732493,
      "loss": 5.1199,
      "mean_token_accuracy": 0.20090494006872178,
      "num_tokens": 23179879.0,
      "step": 10105
    },
    {
      "entropy": 5.316896104812622,
      "epoch": 0.9711815561959655,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004913866907594774,
      "loss": 5.24,
      "mean_token_accuracy": 0.19927761256694793,
      "num_tokens": 23190632.0,
      "step": 10110
    },
    {
      "entropy": 5.441259956359863,
      "epoch": 0.9716618635926993,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004913772932218491,
      "loss": 5.288,
      "mean_token_accuracy": 0.1860272616147995,
      "num_tokens": 23202060.0,
      "step": 10115
    },
    {
      "entropy": 5.429521894454956,
      "epoch": 0.9721421709894332,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004913678906605825,
      "loss": 5.2725,
      "mean_token_accuracy": 0.1872340440750122,
      "num_tokens": 23214103.0,
      "step": 10120
    },
    {
      "entropy": 5.358136367797852,
      "epoch": 0.9726224783861671,
      "grad_norm": 1.625,
      "learning_rate": 0.0004913584830758961,
      "loss": 5.2168,
      "mean_token_accuracy": 0.20135337710380555,
      "num_tokens": 23224676.0,
      "step": 10125
    },
    {
      "entropy": 5.388266372680664,
      "epoch": 0.9731027857829011,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004913490704680081,
      "loss": 5.2382,
      "mean_token_accuracy": 0.20236930549144744,
      "num_tokens": 23235478.0,
      "step": 10130
    },
    {
      "entropy": 5.410058832168579,
      "epoch": 0.973583093179635,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004913396528371371,
      "loss": 5.2772,
      "mean_token_accuracy": 0.19678158015012742,
      "num_tokens": 23246913.0,
      "step": 10135
    },
    {
      "entropy": 5.343859481811523,
      "epoch": 0.9740634005763689,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004913302301835018,
      "loss": 5.2407,
      "mean_token_accuracy": 0.1970997631549835,
      "num_tokens": 23258472.0,
      "step": 10140
    },
    {
      "entropy": 5.438864469528198,
      "epoch": 0.9745437079731027,
      "grad_norm": 2.75,
      "learning_rate": 0.000491320802507321,
      "loss": 5.3598,
      "mean_token_accuracy": 0.18568275570869447,
      "num_tokens": 23270443.0,
      "step": 10145
    },
    {
      "entropy": 5.4093412399292,
      "epoch": 0.9750240153698367,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004913113698088133,
      "loss": 5.2069,
      "mean_token_accuracy": 0.20590991377830506,
      "num_tokens": 23281199.0,
      "step": 10150
    },
    {
      "entropy": 5.296328687667847,
      "epoch": 0.9755043227665706,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000491301932088198,
      "loss": 5.2072,
      "mean_token_accuracy": 0.19928632229566573,
      "num_tokens": 23291980.0,
      "step": 10155
    },
    {
      "entropy": 5.439354133605957,
      "epoch": 0.9759846301633045,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004912924893456942,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1918912336230278,
      "num_tokens": 23303673.0,
      "step": 10160
    },
    {
      "entropy": 5.50971531867981,
      "epoch": 0.9764649375600384,
      "grad_norm": 1.5625,
      "learning_rate": 0.000491283041581521,
      "loss": 5.2719,
      "mean_token_accuracy": 0.1940137818455696,
      "num_tokens": 23315830.0,
      "step": 10165
    },
    {
      "entropy": 5.426065301895141,
      "epoch": 0.9769452449567724,
      "grad_norm": 13.4375,
      "learning_rate": 0.0004912735887958978,
      "loss": 5.3282,
      "mean_token_accuracy": 0.18928514122962953,
      "num_tokens": 23329860.0,
      "step": 10170
    },
    {
      "entropy": 5.334433317184448,
      "epoch": 0.9774255523535063,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004912641309890441,
      "loss": 5.2079,
      "mean_token_accuracy": 0.20020253956317902,
      "num_tokens": 23340610.0,
      "step": 10175
    },
    {
      "entropy": 5.337919902801514,
      "epoch": 0.9779058597502401,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004912546681611794,
      "loss": 5.2325,
      "mean_token_accuracy": 0.19779051393270491,
      "num_tokens": 23352153.0,
      "step": 10180
    },
    {
      "entropy": 5.288594341278076,
      "epoch": 0.978386167146974,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004912452003125234,
      "loss": 5.1663,
      "mean_token_accuracy": 0.20293716341257095,
      "num_tokens": 23363630.0,
      "step": 10185
    },
    {
      "entropy": 5.359788608551026,
      "epoch": 0.978866474543708,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000491235727443296,
      "loss": 5.1841,
      "mean_token_accuracy": 0.19922154992818833,
      "num_tokens": 23375922.0,
      "step": 10190
    },
    {
      "entropy": 5.327295780181885,
      "epoch": 0.9793467819404419,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004912262495537171,
      "loss": 5.1448,
      "mean_token_accuracy": 0.20141881704330444,
      "num_tokens": 23386446.0,
      "step": 10195
    },
    {
      "entropy": 5.300418710708618,
      "epoch": 0.9798270893371758,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004912167666440068,
      "loss": 5.1345,
      "mean_token_accuracy": 0.20769673734903335,
      "num_tokens": 23398458.0,
      "step": 10200
    },
    {
      "entropy": 5.324464321136475,
      "epoch": 0.9803073967339097,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004912072787143852,
      "loss": 5.1996,
      "mean_token_accuracy": 0.19903348833322526,
      "num_tokens": 23409725.0,
      "step": 10205
    },
    {
      "entropy": 5.3119487285614015,
      "epoch": 0.9807877041306436,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004911977857650725,
      "loss": 5.236,
      "mean_token_accuracy": 0.20016496926546096,
      "num_tokens": 23420468.0,
      "step": 10210
    },
    {
      "entropy": 5.387360954284668,
      "epoch": 0.9812680115273775,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004911882877962893,
      "loss": 5.2133,
      "mean_token_accuracy": 0.19513286501169205,
      "num_tokens": 23431386.0,
      "step": 10215
    },
    {
      "entropy": 5.440367603302002,
      "epoch": 0.9817483189241114,
      "grad_norm": 3.1875,
      "learning_rate": 0.0004911787848082559,
      "loss": 5.2552,
      "mean_token_accuracy": 0.19321776330471038,
      "num_tokens": 23442759.0,
      "step": 10220
    },
    {
      "entropy": 5.381049060821534,
      "epoch": 0.9822286263208453,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004911692768011931,
      "loss": 5.1918,
      "mean_token_accuracy": 0.2051030680537224,
      "num_tokens": 23454372.0,
      "step": 10225
    },
    {
      "entropy": 5.42848687171936,
      "epoch": 0.9827089337175793,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004911597637753217,
      "loss": 5.279,
      "mean_token_accuracy": 0.19997452348470687,
      "num_tokens": 23465557.0,
      "step": 10230
    },
    {
      "entropy": 5.286953687667847,
      "epoch": 0.9831892411143132,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004911502457308623,
      "loss": 5.0811,
      "mean_token_accuracy": 0.20363888442516326,
      "num_tokens": 23475850.0,
      "step": 10235
    },
    {
      "entropy": 5.30987434387207,
      "epoch": 0.983669548511047,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000491140722668036,
      "loss": 5.1328,
      "mean_token_accuracy": 0.2009579584002495,
      "num_tokens": 23487145.0,
      "step": 10240
    },
    {
      "entropy": 5.346234607696533,
      "epoch": 0.984149855907781,
      "grad_norm": 1.59375,
      "learning_rate": 0.000491131194587064,
      "loss": 5.2324,
      "mean_token_accuracy": 0.19113098382949828,
      "num_tokens": 23498112.0,
      "step": 10245
    },
    {
      "entropy": 5.351706504821777,
      "epoch": 0.9846301633045149,
      "grad_norm": 1.625,
      "learning_rate": 0.0004911216614881675,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20384518206119537,
      "num_tokens": 23509865.0,
      "step": 10250
    },
    {
      "entropy": 5.413609361648559,
      "epoch": 0.9851104707012488,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004911121233715677,
      "loss": 5.3172,
      "mean_token_accuracy": 0.18676220774650573,
      "num_tokens": 23521153.0,
      "step": 10255
    },
    {
      "entropy": 5.362878274917603,
      "epoch": 0.9855907780979827,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004911025802374861,
      "loss": 5.3476,
      "mean_token_accuracy": 0.19076319932937622,
      "num_tokens": 23532372.0,
      "step": 10260
    },
    {
      "entropy": 5.381707096099854,
      "epoch": 0.9860710854947167,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004910930320861442,
      "loss": 5.3006,
      "mean_token_accuracy": 0.19311443865299224,
      "num_tokens": 23543494.0,
      "step": 10265
    },
    {
      "entropy": 5.4404397964477536,
      "epoch": 0.9865513928914506,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004910834789177639,
      "loss": 5.2111,
      "mean_token_accuracy": 0.1928661823272705,
      "num_tokens": 23554020.0,
      "step": 10270
    },
    {
      "entropy": 5.481026268005371,
      "epoch": 0.9870317002881844,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004910739207325668,
      "loss": 5.3725,
      "mean_token_accuracy": 0.1903602033853531,
      "num_tokens": 23565032.0,
      "step": 10275
    },
    {
      "entropy": 5.308665561676025,
      "epoch": 0.9875120076849183,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004910643575307749,
      "loss": 5.2366,
      "mean_token_accuracy": 0.19494978934526444,
      "num_tokens": 23576243.0,
      "step": 10280
    },
    {
      "entropy": 5.369306802749634,
      "epoch": 0.9879923150816523,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004910547893126102,
      "loss": 5.281,
      "mean_token_accuracy": 0.19268425107002257,
      "num_tokens": 23586111.0,
      "step": 10285
    },
    {
      "entropy": 5.457032108306885,
      "epoch": 0.9884726224783862,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004910452160782948,
      "loss": 5.3066,
      "mean_token_accuracy": 0.19325252920389174,
      "num_tokens": 23597718.0,
      "step": 10290
    },
    {
      "entropy": 5.493101644515991,
      "epoch": 0.9889529298751201,
      "grad_norm": 2.046875,
      "learning_rate": 0.000491035637828051,
      "loss": 5.2978,
      "mean_token_accuracy": 0.18808236122131347,
      "num_tokens": 23608241.0,
      "step": 10295
    },
    {
      "entropy": 5.420267486572266,
      "epoch": 0.989433237271854,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004910260545621012,
      "loss": 5.2874,
      "mean_token_accuracy": 0.19275065660476684,
      "num_tokens": 23619120.0,
      "step": 10300
    },
    {
      "entropy": 5.32737455368042,
      "epoch": 0.9899135446685879,
      "grad_norm": 2.125,
      "learning_rate": 0.0004910164662806679,
      "loss": 5.1483,
      "mean_token_accuracy": 0.20073831379413604,
      "num_tokens": 23630642.0,
      "step": 10305
    },
    {
      "entropy": 5.328326606750489,
      "epoch": 0.9903938520653218,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004910068729839736,
      "loss": 5.1964,
      "mean_token_accuracy": 0.20349667221307755,
      "num_tokens": 23641987.0,
      "step": 10310
    },
    {
      "entropy": 5.42763614654541,
      "epoch": 0.9908741594620557,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004909972746722413,
      "loss": 5.339,
      "mean_token_accuracy": 0.19256781339645385,
      "num_tokens": 23653144.0,
      "step": 10315
    },
    {
      "entropy": 5.386298656463623,
      "epoch": 0.9913544668587896,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004909876713456935,
      "loss": 5.2544,
      "mean_token_accuracy": 0.19387101531028747,
      "num_tokens": 23663807.0,
      "step": 10320
    },
    {
      "entropy": 5.3090744972229,
      "epoch": 0.9918347742555236,
      "grad_norm": 2.6875,
      "learning_rate": 0.0004909780630045534,
      "loss": 5.1819,
      "mean_token_accuracy": 0.1971955418586731,
      "num_tokens": 23674341.0,
      "step": 10325
    },
    {
      "entropy": 5.359406900405884,
      "epoch": 0.9923150816522575,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000490968449649044,
      "loss": 5.2883,
      "mean_token_accuracy": 0.19592582285404206,
      "num_tokens": 23685415.0,
      "step": 10330
    },
    {
      "entropy": 5.4454261302948,
      "epoch": 0.9927953890489913,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004909588312793884,
      "loss": 5.2819,
      "mean_token_accuracy": 0.19722168296575546,
      "num_tokens": 23696652.0,
      "step": 10335
    },
    {
      "entropy": 5.3882434368133545,
      "epoch": 0.9932756964457252,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004909492078958101,
      "loss": 5.2237,
      "mean_token_accuracy": 0.19429783523082733,
      "num_tokens": 23708939.0,
      "step": 10340
    },
    {
      "entropy": 5.369251489639282,
      "epoch": 0.9937560038424592,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004909395794985324,
      "loss": 5.2102,
      "mean_token_accuracy": 0.1963334009051323,
      "num_tokens": 23720587.0,
      "step": 10345
    },
    {
      "entropy": 5.279649543762207,
      "epoch": 0.9942363112391931,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004909299460877788,
      "loss": 5.1716,
      "mean_token_accuracy": 0.19969542175531388,
      "num_tokens": 23732345.0,
      "step": 10350
    },
    {
      "entropy": 5.382672357559204,
      "epoch": 0.994716618635927,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004909203076637732,
      "loss": 5.2306,
      "mean_token_accuracy": 0.19261985272169113,
      "num_tokens": 23743149.0,
      "step": 10355
    },
    {
      "entropy": 5.4967328071594235,
      "epoch": 0.9951969260326609,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004909106642267392,
      "loss": 5.2627,
      "mean_token_accuracy": 0.198738257586956,
      "num_tokens": 23755508.0,
      "step": 10360
    },
    {
      "entropy": 5.356519365310669,
      "epoch": 0.9956772334293948,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004909010157769006,
      "loss": 5.1287,
      "mean_token_accuracy": 0.1936712607741356,
      "num_tokens": 23766313.0,
      "step": 10365
    },
    {
      "entropy": 5.263210487365723,
      "epoch": 0.9961575408261287,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004908913623144814,
      "loss": 5.1291,
      "mean_token_accuracy": 0.2092423692345619,
      "num_tokens": 23777776.0,
      "step": 10370
    },
    {
      "entropy": 5.378554487228394,
      "epoch": 0.9966378482228626,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000490881703839706,
      "loss": 5.3086,
      "mean_token_accuracy": 0.19899863004684448,
      "num_tokens": 23791143.0,
      "step": 10375
    },
    {
      "entropy": 5.503311586380005,
      "epoch": 0.9971181556195965,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004908720403527984,
      "loss": 5.2785,
      "mean_token_accuracy": 0.1919129014015198,
      "num_tokens": 23802446.0,
      "step": 10380
    },
    {
      "entropy": 5.299834775924682,
      "epoch": 0.9975984630163305,
      "grad_norm": 1.453125,
      "learning_rate": 0.000490862371853983,
      "loss": 5.1435,
      "mean_token_accuracy": 0.20518070012331008,
      "num_tokens": 23813757.0,
      "step": 10385
    },
    {
      "entropy": 5.311198663711548,
      "epoch": 0.9980787704130644,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004908526983434844,
      "loss": 5.2307,
      "mean_token_accuracy": 0.1979399248957634,
      "num_tokens": 23824934.0,
      "step": 10390
    },
    {
      "entropy": 5.440586757659912,
      "epoch": 0.9985590778097982,
      "grad_norm": 4.3125,
      "learning_rate": 0.000490843019821527,
      "loss": 5.3278,
      "mean_token_accuracy": 0.1918754205107689,
      "num_tokens": 23836872.0,
      "step": 10395
    },
    {
      "entropy": 5.490906286239624,
      "epoch": 0.9990393852065321,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004908333362883358,
      "loss": 5.2426,
      "mean_token_accuracy": 0.19100466817617417,
      "num_tokens": 23847853.0,
      "step": 10400
    },
    {
      "entropy": 5.298523759841919,
      "epoch": 0.9995196926032661,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004908236477441353,
      "loss": 5.1478,
      "mean_token_accuracy": 0.2047051355242729,
      "num_tokens": 23859208.0,
      "step": 10405
    },
    {
      "entropy": 5.25507493019104,
      "epoch": 1.0,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004908139541891505,
      "loss": 5.117,
      "mean_token_accuracy": 0.1997172936797142,
      "num_tokens": 23868536.0,
      "step": 10410
    },
    {
      "entropy": 5.35916576385498,
      "epoch": 1.0004803073967339,
      "grad_norm": 2.5,
      "learning_rate": 0.0004908042556236066,
      "loss": 5.1458,
      "mean_token_accuracy": 0.1983928233385086,
      "num_tokens": 23881335.0,
      "step": 10415
    },
    {
      "entropy": 5.30327844619751,
      "epoch": 1.0009606147934678,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004907945520477286,
      "loss": 5.0577,
      "mean_token_accuracy": 0.20289318710565568,
      "num_tokens": 23891694.0,
      "step": 10420
    },
    {
      "entropy": 5.288583993911743,
      "epoch": 1.0014409221902016,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004907848434617419,
      "loss": 5.1372,
      "mean_token_accuracy": 0.19711353480815888,
      "num_tokens": 23903560.0,
      "step": 10425
    },
    {
      "entropy": 5.47208800315857,
      "epoch": 1.0019212295869357,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000490775129865872,
      "loss": 5.3221,
      "mean_token_accuracy": 0.18646362125873567,
      "num_tokens": 23914525.0,
      "step": 10430
    },
    {
      "entropy": 5.452016878128052,
      "epoch": 1.0024015369836696,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004907654112603442,
      "loss": 5.1886,
      "mean_token_accuracy": 0.20110237300395967,
      "num_tokens": 23926492.0,
      "step": 10435
    },
    {
      "entropy": 5.402464389801025,
      "epoch": 1.0028818443804035,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004907556876453843,
      "loss": 5.1648,
      "mean_token_accuracy": 0.19879360347986222,
      "num_tokens": 23936326.0,
      "step": 10440
    },
    {
      "entropy": 5.201600027084351,
      "epoch": 1.0033621517771374,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000490745959021218,
      "loss": 4.9686,
      "mean_token_accuracy": 0.21124709844589235,
      "num_tokens": 23947173.0,
      "step": 10445
    },
    {
      "entropy": 5.371070861816406,
      "epoch": 1.0038424591738713,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004907362253880711,
      "loss": 5.1493,
      "mean_token_accuracy": 0.2001799076795578,
      "num_tokens": 23959771.0,
      "step": 10450
    },
    {
      "entropy": 5.351351308822632,
      "epoch": 1.0043227665706052,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004907264867461697,
      "loss": 5.2257,
      "mean_token_accuracy": 0.19711594581604003,
      "num_tokens": 23970908.0,
      "step": 10455
    },
    {
      "entropy": 5.33805136680603,
      "epoch": 1.004803073967339,
      "grad_norm": 3.953125,
      "learning_rate": 0.0004907167430957399,
      "loss": 5.1381,
      "mean_token_accuracy": 0.19941968470811844,
      "num_tokens": 23981640.0,
      "step": 10460
    },
    {
      "entropy": 5.437816190719604,
      "epoch": 1.005283381364073,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004907069944370077,
      "loss": 5.1679,
      "mean_token_accuracy": 0.19937537610530853,
      "num_tokens": 23992913.0,
      "step": 10465
    },
    {
      "entropy": 5.343357515335083,
      "epoch": 1.005763688760807,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004906972407701998,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19388407915830613,
      "num_tokens": 24003819.0,
      "step": 10470
    },
    {
      "entropy": 5.362512111663818,
      "epoch": 1.006243996157541,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004906874820955423,
      "loss": 5.1722,
      "mean_token_accuracy": 0.19612828493118287,
      "num_tokens": 24013997.0,
      "step": 10475
    },
    {
      "entropy": 5.324533557891845,
      "epoch": 1.0067243035542748,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004906777184132621,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19982564747333526,
      "num_tokens": 24025244.0,
      "step": 10480
    },
    {
      "entropy": 5.258550691604614,
      "epoch": 1.0072046109510087,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004906679497235856,
      "loss": 5.1285,
      "mean_token_accuracy": 0.19847694635391236,
      "num_tokens": 24037701.0,
      "step": 10485
    },
    {
      "entropy": 5.407458353042602,
      "epoch": 1.0076849183477425,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004906581760267397,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19189472645521163,
      "num_tokens": 24049369.0,
      "step": 10490
    },
    {
      "entropy": 5.356171464920044,
      "epoch": 1.0081652257444764,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004906483973229513,
      "loss": 5.1522,
      "mean_token_accuracy": 0.19589548408985138,
      "num_tokens": 24060625.0,
      "step": 10495
    },
    {
      "entropy": 5.423621988296508,
      "epoch": 1.0086455331412103,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004906386136124476,
      "loss": 5.1949,
      "mean_token_accuracy": 0.19404211193323134,
      "num_tokens": 24072527.0,
      "step": 10500
    },
    {
      "entropy": 5.402671384811401,
      "epoch": 1.0091258405379442,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004906288248954554,
      "loss": 5.2194,
      "mean_token_accuracy": 0.19060073643922806,
      "num_tokens": 24083979.0,
      "step": 10505
    },
    {
      "entropy": 5.330675411224365,
      "epoch": 1.0096061479346783,
      "grad_norm": 1.5,
      "learning_rate": 0.0004906190311722023,
      "loss": 5.1869,
      "mean_token_accuracy": 0.20352004915475846,
      "num_tokens": 24095926.0,
      "step": 10510
    },
    {
      "entropy": 5.304249668121338,
      "epoch": 1.0100864553314122,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004906092324429155,
      "loss": 5.0919,
      "mean_token_accuracy": 0.205551378428936,
      "num_tokens": 24107898.0,
      "step": 10515
    },
    {
      "entropy": 5.318305540084839,
      "epoch": 1.010566762728146,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004905994287078227,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20388976484537125,
      "num_tokens": 24118713.0,
      "step": 10520
    },
    {
      "entropy": 5.376314926147461,
      "epoch": 1.01104707012488,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004905896199671512,
      "loss": 5.1724,
      "mean_token_accuracy": 0.20016107857227325,
      "num_tokens": 24130410.0,
      "step": 10525
    },
    {
      "entropy": 5.379752540588379,
      "epoch": 1.0115273775216138,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000490579806221129,
      "loss": 5.1599,
      "mean_token_accuracy": 0.19758369624614716,
      "num_tokens": 24140961.0,
      "step": 10530
    },
    {
      "entropy": 5.407641839981079,
      "epoch": 1.0120076849183477,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004905699874699838,
      "loss": 5.2384,
      "mean_token_accuracy": 0.19086066782474517,
      "num_tokens": 24153467.0,
      "step": 10535
    },
    {
      "entropy": 5.302155494689941,
      "epoch": 1.0124879923150816,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004905601637139436,
      "loss": 5.0573,
      "mean_token_accuracy": 0.20565125942230225,
      "num_tokens": 24163873.0,
      "step": 10540
    },
    {
      "entropy": 5.225621843338013,
      "epoch": 1.0129682997118155,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004905503349532365,
      "loss": 5.1595,
      "mean_token_accuracy": 0.20568412989377977,
      "num_tokens": 24175852.0,
      "step": 10545
    },
    {
      "entropy": 5.357414054870605,
      "epoch": 1.0134486071085496,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004905405011880906,
      "loss": 5.2056,
      "mean_token_accuracy": 0.19606508165597916,
      "num_tokens": 24187759.0,
      "step": 10550
    },
    {
      "entropy": 5.385495281219482,
      "epoch": 1.0139289145052834,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004905306624187343,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20722429007291793,
      "num_tokens": 24199725.0,
      "step": 10555
    },
    {
      "entropy": 5.41778073310852,
      "epoch": 1.0144092219020173,
      "grad_norm": 1.625,
      "learning_rate": 0.0004905208186453961,
      "loss": 5.2205,
      "mean_token_accuracy": 0.1956380143761635,
      "num_tokens": 24210739.0,
      "step": 10560
    },
    {
      "entropy": 5.416292524337768,
      "epoch": 1.0148895292987512,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004905109698683044,
      "loss": 5.3134,
      "mean_token_accuracy": 0.18740409463644028,
      "num_tokens": 24221873.0,
      "step": 10565
    },
    {
      "entropy": 5.332883071899414,
      "epoch": 1.015369836695485,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004905011160876878,
      "loss": 5.1674,
      "mean_token_accuracy": 0.19635324627161027,
      "num_tokens": 24233503.0,
      "step": 10570
    },
    {
      "entropy": 5.334778261184693,
      "epoch": 1.015850144092219,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004904912573037753,
      "loss": 5.1363,
      "mean_token_accuracy": 0.19603917598724366,
      "num_tokens": 24244465.0,
      "step": 10575
    },
    {
      "entropy": 5.276717329025269,
      "epoch": 1.0163304514889528,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004904813935167957,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20925995111465454,
      "num_tokens": 24255807.0,
      "step": 10580
    },
    {
      "entropy": 5.411048841476441,
      "epoch": 1.0168107588856867,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004904715247269779,
      "loss": 5.2579,
      "mean_token_accuracy": 0.18983708173036576,
      "num_tokens": 24268195.0,
      "step": 10585
    },
    {
      "entropy": 5.2618835926055905,
      "epoch": 1.0172910662824208,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004904616509345514,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20803383737802505,
      "num_tokens": 24278915.0,
      "step": 10590
    },
    {
      "entropy": 5.349010944366455,
      "epoch": 1.0177713736791547,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004904517721397449,
      "loss": 5.1807,
      "mean_token_accuracy": 0.20545896142721176,
      "num_tokens": 24289115.0,
      "step": 10595
    },
    {
      "entropy": 5.3358502864837645,
      "epoch": 1.0182516810758886,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004904418883427881,
      "loss": 5.1745,
      "mean_token_accuracy": 0.19754030108451842,
      "num_tokens": 24300646.0,
      "step": 10600
    },
    {
      "entropy": 5.518046188354492,
      "epoch": 1.0187319884726225,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004904319995439104,
      "loss": 5.2938,
      "mean_token_accuracy": 0.1906458780169487,
      "num_tokens": 24312344.0,
      "step": 10605
    },
    {
      "entropy": 5.3027870655059814,
      "epoch": 1.0192122958693564,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004904221057433412,
      "loss": 5.0919,
      "mean_token_accuracy": 0.20173458606004716,
      "num_tokens": 24323382.0,
      "step": 10610
    },
    {
      "entropy": 5.296759748458863,
      "epoch": 1.0196926032660902,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004904122069413105,
      "loss": 5.1023,
      "mean_token_accuracy": 0.19988326281309127,
      "num_tokens": 24336820.0,
      "step": 10615
    },
    {
      "entropy": 5.396545362472534,
      "epoch": 1.0201729106628241,
      "grad_norm": 1.796875,
      "learning_rate": 0.000490402303138048,
      "loss": 5.1375,
      "mean_token_accuracy": 0.20109147131443023,
      "num_tokens": 24348547.0,
      "step": 10620
    },
    {
      "entropy": 5.443462705612182,
      "epoch": 1.0206532180595582,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004903923943337836,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19088246300816536,
      "num_tokens": 24359586.0,
      "step": 10625
    },
    {
      "entropy": 5.390566301345825,
      "epoch": 1.021133525456292,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004903824805287475,
      "loss": 5.2283,
      "mean_token_accuracy": 0.1936145007610321,
      "num_tokens": 24371647.0,
      "step": 10630
    },
    {
      "entropy": 5.3553379535675045,
      "epoch": 1.021613832853026,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004903725617231696,
      "loss": 5.0796,
      "mean_token_accuracy": 0.19957351982593535,
      "num_tokens": 24382308.0,
      "step": 10635
    },
    {
      "entropy": 5.319635534286499,
      "epoch": 1.0220941402497599,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004903626379172805,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19824805557727815,
      "num_tokens": 24393224.0,
      "step": 10640
    },
    {
      "entropy": 5.294248247146607,
      "epoch": 1.0225744476464937,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004903527091113102,
      "loss": 5.181,
      "mean_token_accuracy": 0.19135986715555192,
      "num_tokens": 24405220.0,
      "step": 10645
    },
    {
      "entropy": 5.418373441696167,
      "epoch": 1.0230547550432276,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004903427753054897,
      "loss": 5.2016,
      "mean_token_accuracy": 0.19548535495996475,
      "num_tokens": 24416406.0,
      "step": 10650
    },
    {
      "entropy": 5.381947326660156,
      "epoch": 1.0235350624399615,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004903328365000492,
      "loss": 5.2388,
      "mean_token_accuracy": 0.19457056373357773,
      "num_tokens": 24428595.0,
      "step": 10655
    },
    {
      "entropy": 5.353590583801269,
      "epoch": 1.0240153698366954,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004903228926952199,
      "loss": 5.1115,
      "mean_token_accuracy": 0.20169458091259002,
      "num_tokens": 24439568.0,
      "step": 10660
    },
    {
      "entropy": 5.366839456558227,
      "epoch": 1.0244956772334295,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004903129438912322,
      "loss": 5.178,
      "mean_token_accuracy": 0.20210112780332565,
      "num_tokens": 24451736.0,
      "step": 10665
    },
    {
      "entropy": 5.327310609817505,
      "epoch": 1.0249759846301634,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004903029900883174,
      "loss": 5.1391,
      "mean_token_accuracy": 0.20306526720523835,
      "num_tokens": 24463463.0,
      "step": 10670
    },
    {
      "entropy": 5.331499528884888,
      "epoch": 1.0254562920268973,
      "grad_norm": 2.125,
      "learning_rate": 0.0004902930312867063,
      "loss": 5.1404,
      "mean_token_accuracy": 0.19781286865472794,
      "num_tokens": 24474425.0,
      "step": 10675
    },
    {
      "entropy": 5.383645009994507,
      "epoch": 1.0259365994236311,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004902830674866306,
      "loss": 5.1152,
      "mean_token_accuracy": 0.20142365843057633,
      "num_tokens": 24486482.0,
      "step": 10680
    },
    {
      "entropy": 5.339239406585693,
      "epoch": 1.026416906820365,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004902730986883211,
      "loss": 5.1245,
      "mean_token_accuracy": 0.19678469002246857,
      "num_tokens": 24497709.0,
      "step": 10685
    },
    {
      "entropy": 5.309704351425171,
      "epoch": 1.026897214217099,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004902631248920096,
      "loss": 5.0944,
      "mean_token_accuracy": 0.20305254310369492,
      "num_tokens": 24508406.0,
      "step": 10690
    },
    {
      "entropy": 5.271453809738159,
      "epoch": 1.0273775216138328,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004902531460979274,
      "loss": 5.1724,
      "mean_token_accuracy": 0.19548801630735396,
      "num_tokens": 24520688.0,
      "step": 10695
    },
    {
      "entropy": 5.267741632461548,
      "epoch": 1.0278578290105667,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004902431623063065,
      "loss": 5.0636,
      "mean_token_accuracy": 0.20485226362943648,
      "num_tokens": 24531473.0,
      "step": 10700
    },
    {
      "entropy": 5.3205039501190186,
      "epoch": 1.0283381364073008,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004902331735173785,
      "loss": 5.1765,
      "mean_token_accuracy": 0.19298290014266967,
      "num_tokens": 24543263.0,
      "step": 10705
    },
    {
      "entropy": 5.366945123672485,
      "epoch": 1.0288184438040346,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004902231797313752,
      "loss": 5.0712,
      "mean_token_accuracy": 0.20559680014848708,
      "num_tokens": 24554067.0,
      "step": 10710
    },
    {
      "entropy": 5.235648059844971,
      "epoch": 1.0292987512007685,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004902131809485288,
      "loss": 5.0856,
      "mean_token_accuracy": 0.2079134076833725,
      "num_tokens": 24566291.0,
      "step": 10715
    },
    {
      "entropy": 5.348055839538574,
      "epoch": 1.0297790585975024,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004902031771690713,
      "loss": 5.2217,
      "mean_token_accuracy": 0.19839471727609634,
      "num_tokens": 24579248.0,
      "step": 10720
    },
    {
      "entropy": 5.304089164733886,
      "epoch": 1.0302593659942363,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004901931683932352,
      "loss": 5.0978,
      "mean_token_accuracy": 0.2068445473909378,
      "num_tokens": 24591047.0,
      "step": 10725
    },
    {
      "entropy": 5.428918838500977,
      "epoch": 1.0307396733909702,
      "grad_norm": 1.5,
      "learning_rate": 0.0004901831546212526,
      "loss": 5.1683,
      "mean_token_accuracy": 0.19905165433883668,
      "num_tokens": 24602454.0,
      "step": 10730
    },
    {
      "entropy": 5.348472166061401,
      "epoch": 1.031219980787704,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004901731358533562,
      "loss": 5.1651,
      "mean_token_accuracy": 0.20431284755468368,
      "num_tokens": 24614168.0,
      "step": 10735
    },
    {
      "entropy": 5.320114803314209,
      "epoch": 1.031700288184438,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004901631120897785,
      "loss": 5.1545,
      "mean_token_accuracy": 0.2016324982047081,
      "num_tokens": 24625588.0,
      "step": 10740
    },
    {
      "entropy": 5.336018323898315,
      "epoch": 1.032180595581172,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004901530833307522,
      "loss": 5.1301,
      "mean_token_accuracy": 0.19757412523031234,
      "num_tokens": 24637744.0,
      "step": 10745
    },
    {
      "entropy": 5.2790380954742435,
      "epoch": 1.032660902977906,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004901430495765103,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20606896728277208,
      "num_tokens": 24649631.0,
      "step": 10750
    },
    {
      "entropy": 5.284367513656616,
      "epoch": 1.0331412103746398,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004901330108272855,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20179992616176606,
      "num_tokens": 24660622.0,
      "step": 10755
    },
    {
      "entropy": 5.371030998229981,
      "epoch": 1.0336215177713737,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004901229670833111,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20080428719520568,
      "num_tokens": 24671072.0,
      "step": 10760
    },
    {
      "entropy": 5.308475732803345,
      "epoch": 1.0341018251681076,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004901129183448201,
      "loss": 5.1889,
      "mean_token_accuracy": 0.19608522802591324,
      "num_tokens": 24683160.0,
      "step": 10765
    },
    {
      "entropy": 5.3464209079742435,
      "epoch": 1.0345821325648414,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004901028646120459,
      "loss": 5.2144,
      "mean_token_accuracy": 0.19651904553174973,
      "num_tokens": 24694703.0,
      "step": 10770
    },
    {
      "entropy": 5.299839019775391,
      "epoch": 1.0350624399615753,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000490092805885222,
      "loss": 5.1222,
      "mean_token_accuracy": 0.20476236790418625,
      "num_tokens": 24706429.0,
      "step": 10775
    },
    {
      "entropy": 5.331740140914917,
      "epoch": 1.0355427473583094,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004900827421645816,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20401365756988527,
      "num_tokens": 24717682.0,
      "step": 10780
    },
    {
      "entropy": 5.410942125320434,
      "epoch": 1.0360230547550433,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004900726734503589,
      "loss": 5.219,
      "mean_token_accuracy": 0.19769208431243895,
      "num_tokens": 24728789.0,
      "step": 10785
    },
    {
      "entropy": 5.268787050247193,
      "epoch": 1.0365033621517772,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004900625997427872,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20358683466911315,
      "num_tokens": 24739788.0,
      "step": 10790
    },
    {
      "entropy": 5.314316129684448,
      "epoch": 1.036983669548511,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004900525210421006,
      "loss": 5.1883,
      "mean_token_accuracy": 0.19738685786724092,
      "num_tokens": 24751828.0,
      "step": 10795
    },
    {
      "entropy": 5.359867000579834,
      "epoch": 1.037463976945245,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004900424373485329,
      "loss": 5.1147,
      "mean_token_accuracy": 0.2034547060728073,
      "num_tokens": 24762061.0,
      "step": 10800
    },
    {
      "entropy": 5.286533308029175,
      "epoch": 1.0379442843419788,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004900323486623185,
      "loss": 5.1306,
      "mean_token_accuracy": 0.20575871765613557,
      "num_tokens": 24774018.0,
      "step": 10805
    },
    {
      "entropy": 5.303113317489624,
      "epoch": 1.0384245917387127,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004900222549836914,
      "loss": 5.101,
      "mean_token_accuracy": 0.20768230706453322,
      "num_tokens": 24785206.0,
      "step": 10810
    },
    {
      "entropy": 5.325172042846679,
      "epoch": 1.0389048991354466,
      "grad_norm": 1.6875,
      "learning_rate": 0.000490012156312886,
      "loss": 5.1724,
      "mean_token_accuracy": 0.20208380371332169,
      "num_tokens": 24797730.0,
      "step": 10815
    },
    {
      "entropy": 5.336231279373169,
      "epoch": 1.0393852065321807,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004900020526501369,
      "loss": 5.2617,
      "mean_token_accuracy": 0.19558039605617522,
      "num_tokens": 24809267.0,
      "step": 10820
    },
    {
      "entropy": 5.36703405380249,
      "epoch": 1.0398655139289146,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004899919439956785,
      "loss": 5.1401,
      "mean_token_accuracy": 0.19523145854473115,
      "num_tokens": 24820461.0,
      "step": 10825
    },
    {
      "entropy": 5.258511400222778,
      "epoch": 1.0403458213256485,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004899818303497455,
      "loss": 5.1526,
      "mean_token_accuracy": 0.1982979491353035,
      "num_tokens": 24831959.0,
      "step": 10830
    },
    {
      "entropy": 5.291241455078125,
      "epoch": 1.0408261287223823,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004899717117125728,
      "loss": 5.1292,
      "mean_token_accuracy": 0.20087361037731172,
      "num_tokens": 24843655.0,
      "step": 10835
    },
    {
      "entropy": 5.3269788265228275,
      "epoch": 1.0413064361191162,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004899615880843953,
      "loss": 5.107,
      "mean_token_accuracy": 0.19905167818069458,
      "num_tokens": 24855643.0,
      "step": 10840
    },
    {
      "entropy": 5.326853799819946,
      "epoch": 1.04178674351585,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004899514594654481,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20727464109659194,
      "num_tokens": 24865745.0,
      "step": 10845
    },
    {
      "entropy": 5.258389139175415,
      "epoch": 1.042267050912584,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004899413258559662,
      "loss": 5.0838,
      "mean_token_accuracy": 0.20829951465129853,
      "num_tokens": 24876567.0,
      "step": 10850
    },
    {
      "entropy": 5.25359320640564,
      "epoch": 1.0427473583093179,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004899311872561849,
      "loss": 5.0611,
      "mean_token_accuracy": 0.20130243450403212,
      "num_tokens": 24888294.0,
      "step": 10855
    },
    {
      "entropy": 5.358793592453003,
      "epoch": 1.043227665706052,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004899210436663398,
      "loss": 5.2583,
      "mean_token_accuracy": 0.19389114379882813,
      "num_tokens": 24899559.0,
      "step": 10860
    },
    {
      "entropy": 5.33980541229248,
      "epoch": 1.0437079731027858,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004899108950866661,
      "loss": 5.1268,
      "mean_token_accuracy": 0.2026054412126541,
      "num_tokens": 24911205.0,
      "step": 10865
    },
    {
      "entropy": 5.3029063701629635,
      "epoch": 1.0441882804995197,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004899007415173997,
      "loss": 5.0924,
      "mean_token_accuracy": 0.20214444547891616,
      "num_tokens": 24922635.0,
      "step": 10870
    },
    {
      "entropy": 5.262282419204712,
      "epoch": 1.0446685878962536,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004898905829587762,
      "loss": 5.0966,
      "mean_token_accuracy": 0.2085036426782608,
      "num_tokens": 24932941.0,
      "step": 10875
    },
    {
      "entropy": 5.378588724136352,
      "epoch": 1.0451488952929875,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004898804194110313,
      "loss": 5.2116,
      "mean_token_accuracy": 0.20182866007089614,
      "num_tokens": 24943980.0,
      "step": 10880
    },
    {
      "entropy": 5.380227136611938,
      "epoch": 1.0456292026897214,
      "grad_norm": 5.03125,
      "learning_rate": 0.0004898702508744012,
      "loss": 5.1787,
      "mean_token_accuracy": 0.19424990564584732,
      "num_tokens": 24955233.0,
      "step": 10885
    },
    {
      "entropy": 5.32433066368103,
      "epoch": 1.0461095100864553,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004898600773491221,
      "loss": 5.202,
      "mean_token_accuracy": 0.19754131138324738,
      "num_tokens": 24966971.0,
      "step": 10890
    },
    {
      "entropy": 5.388955116271973,
      "epoch": 1.0465898174831891,
      "grad_norm": 1.5,
      "learning_rate": 0.0004898498988354297,
      "loss": 5.1728,
      "mean_token_accuracy": 0.19758228659629823,
      "num_tokens": 24977515.0,
      "step": 10895
    },
    {
      "entropy": 5.365273952484131,
      "epoch": 1.0470701248799232,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004898397153335608,
      "loss": 5.1113,
      "mean_token_accuracy": 0.20484659522771836,
      "num_tokens": 24988131.0,
      "step": 10900
    },
    {
      "entropy": 5.382366371154785,
      "epoch": 1.0475504322766571,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004898295268437517,
      "loss": 5.1783,
      "mean_token_accuracy": 0.19517035633325577,
      "num_tokens": 24999238.0,
      "step": 10905
    },
    {
      "entropy": 5.3594996452331545,
      "epoch": 1.048030739673391,
      "grad_norm": 3.21875,
      "learning_rate": 0.0004898193333662388,
      "loss": 5.2189,
      "mean_token_accuracy": 0.19282138347625732,
      "num_tokens": 25011443.0,
      "step": 10910
    },
    {
      "entropy": 5.324726533889771,
      "epoch": 1.0485110470701249,
      "grad_norm": 1.625,
      "learning_rate": 0.0004898091349012588,
      "loss": 5.0608,
      "mean_token_accuracy": 0.2023757666349411,
      "num_tokens": 25022151.0,
      "step": 10915
    },
    {
      "entropy": 5.3478271484375,
      "epoch": 1.0489913544668588,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004897989314490486,
      "loss": 5.1504,
      "mean_token_accuracy": 0.20534102469682694,
      "num_tokens": 25032730.0,
      "step": 10920
    },
    {
      "entropy": 5.261315107345581,
      "epoch": 1.0494716618635926,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004897887230098451,
      "loss": 5.0971,
      "mean_token_accuracy": 0.20467284321784973,
      "num_tokens": 25044455.0,
      "step": 10925
    },
    {
      "entropy": 5.2687413692474365,
      "epoch": 1.0499519692603265,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004897785095838852,
      "loss": 5.0679,
      "mean_token_accuracy": 0.21040538251399993,
      "num_tokens": 25056647.0,
      "step": 10930
    },
    {
      "entropy": 5.285398197174072,
      "epoch": 1.0504322766570606,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004897682911714061,
      "loss": 5.0816,
      "mean_token_accuracy": 0.2036648541688919,
      "num_tokens": 25067618.0,
      "step": 10935
    },
    {
      "entropy": 5.316044807434082,
      "epoch": 1.0509125840537945,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000489758067772645,
      "loss": 5.0834,
      "mean_token_accuracy": 0.2035629704594612,
      "num_tokens": 25079753.0,
      "step": 10940
    },
    {
      "entropy": 5.283793640136719,
      "epoch": 1.0513928914505284,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004897478393878392,
      "loss": 5.1555,
      "mean_token_accuracy": 0.19937746673822404,
      "num_tokens": 25091023.0,
      "step": 10945
    },
    {
      "entropy": 5.331173467636108,
      "epoch": 1.0518731988472623,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004897376060172264,
      "loss": 5.1503,
      "mean_token_accuracy": 0.19686004221439363,
      "num_tokens": 25101396.0,
      "step": 10950
    },
    {
      "entropy": 5.344325304031372,
      "epoch": 1.0523535062439962,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004897273676610438,
      "loss": 5.1548,
      "mean_token_accuracy": 0.1974998876452446,
      "num_tokens": 25112562.0,
      "step": 10955
    },
    {
      "entropy": 5.333099269866944,
      "epoch": 1.05283381364073,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004897171243195295,
      "loss": 5.1934,
      "mean_token_accuracy": 0.20290710031986237,
      "num_tokens": 25124401.0,
      "step": 10960
    },
    {
      "entropy": 5.302322673797607,
      "epoch": 1.053314121037464,
      "grad_norm": 1.734375,
      "learning_rate": 0.000489706875992921,
      "loss": 5.1849,
      "mean_token_accuracy": 0.19728537499904633,
      "num_tokens": 25137345.0,
      "step": 10965
    },
    {
      "entropy": 5.349043846130371,
      "epoch": 1.0537944284341978,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004896966226814565,
      "loss": 5.1801,
      "mean_token_accuracy": 0.19820962399244307,
      "num_tokens": 25149300.0,
      "step": 10970
    },
    {
      "entropy": 5.3485993385314945,
      "epoch": 1.054274735830932,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004896863643853739,
      "loss": 5.0985,
      "mean_token_accuracy": 0.20550342947244643,
      "num_tokens": 25159510.0,
      "step": 10975
    },
    {
      "entropy": 5.36569390296936,
      "epoch": 1.0547550432276658,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004896761011049114,
      "loss": 5.1091,
      "mean_token_accuracy": 0.19641568064689635,
      "num_tokens": 25172346.0,
      "step": 10980
    },
    {
      "entropy": 5.395049571990967,
      "epoch": 1.0552353506243997,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004896658328403074,
      "loss": 5.2457,
      "mean_token_accuracy": 0.2000742509961128,
      "num_tokens": 25183471.0,
      "step": 10985
    },
    {
      "entropy": 5.3808046817779545,
      "epoch": 1.0557156580211335,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004896555595918001,
      "loss": 5.1629,
      "mean_token_accuracy": 0.20059962719678878,
      "num_tokens": 25195067.0,
      "step": 10990
    },
    {
      "entropy": 5.3311121463775635,
      "epoch": 1.0561959654178674,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004896452813596281,
      "loss": 5.1882,
      "mean_token_accuracy": 0.19561705291271209,
      "num_tokens": 25206165.0,
      "step": 10995
    },
    {
      "entropy": 5.315497589111328,
      "epoch": 1.0566762728146013,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004896349981440301,
      "loss": 5.1,
      "mean_token_accuracy": 0.20908810794353486,
      "num_tokens": 25217229.0,
      "step": 11000
    },
    {
      "entropy": 5.408291530609131,
      "epoch": 1.0571565802113352,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004896247099452447,
      "loss": 5.248,
      "mean_token_accuracy": 0.19362145364284516,
      "num_tokens": 25229349.0,
      "step": 11005
    },
    {
      "entropy": 5.351105976104736,
      "epoch": 1.057636887608069,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004896144167635108,
      "loss": 5.1194,
      "mean_token_accuracy": 0.20179051756858826,
      "num_tokens": 25240367.0,
      "step": 11010
    },
    {
      "entropy": 5.346260929107666,
      "epoch": 1.0581171950048032,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004896041185990675,
      "loss": 5.2133,
      "mean_token_accuracy": 0.1944615662097931,
      "num_tokens": 25252025.0,
      "step": 11015
    },
    {
      "entropy": 5.35924391746521,
      "epoch": 1.058597502401537,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004895938154521538,
      "loss": 5.1325,
      "mean_token_accuracy": 0.20582169741392137,
      "num_tokens": 25264846.0,
      "step": 11020
    },
    {
      "entropy": 5.307664060592652,
      "epoch": 1.059077809798271,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004895835073230089,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20170839875936508,
      "num_tokens": 25276150.0,
      "step": 11025
    },
    {
      "entropy": 5.334551429748535,
      "epoch": 1.0595581171950048,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004895731942118722,
      "loss": 5.1144,
      "mean_token_accuracy": 0.20394746214151382,
      "num_tokens": 25287819.0,
      "step": 11030
    },
    {
      "entropy": 5.361875438690186,
      "epoch": 1.0600384245917387,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004895628761189829,
      "loss": 5.1969,
      "mean_token_accuracy": 0.1970707193017006,
      "num_tokens": 25300467.0,
      "step": 11035
    },
    {
      "entropy": 5.241466903686524,
      "epoch": 1.0605187319884726,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004895525530445809,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20045005530118942,
      "num_tokens": 25312447.0,
      "step": 11040
    },
    {
      "entropy": 5.337131118774414,
      "epoch": 1.0609990393852065,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004895422249889057,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19588236808776854,
      "num_tokens": 25324160.0,
      "step": 11045
    },
    {
      "entropy": 5.342945718765259,
      "epoch": 1.0614793467819403,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004895318919521971,
      "loss": 5.0439,
      "mean_token_accuracy": 0.21145522445440293,
      "num_tokens": 25335106.0,
      "step": 11050
    },
    {
      "entropy": 5.2834875106811525,
      "epoch": 1.0619596541786744,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004895215539346949,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20613888949155806,
      "num_tokens": 25345246.0,
      "step": 11055
    },
    {
      "entropy": 5.288952732086182,
      "epoch": 1.0624399615754083,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004895112109366393,
      "loss": 5.1687,
      "mean_token_accuracy": 0.20038184523582458,
      "num_tokens": 25356665.0,
      "step": 11060
    },
    {
      "entropy": 5.320684862136841,
      "epoch": 1.0629202689721422,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004895008629582703,
      "loss": 5.1015,
      "mean_token_accuracy": 0.20399431735277176,
      "num_tokens": 25368128.0,
      "step": 11065
    },
    {
      "entropy": 5.256019353866577,
      "epoch": 1.063400576368876,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004894905099998283,
      "loss": 5.0985,
      "mean_token_accuracy": 0.20649359077215196,
      "num_tokens": 25379061.0,
      "step": 11070
    },
    {
      "entropy": 5.323419713973999,
      "epoch": 1.06388088376561,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004894801520615535,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20609591603279115,
      "num_tokens": 25390157.0,
      "step": 11075
    },
    {
      "entropy": 5.295299339294433,
      "epoch": 1.0643611911623438,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004894697891436863,
      "loss": 5.1187,
      "mean_token_accuracy": 0.2000119388103485,
      "num_tokens": 25400446.0,
      "step": 11080
    },
    {
      "entropy": 5.354647254943847,
      "epoch": 1.0648414985590777,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004894594212464676,
      "loss": 5.1459,
      "mean_token_accuracy": 0.19363622069358827,
      "num_tokens": 25412064.0,
      "step": 11085
    },
    {
      "entropy": 5.314082050323487,
      "epoch": 1.0653218059558118,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004894490483701381,
      "loss": 5.1484,
      "mean_token_accuracy": 0.20125600844621658,
      "num_tokens": 25422898.0,
      "step": 11090
    },
    {
      "entropy": 5.329068040847778,
      "epoch": 1.0658021133525457,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004894386705149382,
      "loss": 5.1283,
      "mean_token_accuracy": 0.19680934250354767,
      "num_tokens": 25434318.0,
      "step": 11095
    },
    {
      "entropy": 5.27101616859436,
      "epoch": 1.0662824207492796,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004894282876811093,
      "loss": 5.0741,
      "mean_token_accuracy": 0.2062190964818001,
      "num_tokens": 25444941.0,
      "step": 11100
    },
    {
      "entropy": 5.35099687576294,
      "epoch": 1.0667627281460135,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004894178998688921,
      "loss": 5.1034,
      "mean_token_accuracy": 0.2039244294166565,
      "num_tokens": 25455971.0,
      "step": 11105
    },
    {
      "entropy": 5.288018321990966,
      "epoch": 1.0672430355427474,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004894075070785281,
      "loss": 5.0985,
      "mean_token_accuracy": 0.19993453323841096,
      "num_tokens": 25467587.0,
      "step": 11110
    },
    {
      "entropy": 5.234090805053711,
      "epoch": 1.0677233429394812,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004893971093102585,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20735440403223038,
      "num_tokens": 25479137.0,
      "step": 11115
    },
    {
      "entropy": 5.4302772045135494,
      "epoch": 1.0682036503362151,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004893867065643245,
      "loss": 5.2502,
      "mean_token_accuracy": 0.19399198442697524,
      "num_tokens": 25491755.0,
      "step": 11120
    },
    {
      "entropy": 5.269463062286377,
      "epoch": 1.068683957732949,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004893762988409678,
      "loss": 5.0449,
      "mean_token_accuracy": 0.20891222655773162,
      "num_tokens": 25503581.0,
      "step": 11125
    },
    {
      "entropy": 5.301801300048828,
      "epoch": 1.069164265129683,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004893658861404301,
      "loss": 5.1632,
      "mean_token_accuracy": 0.2034335657954216,
      "num_tokens": 25515032.0,
      "step": 11130
    },
    {
      "entropy": 5.270562839508057,
      "epoch": 1.069644572526417,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004893554684629529,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20247771888971328,
      "num_tokens": 25526347.0,
      "step": 11135
    },
    {
      "entropy": 5.378921699523926,
      "epoch": 1.0701248799231509,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004893450458087784,
      "loss": 5.165,
      "mean_token_accuracy": 0.2016587108373642,
      "num_tokens": 25538287.0,
      "step": 11140
    },
    {
      "entropy": 5.190388250350952,
      "epoch": 1.0706051873198847,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004893346181781483,
      "loss": 5.0873,
      "mean_token_accuracy": 0.21713515669107436,
      "num_tokens": 25548431.0,
      "step": 11145
    },
    {
      "entropy": 5.284275913238526,
      "epoch": 1.0710854947166186,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004893241855713048,
      "loss": 5.2511,
      "mean_token_accuracy": 0.19596656560897827,
      "num_tokens": 25558706.0,
      "step": 11150
    },
    {
      "entropy": 5.369014835357666,
      "epoch": 1.0715658021133525,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004893137479884903,
      "loss": 5.1334,
      "mean_token_accuracy": 0.19954039752483368,
      "num_tokens": 25569920.0,
      "step": 11155
    },
    {
      "entropy": 5.312915992736817,
      "epoch": 1.0720461095100864,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004893033054299468,
      "loss": 5.0683,
      "mean_token_accuracy": 0.21186500042676926,
      "num_tokens": 25581340.0,
      "step": 11160
    },
    {
      "entropy": 5.303180742263794,
      "epoch": 1.0725264169068203,
      "grad_norm": 2.078125,
      "learning_rate": 0.000489292857895917,
      "loss": 5.1639,
      "mean_token_accuracy": 0.19763920754194259,
      "num_tokens": 25592477.0,
      "step": 11165
    },
    {
      "entropy": 5.361607789993286,
      "epoch": 1.0730067243035544,
      "grad_norm": 1.5,
      "learning_rate": 0.0004892824053866432,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20616717934608458,
      "num_tokens": 25604583.0,
      "step": 11170
    },
    {
      "entropy": 5.438195610046387,
      "epoch": 1.0734870317002883,
      "grad_norm": 2.0,
      "learning_rate": 0.0004892719479023683,
      "loss": 5.2276,
      "mean_token_accuracy": 0.19459341019392012,
      "num_tokens": 25614743.0,
      "step": 11175
    },
    {
      "entropy": 5.242146968841553,
      "epoch": 1.0739673390970221,
      "grad_norm": 1.71875,
      "learning_rate": 0.000489261485443335,
      "loss": 4.9834,
      "mean_token_accuracy": 0.21015122681856155,
      "num_tokens": 25626796.0,
      "step": 11180
    },
    {
      "entropy": 5.30523681640625,
      "epoch": 1.074447646493756,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004892510180097863,
      "loss": 5.1134,
      "mean_token_accuracy": 0.20219666808843612,
      "num_tokens": 25638014.0,
      "step": 11185
    },
    {
      "entropy": 5.284056901931763,
      "epoch": 1.07492795389049,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004892405456019651,
      "loss": 5.1581,
      "mean_token_accuracy": 0.20499511957168579,
      "num_tokens": 25649555.0,
      "step": 11190
    },
    {
      "entropy": 5.376839256286621,
      "epoch": 1.0754082612872238,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004892300682201147,
      "loss": 5.2426,
      "mean_token_accuracy": 0.19616783708333968,
      "num_tokens": 25662886.0,
      "step": 11195
    },
    {
      "entropy": 5.367648506164551,
      "epoch": 1.0758885686839577,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004892195858644782,
      "loss": 5.0948,
      "mean_token_accuracy": 0.20881618708372116,
      "num_tokens": 25673802.0,
      "step": 11200
    },
    {
      "entropy": 5.28165488243103,
      "epoch": 1.0763688760806915,
      "grad_norm": 1.7890625,
      "learning_rate": 0.000489209098535299,
      "loss": 5.1293,
      "mean_token_accuracy": 0.20603721886873244,
      "num_tokens": 25685545.0,
      "step": 11205
    },
    {
      "entropy": 5.269009351730347,
      "epoch": 1.0768491834774256,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004891986062328205,
      "loss": 5.072,
      "mean_token_accuracy": 0.2044074445962906,
      "num_tokens": 25697955.0,
      "step": 11210
    },
    {
      "entropy": 5.363277864456177,
      "epoch": 1.0773294908741595,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004891881089572865,
      "loss": 5.1483,
      "mean_token_accuracy": 0.19669867306947708,
      "num_tokens": 25709707.0,
      "step": 11215
    },
    {
      "entropy": 5.349743318557739,
      "epoch": 1.0778097982708934,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004891776067089406,
      "loss": 5.1929,
      "mean_token_accuracy": 0.19733243435621262,
      "num_tokens": 25720576.0,
      "step": 11220
    },
    {
      "entropy": 5.282660341262817,
      "epoch": 1.0782901056676273,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004891670994880266,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20656931549310684,
      "num_tokens": 25731387.0,
      "step": 11225
    },
    {
      "entropy": 5.293915271759033,
      "epoch": 1.0787704130643612,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004891565872947888,
      "loss": 5.048,
      "mean_token_accuracy": 0.2143150970339775,
      "num_tokens": 25741813.0,
      "step": 11230
    },
    {
      "entropy": 5.3240571975708,
      "epoch": 1.079250720461095,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004891460701294706,
      "loss": 5.1486,
      "mean_token_accuracy": 0.1984674945473671,
      "num_tokens": 25753763.0,
      "step": 11235
    },
    {
      "entropy": 5.302602529525757,
      "epoch": 1.079731027857829,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004891355479923167,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20513346344232558,
      "num_tokens": 25765269.0,
      "step": 11240
    },
    {
      "entropy": 5.304611253738403,
      "epoch": 1.080211335254563,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004891250208835712,
      "loss": 5.1672,
      "mean_token_accuracy": 0.2030718818306923,
      "num_tokens": 25776841.0,
      "step": 11245
    },
    {
      "entropy": 5.368564414978027,
      "epoch": 1.080691642651297,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004891144888034784,
      "loss": 5.2038,
      "mean_token_accuracy": 0.18602197021245956,
      "num_tokens": 25788151.0,
      "step": 11250
    },
    {
      "entropy": 5.337092876434326,
      "epoch": 1.0811719500480308,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004891039517522832,
      "loss": 5.0752,
      "mean_token_accuracy": 0.2106889858841896,
      "num_tokens": 25799194.0,
      "step": 11255
    },
    {
      "entropy": 5.287901401519775,
      "epoch": 1.0816522574447647,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004890934097302299,
      "loss": 5.111,
      "mean_token_accuracy": 0.19977914839982985,
      "num_tokens": 25809483.0,
      "step": 11260
    },
    {
      "entropy": 5.252952909469604,
      "epoch": 1.0821325648414986,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004890828627375632,
      "loss": 5.0917,
      "mean_token_accuracy": 0.19584571123123168,
      "num_tokens": 25820781.0,
      "step": 11265
    },
    {
      "entropy": 5.362701177597046,
      "epoch": 1.0826128722382324,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004890723107745283,
      "loss": 5.1824,
      "mean_token_accuracy": 0.1962701588869095,
      "num_tokens": 25832111.0,
      "step": 11270
    },
    {
      "entropy": 5.428535985946655,
      "epoch": 1.0830931796349663,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004890617538413699,
      "loss": 5.1861,
      "mean_token_accuracy": 0.20441062301397322,
      "num_tokens": 25843300.0,
      "step": 11275
    },
    {
      "entropy": 5.380526542663574,
      "epoch": 1.0835734870317002,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004890511919383333,
      "loss": 5.1322,
      "mean_token_accuracy": 0.19422634094953536,
      "num_tokens": 25855337.0,
      "step": 11280
    },
    {
      "entropy": 5.4125570297241214,
      "epoch": 1.084053794428434,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004890406250656636,
      "loss": 5.2178,
      "mean_token_accuracy": 0.19568423926830292,
      "num_tokens": 25866683.0,
      "step": 11285
    },
    {
      "entropy": 5.261790180206299,
      "epoch": 1.0845341018251682,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004890300532236062,
      "loss": 5.1019,
      "mean_token_accuracy": 0.20246828198432923,
      "num_tokens": 25878041.0,
      "step": 11290
    },
    {
      "entropy": 5.398085498809815,
      "epoch": 1.085014409221902,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004890194764124064,
      "loss": 5.1364,
      "mean_token_accuracy": 0.19958120584487915,
      "num_tokens": 25889059.0,
      "step": 11295
    },
    {
      "entropy": 5.281335115432739,
      "epoch": 1.085494716618636,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004890088946323099,
      "loss": 5.03,
      "mean_token_accuracy": 0.21608632951974868,
      "num_tokens": 25900062.0,
      "step": 11300
    },
    {
      "entropy": 5.316459131240845,
      "epoch": 1.0859750240153698,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004889983078835623,
      "loss": 5.2666,
      "mean_token_accuracy": 0.19546637386083604,
      "num_tokens": 25911813.0,
      "step": 11305
    },
    {
      "entropy": 5.36879415512085,
      "epoch": 1.0864553314121037,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004889877161664096,
      "loss": 5.0543,
      "mean_token_accuracy": 0.20611478239297867,
      "num_tokens": 25922321.0,
      "step": 11310
    },
    {
      "entropy": 5.315264940261841,
      "epoch": 1.0869356388088376,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004889771194810974,
      "loss": 5.1647,
      "mean_token_accuracy": 0.20887239575386046,
      "num_tokens": 25933631.0,
      "step": 11315
    },
    {
      "entropy": 5.375712156295776,
      "epoch": 1.0874159462055715,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004889665178278719,
      "loss": 5.1903,
      "mean_token_accuracy": 0.20074028968811036,
      "num_tokens": 25945190.0,
      "step": 11320
    },
    {
      "entropy": 5.365614700317383,
      "epoch": 1.0878962536023056,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004889559112069792,
      "loss": 5.1266,
      "mean_token_accuracy": 0.20831136107444764,
      "num_tokens": 25956817.0,
      "step": 11325
    },
    {
      "entropy": 5.401874113082886,
      "epoch": 1.0883765609990395,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004889452996186657,
      "loss": 5.1759,
      "mean_token_accuracy": 0.2048810377717018,
      "num_tokens": 25968134.0,
      "step": 11330
    },
    {
      "entropy": 5.25398621559143,
      "epoch": 1.0888568683957733,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004889346830631774,
      "loss": 5.1142,
      "mean_token_accuracy": 0.20354647785425187,
      "num_tokens": 25981825.0,
      "step": 11335
    },
    {
      "entropy": 5.364640998840332,
      "epoch": 1.0893371757925072,
      "grad_norm": 1.5625,
      "learning_rate": 0.000488924061540761,
      "loss": 5.2423,
      "mean_token_accuracy": 0.18749297857284547,
      "num_tokens": 25993962.0,
      "step": 11340
    },
    {
      "entropy": 5.307242155075073,
      "epoch": 1.089817483189241,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004889134350516633,
      "loss": 5.0934,
      "mean_token_accuracy": 0.20552517622709274,
      "num_tokens": 26005705.0,
      "step": 11345
    },
    {
      "entropy": 5.316702270507813,
      "epoch": 1.090297790585975,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004889028035961308,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20994055718183519,
      "num_tokens": 26017111.0,
      "step": 11350
    },
    {
      "entropy": 5.319946193695069,
      "epoch": 1.0907780979827089,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004888921671744103,
      "loss": 5.1722,
      "mean_token_accuracy": 0.19824621826410294,
      "num_tokens": 26028097.0,
      "step": 11355
    },
    {
      "entropy": 5.317778205871582,
      "epoch": 1.0912584053794427,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004888815257867488,
      "loss": 5.1053,
      "mean_token_accuracy": 0.2026633307337761,
      "num_tokens": 26039281.0,
      "step": 11360
    },
    {
      "entropy": 5.342022180557251,
      "epoch": 1.0917387127761768,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004888708794333934,
      "loss": 5.1874,
      "mean_token_accuracy": 0.19024603068828583,
      "num_tokens": 26051133.0,
      "step": 11365
    },
    {
      "entropy": 5.314816284179687,
      "epoch": 1.0922190201729107,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004888602281145913,
      "loss": 5.1648,
      "mean_token_accuracy": 0.1988406077027321,
      "num_tokens": 26062220.0,
      "step": 11370
    },
    {
      "entropy": 5.314972734451294,
      "epoch": 1.0926993275696446,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004888495718305897,
      "loss": 5.0783,
      "mean_token_accuracy": 0.21319786161184312,
      "num_tokens": 26072556.0,
      "step": 11375
    },
    {
      "entropy": 5.380842208862305,
      "epoch": 1.0931796349663785,
      "grad_norm": 1.609375,
      "learning_rate": 0.000488838910581636,
      "loss": 5.2178,
      "mean_token_accuracy": 0.19425466656684875,
      "num_tokens": 26084218.0,
      "step": 11380
    },
    {
      "entropy": 5.4592187881469725,
      "epoch": 1.0936599423631124,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004888282443679777,
      "loss": 5.3343,
      "mean_token_accuracy": 0.19081743508577348,
      "num_tokens": 26096272.0,
      "step": 11385
    },
    {
      "entropy": 5.376057863235474,
      "epoch": 1.0941402497598463,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004888175731898627,
      "loss": 5.2093,
      "mean_token_accuracy": 0.1963579311966896,
      "num_tokens": 26109247.0,
      "step": 11390
    },
    {
      "entropy": 5.3417774677276615,
      "epoch": 1.0946205571565801,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004888068970475384,
      "loss": 5.1102,
      "mean_token_accuracy": 0.20040577352046968,
      "num_tokens": 26121340.0,
      "step": 11395
    },
    {
      "entropy": 5.357487201690674,
      "epoch": 1.0951008645533142,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004887962159412529,
      "loss": 5.1395,
      "mean_token_accuracy": 0.19647157937288284,
      "num_tokens": 26133135.0,
      "step": 11400
    },
    {
      "entropy": 5.256601476669312,
      "epoch": 1.0955811719500481,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004887855298712541,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20414958596229554,
      "num_tokens": 26145324.0,
      "step": 11405
    },
    {
      "entropy": 5.238174772262573,
      "epoch": 1.096061479346782,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048877483883779,
      "loss": 5.0418,
      "mean_token_accuracy": 0.20625945925712585,
      "num_tokens": 26156739.0,
      "step": 11410
    },
    {
      "entropy": 5.3349381446838375,
      "epoch": 1.0965417867435159,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004887641428411091,
      "loss": 5.1561,
      "mean_token_accuracy": 0.1959335044026375,
      "num_tokens": 26169543.0,
      "step": 11415
    },
    {
      "entropy": 5.362424612045288,
      "epoch": 1.0970220941402498,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004887534418814595,
      "loss": 5.1478,
      "mean_token_accuracy": 0.20367608070373536,
      "num_tokens": 26181317.0,
      "step": 11420
    },
    {
      "entropy": 5.394422101974487,
      "epoch": 1.0975024015369836,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004887427359590897,
      "loss": 5.227,
      "mean_token_accuracy": 0.19395215213298797,
      "num_tokens": 26193402.0,
      "step": 11425
    },
    {
      "entropy": 5.307780361175537,
      "epoch": 1.0979827089337175,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004887320250742482,
      "loss": 5.0916,
      "mean_token_accuracy": 0.20511206686496736,
      "num_tokens": 26204660.0,
      "step": 11430
    },
    {
      "entropy": 5.383396196365356,
      "epoch": 1.0984630163304514,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004887213092271838,
      "loss": 5.1528,
      "mean_token_accuracy": 0.1984705463051796,
      "num_tokens": 26215593.0,
      "step": 11435
    },
    {
      "entropy": 5.260968637466431,
      "epoch": 1.0989433237271853,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004887105884181451,
      "loss": 5.053,
      "mean_token_accuracy": 0.21085789948701858,
      "num_tokens": 26227481.0,
      "step": 11440
    },
    {
      "entropy": 5.341152286529541,
      "epoch": 1.0994236311239194,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004886998626473813,
      "loss": 5.2255,
      "mean_token_accuracy": 0.19789917320013045,
      "num_tokens": 26238260.0,
      "step": 11445
    },
    {
      "entropy": 5.365911197662354,
      "epoch": 1.0999039385206533,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004886891319151411,
      "loss": 5.1328,
      "mean_token_accuracy": 0.2019819512963295,
      "num_tokens": 26249749.0,
      "step": 11450
    },
    {
      "entropy": 5.36580228805542,
      "epoch": 1.1003842459173871,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004886783962216738,
      "loss": 5.1212,
      "mean_token_accuracy": 0.2028710126876831,
      "num_tokens": 26260639.0,
      "step": 11455
    },
    {
      "entropy": 5.286028146743774,
      "epoch": 1.100864553314121,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004886676555672287,
      "loss": 5.145,
      "mean_token_accuracy": 0.1980138450860977,
      "num_tokens": 26271797.0,
      "step": 11460
    },
    {
      "entropy": 5.326308965682983,
      "epoch": 1.101344860710855,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004886569099520551,
      "loss": 5.1303,
      "mean_token_accuracy": 0.19022870510816575,
      "num_tokens": 26283298.0,
      "step": 11465
    },
    {
      "entropy": 5.240048599243164,
      "epoch": 1.1018251681075888,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004886461593764024,
      "loss": 4.9702,
      "mean_token_accuracy": 0.20655218809843062,
      "num_tokens": 26293632.0,
      "step": 11470
    },
    {
      "entropy": 5.30937933921814,
      "epoch": 1.1023054755043227,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004886354038405204,
      "loss": 5.2134,
      "mean_token_accuracy": 0.1942225769162178,
      "num_tokens": 26307053.0,
      "step": 11475
    },
    {
      "entropy": 5.308082342147827,
      "epoch": 1.1027857829010568,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004886246433446586,
      "loss": 5.0461,
      "mean_token_accuracy": 0.21324211806058885,
      "num_tokens": 26317136.0,
      "step": 11480
    },
    {
      "entropy": 5.310516214370727,
      "epoch": 1.1032660902977907,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004886138778890669,
      "loss": 5.1588,
      "mean_token_accuracy": 0.20462871193885804,
      "num_tokens": 26328865.0,
      "step": 11485
    },
    {
      "entropy": 5.286093759536743,
      "epoch": 1.1037463976945245,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004886031074739953,
      "loss": 5.1544,
      "mean_token_accuracy": 0.20698221772909164,
      "num_tokens": 26341010.0,
      "step": 11490
    },
    {
      "entropy": 5.3811126232147215,
      "epoch": 1.1042267050912584,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004885923320996938,
      "loss": 5.1995,
      "mean_token_accuracy": 0.1976682350039482,
      "num_tokens": 26352203.0,
      "step": 11495
    },
    {
      "entropy": 5.369465780258179,
      "epoch": 1.1047070124879923,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004885815517664127,
      "loss": 5.1608,
      "mean_token_accuracy": 0.1991453006863594,
      "num_tokens": 26363096.0,
      "step": 11500
    },
    {
      "entropy": 5.235171985626221,
      "epoch": 1.1051873198847262,
      "grad_norm": 2.84375,
      "learning_rate": 0.000488570766474402,
      "loss": 5.0935,
      "mean_token_accuracy": 0.20452217906713485,
      "num_tokens": 26374776.0,
      "step": 11505
    },
    {
      "entropy": 5.288644599914551,
      "epoch": 1.10566762728146,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004885599762239124,
      "loss": 5.1729,
      "mean_token_accuracy": 0.2004079893231392,
      "num_tokens": 26386654.0,
      "step": 11510
    },
    {
      "entropy": 5.281540012359619,
      "epoch": 1.106147934678194,
      "grad_norm": 3.0625,
      "learning_rate": 0.0004885491810151943,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20224497318267823,
      "num_tokens": 26400022.0,
      "step": 11515
    },
    {
      "entropy": 5.292015504837036,
      "epoch": 1.106628242074928,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004885383808484982,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20151816606521605,
      "num_tokens": 26412308.0,
      "step": 11520
    },
    {
      "entropy": 5.387752437591553,
      "epoch": 1.107108549471662,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004885275757240751,
      "loss": 5.2164,
      "mean_token_accuracy": 0.1877727374434471,
      "num_tokens": 26424506.0,
      "step": 11525
    },
    {
      "entropy": 5.326724243164063,
      "epoch": 1.1075888568683958,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004885167656421757,
      "loss": 5.1161,
      "mean_token_accuracy": 0.20713104903697968,
      "num_tokens": 26435020.0,
      "step": 11530
    },
    {
      "entropy": 5.194391345977783,
      "epoch": 1.1080691642651297,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000488505950603051,
      "loss": 5.0489,
      "mean_token_accuracy": 0.20736747235059738,
      "num_tokens": 26447952.0,
      "step": 11535
    },
    {
      "entropy": 5.345295095443726,
      "epoch": 1.1085494716618636,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000488495130606952,
      "loss": 5.1113,
      "mean_token_accuracy": 0.20188061892986298,
      "num_tokens": 26460001.0,
      "step": 11540
    },
    {
      "entropy": 5.325395345687866,
      "epoch": 1.1090297790585975,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004884843056541302,
      "loss": 5.099,
      "mean_token_accuracy": 0.2056600719690323,
      "num_tokens": 26472139.0,
      "step": 11545
    },
    {
      "entropy": 5.324504566192627,
      "epoch": 1.1095100864553313,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004884734757448367,
      "loss": 5.1583,
      "mean_token_accuracy": 0.20389454215765,
      "num_tokens": 26483855.0,
      "step": 11550
    },
    {
      "entropy": 5.380694341659546,
      "epoch": 1.1099903938520654,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000488462640879323,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1924385756254196,
      "num_tokens": 26496529.0,
      "step": 11555
    },
    {
      "entropy": 5.328584671020508,
      "epoch": 1.1104707012487993,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004884518010578405,
      "loss": 5.079,
      "mean_token_accuracy": 0.21022463589906693,
      "num_tokens": 26506663.0,
      "step": 11560
    },
    {
      "entropy": 5.344662618637085,
      "epoch": 1.1109510086455332,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004884409562806411,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19380222856998444,
      "num_tokens": 26519091.0,
      "step": 11565
    },
    {
      "entropy": 5.3521080493927,
      "epoch": 1.111431316042267,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004884301065479765,
      "loss": 5.161,
      "mean_token_accuracy": 0.19393169283866882,
      "num_tokens": 26530145.0,
      "step": 11570
    },
    {
      "entropy": 5.370323705673218,
      "epoch": 1.111911623439001,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004884192518600986,
      "loss": 5.0848,
      "mean_token_accuracy": 0.19910329729318618,
      "num_tokens": 26541228.0,
      "step": 11575
    },
    {
      "entropy": 5.36710524559021,
      "epoch": 1.1123919308357348,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004884083922172593,
      "loss": 5.136,
      "mean_token_accuracy": 0.19784497767686843,
      "num_tokens": 26551359.0,
      "step": 11580
    },
    {
      "entropy": 5.198082780838012,
      "epoch": 1.1128722382324687,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004883975276197108,
      "loss": 5.1308,
      "mean_token_accuracy": 0.2030889168381691,
      "num_tokens": 26562911.0,
      "step": 11585
    },
    {
      "entropy": 5.3775718212127686,
      "epoch": 1.1133525456292026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004883866580677055,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20831287354230882,
      "num_tokens": 26573198.0,
      "step": 11590
    },
    {
      "entropy": 5.297350549697876,
      "epoch": 1.1138328530259365,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004883757835614956,
      "loss": 5.1507,
      "mean_token_accuracy": 0.1993216633796692,
      "num_tokens": 26584932.0,
      "step": 11595
    },
    {
      "entropy": 5.391242742538452,
      "epoch": 1.1143131604226706,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004883649041013335,
      "loss": 5.2766,
      "mean_token_accuracy": 0.19492211639881135,
      "num_tokens": 26596042.0,
      "step": 11600
    },
    {
      "entropy": 5.391051769256592,
      "epoch": 1.1147934678194045,
      "grad_norm": 1.75,
      "learning_rate": 0.000488354019687472,
      "loss": 5.1624,
      "mean_token_accuracy": 0.1957885205745697,
      "num_tokens": 26607359.0,
      "step": 11605
    },
    {
      "entropy": 5.418501377105713,
      "epoch": 1.1152737752161384,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004883431303201636,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20717217326164244,
      "num_tokens": 26617095.0,
      "step": 11610
    },
    {
      "entropy": 5.34371018409729,
      "epoch": 1.1157540826128722,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004883322359996613,
      "loss": 5.1775,
      "mean_token_accuracy": 0.20002456605434418,
      "num_tokens": 26628450.0,
      "step": 11615
    },
    {
      "entropy": 5.275143957138061,
      "epoch": 1.1162343900096061,
      "grad_norm": 2.828125,
      "learning_rate": 0.0004883213367262179,
      "loss": 5.0736,
      "mean_token_accuracy": 0.20413772612810135,
      "num_tokens": 26639428.0,
      "step": 11620
    },
    {
      "entropy": 5.212841367721557,
      "epoch": 1.11671469740634,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004883104325000866,
      "loss": 5.1012,
      "mean_token_accuracy": 0.20164805948734282,
      "num_tokens": 26650730.0,
      "step": 11625
    },
    {
      "entropy": 5.359587049484253,
      "epoch": 1.1171950048030739,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004882995233215203,
      "loss": 5.0802,
      "mean_token_accuracy": 0.19986220002174376,
      "num_tokens": 26661745.0,
      "step": 11630
    },
    {
      "entropy": 5.304432821273804,
      "epoch": 1.117675312199808,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004882886091907726,
      "loss": 5.0299,
      "mean_token_accuracy": 0.21270100474357606,
      "num_tokens": 26672976.0,
      "step": 11635
    },
    {
      "entropy": 5.3710681915283205,
      "epoch": 1.1181556195965419,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048827769010809666,
      "loss": 5.1715,
      "mean_token_accuracy": 0.20010971128940583,
      "num_tokens": 26683503.0,
      "step": 11640
    },
    {
      "entropy": 5.333467483520508,
      "epoch": 1.1186359269932757,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00048826676607374606,
      "loss": 5.1258,
      "mean_token_accuracy": 0.19441845864057541,
      "num_tokens": 26695160.0,
      "step": 11645
    },
    {
      "entropy": 5.298226356506348,
      "epoch": 1.1191162343900096,
      "grad_norm": 2.59375,
      "learning_rate": 0.00048825583708797434,
      "loss": 5.1234,
      "mean_token_accuracy": 0.2089619129896164,
      "num_tokens": 26706514.0,
      "step": 11650
    },
    {
      "entropy": 5.269511890411377,
      "epoch": 1.1195965417867435,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004882449031510354,
      "loss": 5.0291,
      "mean_token_accuracy": 0.21192042529582977,
      "num_tokens": 26717268.0,
      "step": 11655
    },
    {
      "entropy": 5.412432432174683,
      "epoch": 1.1200768491834774,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000488233964263183,
      "loss": 5.2831,
      "mean_token_accuracy": 0.1944217398762703,
      "num_tokens": 26728821.0,
      "step": 11660
    },
    {
      "entropy": 5.372755098342895,
      "epoch": 1.1205571565802113,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048822302042467115,
      "loss": 5.2344,
      "mean_token_accuracy": 0.1952611729502678,
      "num_tokens": 26740476.0,
      "step": 11665
    },
    {
      "entropy": 5.358553028106689,
      "epoch": 1.1210374639769451,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004882120716357539,
      "loss": 5.2108,
      "mean_token_accuracy": 0.2008669763803482,
      "num_tokens": 26751289.0,
      "step": 11670
    },
    {
      "entropy": 5.260187387466431,
      "epoch": 1.1215177713736793,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004882011178966854,
      "loss": 5.1211,
      "mean_token_accuracy": 0.20536865890026093,
      "num_tokens": 26762115.0,
      "step": 11675
    },
    {
      "entropy": 5.263187265396118,
      "epoch": 1.1219980787704131,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004881901592077201,
      "loss": 5.0622,
      "mean_token_accuracy": 0.2028586134314537,
      "num_tokens": 26772837.0,
      "step": 11680
    },
    {
      "entropy": 5.413993549346924,
      "epoch": 1.122478386167147,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004881791955691123,
      "loss": 5.1597,
      "mean_token_accuracy": 0.20203806608915328,
      "num_tokens": 26783721.0,
      "step": 11685
    },
    {
      "entropy": 5.288437223434448,
      "epoch": 1.122958693563881,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048816822698111655,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20836410373449327,
      "num_tokens": 26794320.0,
      "step": 11690
    },
    {
      "entropy": 5.328686952590942,
      "epoch": 1.1234390009606148,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048815725344398766,
      "loss": 5.2619,
      "mean_token_accuracy": 0.18543754518032074,
      "num_tokens": 26807053.0,
      "step": 11695
    },
    {
      "entropy": 5.432075452804566,
      "epoch": 1.1239193083573487,
      "grad_norm": 2.609375,
      "learning_rate": 0.00048814627495798017,
      "loss": 5.162,
      "mean_token_accuracy": 0.19605380892753602,
      "num_tokens": 26817245.0,
      "step": 11700
    },
    {
      "entropy": 5.318490743637085,
      "epoch": 1.1243996157540825,
      "grad_norm": 2.125,
      "learning_rate": 0.0004881352915233492,
      "loss": 5.2033,
      "mean_token_accuracy": 0.19369176179170608,
      "num_tokens": 26829498.0,
      "step": 11705
    },
    {
      "entropy": 5.269003677368164,
      "epoch": 1.1248799231508164,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048812430314034956,
      "loss": 5.0602,
      "mean_token_accuracy": 0.20862208157777787,
      "num_tokens": 26840534.0,
      "step": 11710
    },
    {
      "entropy": 5.345874357223511,
      "epoch": 1.1253602305475505,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004881133098092365,
      "loss": 5.1323,
      "mean_token_accuracy": 0.19713760316371917,
      "num_tokens": 26851305.0,
      "step": 11715
    },
    {
      "entropy": 5.286346864700318,
      "epoch": 1.1258405379442844,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004881023115302652,
      "loss": 5.1341,
      "mean_token_accuracy": 0.19811634868383407,
      "num_tokens": 26862620.0,
      "step": 11720
    },
    {
      "entropy": 5.341561412811279,
      "epoch": 1.1263208453410183,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000488091308303691,
      "loss": 5.1564,
      "mean_token_accuracy": 0.19700847566127777,
      "num_tokens": 26874108.0,
      "step": 11725
    },
    {
      "entropy": 5.379293823242188,
      "epoch": 1.1268011527377522,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004880803001297694,
      "loss": 5.1602,
      "mean_token_accuracy": 0.19801948666572572,
      "num_tokens": 26886523.0,
      "step": 11730
    },
    {
      "entropy": 5.268254137039184,
      "epoch": 1.127281460134486,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004880692870087559,
      "loss": 5.0595,
      "mean_token_accuracy": 0.20859607309103012,
      "num_tokens": 26897316.0,
      "step": 11735
    },
    {
      "entropy": 5.292900800704956,
      "epoch": 1.12776176753122,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048805826894090626,
      "loss": 5.111,
      "mean_token_accuracy": 0.1993417963385582,
      "num_tokens": 26909807.0,
      "step": 11740
    },
    {
      "entropy": 5.404553413391113,
      "epoch": 1.1282420749279538,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00048804724592647626,
      "loss": 5.2309,
      "mean_token_accuracy": 0.1930053010582924,
      "num_tokens": 26922054.0,
      "step": 11745
    },
    {
      "entropy": 5.2868565082550045,
      "epoch": 1.1287223823246877,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004880362179657218,
      "loss": 5.0878,
      "mean_token_accuracy": 0.19807282090187073,
      "num_tokens": 26933838.0,
      "step": 11750
    },
    {
      "entropy": 5.286669588088989,
      "epoch": 1.1292026897214218,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048802518505889904,
      "loss": 5.1608,
      "mean_token_accuracy": 0.19519306868314742,
      "num_tokens": 26945396.0,
      "step": 11755
    },
    {
      "entropy": 5.4107646465301515,
      "epoch": 1.1296829971181557,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048801414720626404,
      "loss": 5.2302,
      "mean_token_accuracy": 0.19581684917211534,
      "num_tokens": 26957809.0,
      "step": 11760
    },
    {
      "entropy": 5.381056547164917,
      "epoch": 1.1301633045148896,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048800310440807294,
      "loss": 5.0706,
      "mean_token_accuracy": 0.213106144964695,
      "num_tokens": 26969024.0,
      "step": 11765
    },
    {
      "entropy": 5.2403723239898685,
      "epoch": 1.1306436119116234,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004879920566645823,
      "loss": 5.1348,
      "mean_token_accuracy": 0.20311659723520278,
      "num_tokens": 26980689.0,
      "step": 11770
    },
    {
      "entropy": 5.381258153915406,
      "epoch": 1.1311239193083573,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004879810039760486,
      "loss": 5.2172,
      "mean_token_accuracy": 0.19316246062517167,
      "num_tokens": 26992115.0,
      "step": 11775
    },
    {
      "entropy": 5.416398000717163,
      "epoch": 1.1316042267050912,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004879699463427284,
      "loss": 5.1464,
      "mean_token_accuracy": 0.2001924678683281,
      "num_tokens": 27003042.0,
      "step": 11780
    },
    {
      "entropy": 5.269796848297119,
      "epoch": 1.132084534101825,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004879588837648785,
      "loss": 5.0768,
      "mean_token_accuracy": 0.20741364657878875,
      "num_tokens": 27013863.0,
      "step": 11785
    },
    {
      "entropy": 5.209153270721435,
      "epoch": 1.1325648414985592,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00048794781624275554,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20994998663663864,
      "num_tokens": 27024423.0,
      "step": 11790
    },
    {
      "entropy": 5.273507118225098,
      "epoch": 1.133045148895293,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048793674377661664,
      "loss": 5.1307,
      "mean_token_accuracy": 0.19897654950618743,
      "num_tokens": 27035460.0,
      "step": 11795
    },
    {
      "entropy": 5.308765459060669,
      "epoch": 1.133525456292027,
      "grad_norm": 4.9375,
      "learning_rate": 0.00048792566636671886,
      "loss": 5.1336,
      "mean_token_accuracy": 0.2021664023399353,
      "num_tokens": 27046731.0,
      "step": 11800
    },
    {
      "entropy": 5.317369031906128,
      "epoch": 1.1340057636887608,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004879145840133194,
      "loss": 5.1276,
      "mean_token_accuracy": 0.20544746965169908,
      "num_tokens": 27057758.0,
      "step": 11805
    },
    {
      "entropy": 5.419113540649414,
      "epoch": 1.1344860710854947,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004879034967166755,
      "loss": 5.1843,
      "mean_token_accuracy": 0.19622780233621598,
      "num_tokens": 27068055.0,
      "step": 11810
    },
    {
      "entropy": 5.274322271347046,
      "epoch": 1.1349663784822286,
      "grad_norm": 1.625,
      "learning_rate": 0.0004878924044770446,
      "loss": 5.0604,
      "mean_token_accuracy": 0.20086952447891235,
      "num_tokens": 27078602.0,
      "step": 11815
    },
    {
      "entropy": 5.241742944717407,
      "epoch": 1.1354466858789625,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004878813072946843,
      "loss": 5.1326,
      "mean_token_accuracy": 0.2024749681353569,
      "num_tokens": 27089944.0,
      "step": 11820
    },
    {
      "entropy": 5.316387081146241,
      "epoch": 1.1359269932756964,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048787020516985203,
      "loss": 5.097,
      "mean_token_accuracy": 0.21009573638439177,
      "num_tokens": 27100486.0,
      "step": 11825
    },
    {
      "entropy": 5.327194261550903,
      "epoch": 1.1364073006724302,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048785909810280576,
      "loss": 5.1464,
      "mean_token_accuracy": 0.2001217260956764,
      "num_tokens": 27111435.0,
      "step": 11830
    },
    {
      "entropy": 5.339106178283691,
      "epoch": 1.1368876080691643,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004878479860938033,
      "loss": 5.1827,
      "mean_token_accuracy": 0.2010580375790596,
      "num_tokens": 27123482.0,
      "step": 11835
    },
    {
      "entropy": 5.316496467590332,
      "epoch": 1.1373679154658982,
      "grad_norm": 2.59375,
      "learning_rate": 0.00048783686914310266,
      "loss": 5.1081,
      "mean_token_accuracy": 0.20054138600826263,
      "num_tokens": 27135009.0,
      "step": 11840
    },
    {
      "entropy": 5.292483854293823,
      "epoch": 1.137848222862632,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004878257472509619,
      "loss": 5.0763,
      "mean_token_accuracy": 0.2079618513584137,
      "num_tokens": 27145230.0,
      "step": 11845
    },
    {
      "entropy": 5.257307815551758,
      "epoch": 1.138328530259366,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004878146204176392,
      "loss": 5.1102,
      "mean_token_accuracy": 0.20189673453569412,
      "num_tokens": 27156490.0,
      "step": 11850
    },
    {
      "entropy": 5.317572021484375,
      "epoch": 1.1388088376560999,
      "grad_norm": 1.84375,
      "learning_rate": 0.000487803488643393,
      "loss": 5.108,
      "mean_token_accuracy": 0.20475540459156036,
      "num_tokens": 27166727.0,
      "step": 11855
    },
    {
      "entropy": 5.336464738845825,
      "epoch": 1.1392891450528337,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048779235192848166,
      "loss": 5.0706,
      "mean_token_accuracy": 0.20334441661834718,
      "num_tokens": 27177969.0,
      "step": 11860
    },
    {
      "entropy": 5.328475856781006,
      "epoch": 1.1397694524495678,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004877812102731638,
      "loss": 5.2097,
      "mean_token_accuracy": 0.2028002917766571,
      "num_tokens": 27189637.0,
      "step": 11865
    },
    {
      "entropy": 5.419164037704467,
      "epoch": 1.1402497598463017,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048777006367769804,
      "loss": 5.3117,
      "mean_token_accuracy": 0.19987482130527495,
      "num_tokens": 27201124.0,
      "step": 11870
    },
    {
      "entropy": 5.4266434669494625,
      "epoch": 1.1407300672430356,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004877589121423432,
      "loss": 5.1971,
      "mean_token_accuracy": 0.1969528913497925,
      "num_tokens": 27212399.0,
      "step": 11875
    },
    {
      "entropy": 5.347195625305176,
      "epoch": 1.1412103746397695,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004877477556673582,
      "loss": 5.1129,
      "mean_token_accuracy": 0.20323327630758287,
      "num_tokens": 27224469.0,
      "step": 11880
    },
    {
      "entropy": 5.340413570404053,
      "epoch": 1.1416906820365034,
      "grad_norm": 1.546875,
      "learning_rate": 0.000487736594253002,
      "loss": 5.0986,
      "mean_token_accuracy": 0.2063771441578865,
      "num_tokens": 27236026.0,
      "step": 11885
    },
    {
      "entropy": 5.325238943099976,
      "epoch": 1.1421709894332372,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00048772542789953384,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19896686524152757,
      "num_tokens": 27248332.0,
      "step": 11890
    },
    {
      "entropy": 5.320242500305175,
      "epoch": 1.1426512968299711,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048771425660721284,
      "loss": 5.0651,
      "mean_token_accuracy": 0.20700283348560333,
      "num_tokens": 27258701.0,
      "step": 11895
    },
    {
      "entropy": 5.341927719116211,
      "epoch": 1.143131604226705,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048770308037629853,
      "loss": 5.1962,
      "mean_token_accuracy": 0.1995152086019516,
      "num_tokens": 27269808.0,
      "step": 11900
    },
    {
      "entropy": 5.358848667144775,
      "epoch": 1.143611911623439,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004876918992070502,
      "loss": 5.2243,
      "mean_token_accuracy": 0.19571249932050705,
      "num_tokens": 27282440.0,
      "step": 11905
    },
    {
      "entropy": 5.430353212356567,
      "epoch": 1.144092219020173,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004876807130997276,
      "loss": 5.2368,
      "mean_token_accuracy": 0.20419214218854903,
      "num_tokens": 27294430.0,
      "step": 11910
    },
    {
      "entropy": 5.386064529418945,
      "epoch": 1.1445725264169069,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004876695220545903,
      "loss": 5.2157,
      "mean_token_accuracy": 0.20013083666563034,
      "num_tokens": 27305896.0,
      "step": 11915
    },
    {
      "entropy": 5.426707410812378,
      "epoch": 1.1450528338136408,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048765832607189824,
      "loss": 5.2013,
      "mean_token_accuracy": 0.19146447330713273,
      "num_tokens": 27318195.0,
      "step": 11920
    },
    {
      "entropy": 5.2909684658050535,
      "epoch": 1.1455331412103746,
      "grad_norm": 1.5,
      "learning_rate": 0.00048764712515191136,
      "loss": 5.1082,
      "mean_token_accuracy": 0.19882248640060424,
      "num_tokens": 27329055.0,
      "step": 11925
    },
    {
      "entropy": 5.309229183197021,
      "epoch": 1.1460134486071085,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048763591929488966,
      "loss": 5.0994,
      "mean_token_accuracy": 0.2096550866961479,
      "num_tokens": 27340296.0,
      "step": 11930
    },
    {
      "entropy": 5.35607123374939,
      "epoch": 1.1464937560038424,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004876247085010933,
      "loss": 5.1326,
      "mean_token_accuracy": 0.19835602343082429,
      "num_tokens": 27350995.0,
      "step": 11935
    },
    {
      "entropy": 5.312476348876953,
      "epoch": 1.1469740634005763,
      "grad_norm": 1.875,
      "learning_rate": 0.00048761349277078253,
      "loss": 5.1943,
      "mean_token_accuracy": 0.19872767776250838,
      "num_tokens": 27362463.0,
      "step": 11940
    },
    {
      "entropy": 5.3264600276947025,
      "epoch": 1.1474543707973104,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048760227210421775,
      "loss": 5.0463,
      "mean_token_accuracy": 0.2142341449856758,
      "num_tokens": 27373579.0,
      "step": 11945
    },
    {
      "entropy": 5.311580657958984,
      "epoch": 1.1479346781940443,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004875910465016596,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20801135301589965,
      "num_tokens": 27383813.0,
      "step": 11950
    },
    {
      "entropy": 5.319390487670899,
      "epoch": 1.1484149855907781,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004875798159633686,
      "loss": 5.1638,
      "mean_token_accuracy": 0.19501575827598572,
      "num_tokens": 27395651.0,
      "step": 11955
    },
    {
      "entropy": 5.360228490829468,
      "epoch": 1.148895292987512,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004875685804896055,
      "loss": 5.1721,
      "mean_token_accuracy": 0.19930996000766754,
      "num_tokens": 27407827.0,
      "step": 11960
    },
    {
      "entropy": 5.359977388381958,
      "epoch": 1.149375600384246,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004875573400806312,
      "loss": 5.1734,
      "mean_token_accuracy": 0.19832769632339478,
      "num_tokens": 27418777.0,
      "step": 11965
    },
    {
      "entropy": 5.3723530769348145,
      "epoch": 1.1498559077809798,
      "grad_norm": 2.3125,
      "learning_rate": 0.00048754609473670654,
      "loss": 5.1921,
      "mean_token_accuracy": 0.19500161111354827,
      "num_tokens": 27429626.0,
      "step": 11970
    },
    {
      "entropy": 5.429681301116943,
      "epoch": 1.1503362151777137,
      "grad_norm": 1.875,
      "learning_rate": 0.0004875348444580927,
      "loss": 5.3649,
      "mean_token_accuracy": 0.19333478510379792,
      "num_tokens": 27442304.0,
      "step": 11975
    },
    {
      "entropy": 5.349698638916015,
      "epoch": 1.1508165225744476,
      "grad_norm": 1.46875,
      "learning_rate": 0.000487523589245051,
      "loss": 5.1181,
      "mean_token_accuracy": 0.2069244235754013,
      "num_tokens": 27454421.0,
      "step": 11980
    },
    {
      "entropy": 5.262469625473022,
      "epoch": 1.1512968299711814,
      "grad_norm": 1.875,
      "learning_rate": 0.0004875123290978425,
      "loss": 5.0257,
      "mean_token_accuracy": 0.20779014974832535,
      "num_tokens": 27465220.0,
      "step": 11985
    },
    {
      "entropy": 5.276208305358887,
      "epoch": 1.1517771373679155,
      "grad_norm": 2.0,
      "learning_rate": 0.00048750106401672876,
      "loss": 5.1378,
      "mean_token_accuracy": 0.1957992672920227,
      "num_tokens": 27476542.0,
      "step": 11990
    },
    {
      "entropy": 5.275929069519043,
      "epoch": 1.1522574447646494,
      "grad_norm": 2.296875,
      "learning_rate": 0.00048748979400197134,
      "loss": 5.0235,
      "mean_token_accuracy": 0.20864951312541963,
      "num_tokens": 27487721.0,
      "step": 11995
    },
    {
      "entropy": 5.3089026927948,
      "epoch": 1.1527377521613833,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048747851905383183,
      "loss": 5.1233,
      "mean_token_accuracy": 0.20202126801013948,
      "num_tokens": 27498982.0,
      "step": 12000
    },
    {
      "epoch": 1.1527377521613833,
      "eval_entropy": 5.155717872282983,
      "eval_loss": 5.226895809173584,
      "eval_mean_token_accuracy": 0.2056008455589203,
      "eval_num_tokens": 27498982.0,
      "eval_runtime": 26.5965,
      "eval_samples_per_second": 1233.81,
      "eval_steps_per_second": 154.231,
      "step": 12000
    },
    {
      "entropy": 5.246145439147949,
      "epoch": 1.1532180595581172,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004874672391725721,
      "loss": 5.0178,
      "mean_token_accuracy": 0.20342948734760286,
      "num_tokens": 27509863.0,
      "step": 12005
    },
    {
      "entropy": 5.28523211479187,
      "epoch": 1.153698366954851,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004874559543584539,
      "loss": 5.1356,
      "mean_token_accuracy": 0.19701607525348663,
      "num_tokens": 27521061.0,
      "step": 12010
    },
    {
      "entropy": 5.293413591384888,
      "epoch": 1.154178674351585,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004874446646117394,
      "loss": 5.135,
      "mean_token_accuracy": 0.20306088775396347,
      "num_tokens": 27532187.0,
      "step": 12015
    },
    {
      "entropy": 5.301681852340698,
      "epoch": 1.154658981748319,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004874333699326906,
      "loss": 5.1252,
      "mean_token_accuracy": 0.20141319185495377,
      "num_tokens": 27543371.0,
      "step": 12020
    },
    {
      "entropy": 5.278606939315796,
      "epoch": 1.155139289145053,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004874220703215697,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20223018079996108,
      "num_tokens": 27554423.0,
      "step": 12025
    },
    {
      "entropy": 5.3352516174316404,
      "epoch": 1.1556195965417868,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004874107657786391,
      "loss": 5.1271,
      "mean_token_accuracy": 0.20298559069633484,
      "num_tokens": 27566397.0,
      "step": 12030
    },
    {
      "entropy": 5.3967146396636965,
      "epoch": 1.1560999039385207,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048739945630416124,
      "loss": 5.1921,
      "mean_token_accuracy": 0.19905346632003784,
      "num_tokens": 27578668.0,
      "step": 12035
    },
    {
      "entropy": 5.32283296585083,
      "epoch": 1.1565802113352546,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004873881418983987,
      "loss": 5.1435,
      "mean_token_accuracy": 0.1951536774635315,
      "num_tokens": 27589345.0,
      "step": 12040
    },
    {
      "entropy": 5.24552698135376,
      "epoch": 1.1570605187319885,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004873768225616141,
      "loss": 5.0615,
      "mean_token_accuracy": 0.20939509123563765,
      "num_tokens": 27600482.0,
      "step": 12045
    },
    {
      "entropy": 5.339636325836182,
      "epoch": 1.1575408261287223,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048736549829407047,
      "loss": 5.1876,
      "mean_token_accuracy": 0.20021516233682632,
      "num_tokens": 27611543.0,
      "step": 12050
    },
    {
      "entropy": 5.403999757766724,
      "epoch": 1.1580211335254562,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004873541690960305,
      "loss": 5.1326,
      "mean_token_accuracy": 0.20268415659666061,
      "num_tokens": 27623503.0,
      "step": 12055
    },
    {
      "entropy": 5.222760534286499,
      "epoch": 1.15850144092219,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004873428349677573,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21606365442276002,
      "num_tokens": 27633593.0,
      "step": 12060
    },
    {
      "entropy": 5.256800699234009,
      "epoch": 1.1589817483189242,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000487331495909514,
      "loss": 5.2127,
      "mean_token_accuracy": 0.19374803304672242,
      "num_tokens": 27646339.0,
      "step": 12065
    },
    {
      "entropy": 5.231207990646363,
      "epoch": 1.159462055715658,
      "grad_norm": 1.625,
      "learning_rate": 0.00048732015192156383,
      "loss": 4.9379,
      "mean_token_accuracy": 0.22210926562547684,
      "num_tokens": 27657314.0,
      "step": 12070
    },
    {
      "entropy": 5.2947710990905765,
      "epoch": 1.159942363112392,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048730880300417015,
      "loss": 5.149,
      "mean_token_accuracy": 0.19903395771980287,
      "num_tokens": 27668680.0,
      "step": 12075
    },
    {
      "entropy": 5.382332134246826,
      "epoch": 1.1604226705091258,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048729744915759657,
      "loss": 5.1347,
      "mean_token_accuracy": 0.20175010114908218,
      "num_tokens": 27679628.0,
      "step": 12080
    },
    {
      "entropy": 5.320165586471558,
      "epoch": 1.1609029779058597,
      "grad_norm": 1.75,
      "learning_rate": 0.00048728609038210655,
      "loss": 5.177,
      "mean_token_accuracy": 0.20453463941812516,
      "num_tokens": 27691205.0,
      "step": 12085
    },
    {
      "entropy": 5.280611133575439,
      "epoch": 1.1613832853025936,
      "grad_norm": 2.125,
      "learning_rate": 0.00048727472667796395,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20722220242023467,
      "num_tokens": 27702903.0,
      "step": 12090
    },
    {
      "entropy": 5.376653718948364,
      "epoch": 1.1618635926993275,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004872633580454325,
      "loss": 5.2375,
      "mean_token_accuracy": 0.18949190229177476,
      "num_tokens": 27716554.0,
      "step": 12095
    },
    {
      "entropy": 5.391836643218994,
      "epoch": 1.1623439000960616,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048725198448477616,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20616166442632675,
      "num_tokens": 27727931.0,
      "step": 12100
    },
    {
      "entropy": 5.368258857727051,
      "epoch": 1.1628242074927955,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048724060599625893,
      "loss": 5.1735,
      "mean_token_accuracy": 0.20043665468692778,
      "num_tokens": 27739483.0,
      "step": 12105
    },
    {
      "entropy": 5.355756378173828,
      "epoch": 1.1633045148895294,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048722922258014506,
      "loss": 5.1472,
      "mean_token_accuracy": 0.20054960399866104,
      "num_tokens": 27751157.0,
      "step": 12110
    },
    {
      "entropy": 5.429017066955566,
      "epoch": 1.1637848222862632,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004872178342366989,
      "loss": 5.2368,
      "mean_token_accuracy": 0.19319243133068084,
      "num_tokens": 27762190.0,
      "step": 12115
    },
    {
      "entropy": 5.271763944625855,
      "epoch": 1.1642651296829971,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048720644096618475,
      "loss": 5.0994,
      "mean_token_accuracy": 0.20084647834300995,
      "num_tokens": 27773684.0,
      "step": 12120
    },
    {
      "entropy": 5.3564873218536375,
      "epoch": 1.164745437079731,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004871950427688672,
      "loss": 5.1707,
      "mean_token_accuracy": 0.1926524221897125,
      "num_tokens": 27785388.0,
      "step": 12125
    },
    {
      "entropy": 5.301265144348145,
      "epoch": 1.1652257444764649,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048718363964501087,
      "loss": 5.0847,
      "mean_token_accuracy": 0.20568531900644302,
      "num_tokens": 27797603.0,
      "step": 12130
    },
    {
      "entropy": 5.4223939895629885,
      "epoch": 1.1657060518731988,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004871722315948805,
      "loss": 5.2177,
      "mean_token_accuracy": 0.19753820598125457,
      "num_tokens": 27809444.0,
      "step": 12135
    },
    {
      "entropy": 5.399594068527222,
      "epoch": 1.1661863592699326,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004871608186187408,
      "loss": 5.2407,
      "mean_token_accuracy": 0.1932898446917534,
      "num_tokens": 27820213.0,
      "step": 12140
    },
    {
      "entropy": 5.280743503570557,
      "epoch": 1.1666666666666667,
      "grad_norm": 1.75,
      "learning_rate": 0.00048714940071685703,
      "loss": 5.0801,
      "mean_token_accuracy": 0.20481188744306564,
      "num_tokens": 27831564.0,
      "step": 12145
    },
    {
      "entropy": 5.342092800140381,
      "epoch": 1.1671469740634006,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048713797788949405,
      "loss": 5.1831,
      "mean_token_accuracy": 0.1955466851592064,
      "num_tokens": 27843686.0,
      "step": 12150
    },
    {
      "entropy": 5.389429998397827,
      "epoch": 1.1676272814601345,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048712655013691714,
      "loss": 5.2056,
      "mean_token_accuracy": 0.2001350089907646,
      "num_tokens": 27855428.0,
      "step": 12155
    },
    {
      "entropy": 5.471636581420898,
      "epoch": 1.1681075888568684,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048711511745939165,
      "loss": 5.3021,
      "mean_token_accuracy": 0.19671549797058105,
      "num_tokens": 27866733.0,
      "step": 12160
    },
    {
      "entropy": 5.326448202133179,
      "epoch": 1.1685878962536023,
      "grad_norm": 1.578125,
      "learning_rate": 0.000487103679857183,
      "loss": 5.151,
      "mean_token_accuracy": 0.200122632086277,
      "num_tokens": 27878639.0,
      "step": 12165
    },
    {
      "entropy": 5.204430294036865,
      "epoch": 1.1690682036503361,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004870922373305567,
      "loss": 5.1984,
      "mean_token_accuracy": 0.1928955778479576,
      "num_tokens": 27892139.0,
      "step": 12170
    },
    {
      "entropy": 5.423600387573242,
      "epoch": 1.1695485110470702,
      "grad_norm": 1.375,
      "learning_rate": 0.00048708078987977837,
      "loss": 5.2253,
      "mean_token_accuracy": 0.196384496986866,
      "num_tokens": 27904395.0,
      "step": 12175
    },
    {
      "entropy": 5.465410995483398,
      "epoch": 1.1700288184438041,
      "grad_norm": 1.90625,
      "learning_rate": 0.00048706933750511394,
      "loss": 5.192,
      "mean_token_accuracy": 0.19910237044095994,
      "num_tokens": 27916444.0,
      "step": 12180
    },
    {
      "entropy": 5.325769853591919,
      "epoch": 1.170509125840538,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004870578802068292,
      "loss": 5.1319,
      "mean_token_accuracy": 0.21037591695785524,
      "num_tokens": 27926914.0,
      "step": 12185
    },
    {
      "entropy": 5.274808597564697,
      "epoch": 1.170989433237272,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00048704641798519006,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20392880886793135,
      "num_tokens": 27938203.0,
      "step": 12190
    },
    {
      "entropy": 5.417373180389404,
      "epoch": 1.1714697406340058,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048703495084046286,
      "loss": 5.1565,
      "mean_token_accuracy": 0.20045826435089112,
      "num_tokens": 27950463.0,
      "step": 12195
    },
    {
      "entropy": 5.347870492935181,
      "epoch": 1.1719500480307397,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004870234787729137,
      "loss": 5.0758,
      "mean_token_accuracy": 0.196811805665493,
      "num_tokens": 27962312.0,
      "step": 12200
    },
    {
      "entropy": 5.275456523895263,
      "epoch": 1.1724303554274735,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004870120017828089,
      "loss": 5.0766,
      "mean_token_accuracy": 0.20652167946100236,
      "num_tokens": 27973723.0,
      "step": 12205
    },
    {
      "entropy": 5.30847020149231,
      "epoch": 1.1729106628242074,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000487000519870415,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20973498672246932,
      "num_tokens": 27983816.0,
      "step": 12210
    },
    {
      "entropy": 5.212904787063598,
      "epoch": 1.1733909702209413,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004869890330359986,
      "loss": 5.1192,
      "mean_token_accuracy": 0.20577560663223265,
      "num_tokens": 27996516.0,
      "step": 12215
    },
    {
      "entropy": 5.329222202301025,
      "epoch": 1.1738712776176754,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004869775412798262,
      "loss": 5.1483,
      "mean_token_accuracy": 0.2021143138408661,
      "num_tokens": 28008256.0,
      "step": 12220
    },
    {
      "entropy": 5.289189958572388,
      "epoch": 1.1743515850144093,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048696604460216476,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21567831486463546,
      "num_tokens": 28018882.0,
      "step": 12225
    },
    {
      "entropy": 5.272204160690308,
      "epoch": 1.1748318924111432,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048695454300328123,
      "loss": 5.1037,
      "mean_token_accuracy": 0.20084813386201858,
      "num_tokens": 28030399.0,
      "step": 12230
    },
    {
      "entropy": 5.33041787147522,
      "epoch": 1.175312199807877,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048694303648344256,
      "loss": 5.1599,
      "mean_token_accuracy": 0.20343490689992905,
      "num_tokens": 28041621.0,
      "step": 12235
    },
    {
      "entropy": 5.338892602920533,
      "epoch": 1.175792507204611,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048693152504291595,
      "loss": 5.0614,
      "mean_token_accuracy": 0.20413374453783034,
      "num_tokens": 28053198.0,
      "step": 12240
    },
    {
      "entropy": 5.289673662185669,
      "epoch": 1.1762728146013448,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004869200086819686,
      "loss": 5.0952,
      "mean_token_accuracy": 0.20140086561441423,
      "num_tokens": 28064095.0,
      "step": 12245
    },
    {
      "entropy": 5.292538404464722,
      "epoch": 1.1767531219980787,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048690848740086796,
      "loss": 5.1562,
      "mean_token_accuracy": 0.196707783639431,
      "num_tokens": 28075665.0,
      "step": 12250
    },
    {
      "entropy": 5.355053424835205,
      "epoch": 1.1772334293948128,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004868969611998814,
      "loss": 5.1979,
      "mean_token_accuracy": 0.19931446462869645,
      "num_tokens": 28089166.0,
      "step": 12255
    },
    {
      "entropy": 5.494907474517822,
      "epoch": 1.1777137367915467,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004868854300792767,
      "loss": 5.2036,
      "mean_token_accuracy": 0.19365310072898864,
      "num_tokens": 28101379.0,
      "step": 12260
    },
    {
      "entropy": 5.355603981018066,
      "epoch": 1.1781940441882806,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048687389403932144,
      "loss": 5.1648,
      "mean_token_accuracy": 0.19845588207244874,
      "num_tokens": 28112836.0,
      "step": 12265
    },
    {
      "entropy": 5.294369173049927,
      "epoch": 1.1786743515850144,
      "grad_norm": 5.0625,
      "learning_rate": 0.0004868623530802835,
      "loss": 5.2434,
      "mean_token_accuracy": 0.1944652482867241,
      "num_tokens": 28125039.0,
      "step": 12270
    },
    {
      "entropy": 5.331742906570435,
      "epoch": 1.1791546589817483,
      "grad_norm": 2.109375,
      "learning_rate": 0.00048685080720243086,
      "loss": 5.1103,
      "mean_token_accuracy": 0.19970922023057938,
      "num_tokens": 28137504.0,
      "step": 12275
    },
    {
      "entropy": 5.380903863906861,
      "epoch": 1.1796349663784822,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004868392564060315,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20417306423187256,
      "num_tokens": 28149483.0,
      "step": 12280
    },
    {
      "entropy": 5.29079155921936,
      "epoch": 1.180115273775216,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004868277006913537,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20020388066768646,
      "num_tokens": 28161001.0,
      "step": 12285
    },
    {
      "entropy": 5.331949186325073,
      "epoch": 1.18059558117195,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004868161400586656,
      "loss": 5.1704,
      "mean_token_accuracy": 0.20341708213090898,
      "num_tokens": 28171146.0,
      "step": 12290
    },
    {
      "entropy": 5.413556718826294,
      "epoch": 1.1810758885686838,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004868045745082357,
      "loss": 5.1979,
      "mean_token_accuracy": 0.19392373114824296,
      "num_tokens": 28182870.0,
      "step": 12295
    },
    {
      "entropy": 5.322707986831665,
      "epoch": 1.181556195965418,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004867930040403326,
      "loss": 5.1897,
      "mean_token_accuracy": 0.2018422529101372,
      "num_tokens": 28194457.0,
      "step": 12300
    },
    {
      "entropy": 5.352476739883423,
      "epoch": 1.1820365033621518,
      "grad_norm": 1.75,
      "learning_rate": 0.00048678142865522475,
      "loss": 5.2136,
      "mean_token_accuracy": 0.19604827910661698,
      "num_tokens": 28205818.0,
      "step": 12305
    },
    {
      "entropy": 5.330062961578369,
      "epoch": 1.1825168107588857,
      "grad_norm": 1.421875,
      "learning_rate": 0.000486769848353181,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20487801730632782,
      "num_tokens": 28216124.0,
      "step": 12310
    },
    {
      "entropy": 5.3179103374481205,
      "epoch": 1.1829971181556196,
      "grad_norm": 3.09375,
      "learning_rate": 0.00048675826313447027,
      "loss": 5.1436,
      "mean_token_accuracy": 0.19761122912168502,
      "num_tokens": 28227485.0,
      "step": 12315
    },
    {
      "entropy": 5.329595279693604,
      "epoch": 1.1834774255523535,
      "grad_norm": 2.265625,
      "learning_rate": 0.00048674667299936135,
      "loss": 5.176,
      "mean_token_accuracy": 0.20214862525463104,
      "num_tokens": 28239007.0,
      "step": 12320
    },
    {
      "entropy": 5.3753503322601315,
      "epoch": 1.1839577329490873,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048673507794812356,
      "loss": 5.1659,
      "mean_token_accuracy": 0.1986163154244423,
      "num_tokens": 28250926.0,
      "step": 12325
    },
    {
      "entropy": 5.365655946731567,
      "epoch": 1.1844380403458212,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004867234779810259,
      "loss": 5.1613,
      "mean_token_accuracy": 0.20166646391153337,
      "num_tokens": 28263182.0,
      "step": 12330
    },
    {
      "entropy": 5.244043493270874,
      "epoch": 1.1849183477425553,
      "grad_norm": 1.875,
      "learning_rate": 0.0004867118730983378,
      "loss": 5.0998,
      "mean_token_accuracy": 0.20676581263542176,
      "num_tokens": 28274666.0,
      "step": 12335
    },
    {
      "entropy": 5.202615261077881,
      "epoch": 1.1853986551392892,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004867002633003286,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21738803386688232,
      "num_tokens": 28286007.0,
      "step": 12340
    },
    {
      "entropy": 5.268519306182862,
      "epoch": 1.185878962536023,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000486688648587268,
      "loss": 5.0781,
      "mean_token_accuracy": 0.20838257074356079,
      "num_tokens": 28298262.0,
      "step": 12345
    },
    {
      "entropy": 5.241210079193115,
      "epoch": 1.186359269932757,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004866770289594256,
      "loss": 5.0825,
      "mean_token_accuracy": 0.20715923607349396,
      "num_tokens": 28309428.0,
      "step": 12350
    },
    {
      "entropy": 5.329868936538697,
      "epoch": 1.1868395773294909,
      "grad_norm": 9.5625,
      "learning_rate": 0.00048666540441707107,
      "loss": 5.1945,
      "mean_token_accuracy": 0.2026727169752121,
      "num_tokens": 28319609.0,
      "step": 12355
    },
    {
      "entropy": 5.402388715744019,
      "epoch": 1.1873198847262247,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004866537749604744,
      "loss": 5.1765,
      "mean_token_accuracy": 0.1958605855703354,
      "num_tokens": 28330526.0,
      "step": 12360
    },
    {
      "entropy": 5.3303382873535154,
      "epoch": 1.1878001921229586,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00048664214058990546,
      "loss": 5.1434,
      "mean_token_accuracy": 0.1984109103679657,
      "num_tokens": 28343919.0,
      "step": 12365
    },
    {
      "entropy": 5.373897695541382,
      "epoch": 1.1882804995196925,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004866305013056346,
      "loss": 5.1447,
      "mean_token_accuracy": 0.201043701171875,
      "num_tokens": 28354850.0,
      "step": 12370
    },
    {
      "entropy": 5.2233014583587645,
      "epoch": 1.1887608069164266,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004866188571079318,
      "loss": 5.0463,
      "mean_token_accuracy": 0.20991920679807663,
      "num_tokens": 28365699.0,
      "step": 12375
    },
    {
      "entropy": 5.21318941116333,
      "epoch": 1.1892411143131605,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004866072079970676,
      "loss": 5.0109,
      "mean_token_accuracy": 0.20985946655273438,
      "num_tokens": 28377053.0,
      "step": 12380
    },
    {
      "entropy": 5.363075876235962,
      "epoch": 1.1897214217098944,
      "grad_norm": 2.203125,
      "learning_rate": 0.00048659555397331236,
      "loss": 5.1248,
      "mean_token_accuracy": 0.20071543008089066,
      "num_tokens": 28387505.0,
      "step": 12385
    },
    {
      "entropy": 5.298946571350098,
      "epoch": 1.1902017291066282,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004865838950369366,
      "loss": 5.1352,
      "mean_token_accuracy": 0.1979212835431099,
      "num_tokens": 28399308.0,
      "step": 12390
    },
    {
      "entropy": 5.2933995723724365,
      "epoch": 1.1906820365033621,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048657223118821116,
      "loss": 5.1183,
      "mean_token_accuracy": 0.21133489906787872,
      "num_tokens": 28410860.0,
      "step": 12395
    },
    {
      "entropy": 5.210184764862061,
      "epoch": 1.191162343900096,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048656056242740665,
      "loss": 4.9748,
      "mean_token_accuracy": 0.20621041357517242,
      "num_tokens": 28421867.0,
      "step": 12400
    },
    {
      "entropy": 5.293907165527344,
      "epoch": 1.19164265129683,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004865488887547942,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20088723450899124,
      "num_tokens": 28433375.0,
      "step": 12405
    },
    {
      "entropy": 5.3554199695587155,
      "epoch": 1.192122958693564,
      "grad_norm": 2.75,
      "learning_rate": 0.0004865372101706446,
      "loss": 5.1599,
      "mean_token_accuracy": 0.19720466285943986,
      "num_tokens": 28445242.0,
      "step": 12410
    },
    {
      "entropy": 5.2408043384552006,
      "epoch": 1.1926032660902979,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004865255266752292,
      "loss": 5.0883,
      "mean_token_accuracy": 0.20338778644800187,
      "num_tokens": 28455904.0,
      "step": 12415
    },
    {
      "entropy": 5.358150815963745,
      "epoch": 1.1930835734870318,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004865138382688191,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20472579896450044,
      "num_tokens": 28465768.0,
      "step": 12420
    },
    {
      "entropy": 5.369869899749756,
      "epoch": 1.1935638808837656,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004865021449516859,
      "loss": 5.1899,
      "mean_token_accuracy": 0.20407237559556962,
      "num_tokens": 28476489.0,
      "step": 12425
    },
    {
      "entropy": 5.269305801391601,
      "epoch": 1.1940441882804995,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004864904467241008,
      "loss": 5.0634,
      "mean_token_accuracy": 0.21202369183301925,
      "num_tokens": 28487684.0,
      "step": 12430
    },
    {
      "entropy": 5.332604837417603,
      "epoch": 1.1945244956772334,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048647874358633556,
      "loss": 5.109,
      "mean_token_accuracy": 0.1990312084555626,
      "num_tokens": 28499083.0,
      "step": 12435
    },
    {
      "entropy": 5.3246994972229,
      "epoch": 1.1950048030739673,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00048646703553866183,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20161052346229552,
      "num_tokens": 28510057.0,
      "step": 12440
    },
    {
      "entropy": 5.237467622756958,
      "epoch": 1.1954851104707012,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004864553225813515,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20591011494398118,
      "num_tokens": 28522787.0,
      "step": 12445
    },
    {
      "entropy": 5.398256015777588,
      "epoch": 1.195965417867435,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004864436047146765,
      "loss": 5.1539,
      "mean_token_accuracy": 0.1992432415485382,
      "num_tokens": 28534353.0,
      "step": 12450
    },
    {
      "entropy": 5.425845432281494,
      "epoch": 1.1964457252641691,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048643188193890874,
      "loss": 5.2272,
      "mean_token_accuracy": 0.19161331802606582,
      "num_tokens": 28546079.0,
      "step": 12455
    },
    {
      "entropy": 5.395985221862793,
      "epoch": 1.196926032660903,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004864201542543206,
      "loss": 5.1972,
      "mean_token_accuracy": 0.19638125449419022,
      "num_tokens": 28557637.0,
      "step": 12460
    },
    {
      "entropy": 5.258239698410034,
      "epoch": 1.197406340057637,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004864084216611843,
      "loss": 5.1088,
      "mean_token_accuracy": 0.21078716963529587,
      "num_tokens": 28568809.0,
      "step": 12465
    },
    {
      "entropy": 5.334320068359375,
      "epoch": 1.1978866474543708,
      "grad_norm": 2.375,
      "learning_rate": 0.00048639668415977207,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19827519804239274,
      "num_tokens": 28579766.0,
      "step": 12470
    },
    {
      "entropy": 5.396507596969604,
      "epoch": 1.1983669548511047,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00048638494175035665,
      "loss": 5.0654,
      "mean_token_accuracy": 0.21104428470134734,
      "num_tokens": 28590898.0,
      "step": 12475
    },
    {
      "entropy": 5.275398302078247,
      "epoch": 1.1988472622478386,
      "grad_norm": 1.75,
      "learning_rate": 0.0004863731944332105,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20094024091959,
      "num_tokens": 28603380.0,
      "step": 12480
    },
    {
      "entropy": 5.314701747894287,
      "epoch": 1.1993275696445724,
      "grad_norm": 2.96875,
      "learning_rate": 0.0004863614422086065,
      "loss": 5.2315,
      "mean_token_accuracy": 0.1967095836997032,
      "num_tokens": 28615509.0,
      "step": 12485
    },
    {
      "entropy": 5.344776725769043,
      "epoch": 1.1998078770413065,
      "grad_norm": 4.15625,
      "learning_rate": 0.0004863496850768174,
      "loss": 5.1615,
      "mean_token_accuracy": 0.20454192161560059,
      "num_tokens": 28627353.0,
      "step": 12490
    },
    {
      "entropy": 5.3155992984771725,
      "epoch": 1.2002881844380404,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004863379230381162,
      "loss": 5.006,
      "mean_token_accuracy": 0.21289974302053452,
      "num_tokens": 28638217.0,
      "step": 12495
    },
    {
      "entropy": 5.299427938461304,
      "epoch": 1.2007684918347743,
      "grad_norm": 1.5625,
      "learning_rate": 0.000486326156092776,
      "loss": 5.0678,
      "mean_token_accuracy": 0.20807949602603912,
      "num_tokens": 28648862.0,
      "step": 12500
    },
    {
      "entropy": 5.300096654891968,
      "epoch": 1.2012487992315082,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00048631438424106985,
      "loss": 5.1712,
      "mean_token_accuracy": 0.20749196857213975,
      "num_tokens": 28661561.0,
      "step": 12505
    },
    {
      "entropy": 5.338607978820801,
      "epoch": 1.201729106628242,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048630260748327124,
      "loss": 5.1918,
      "mean_token_accuracy": 0.2001659482717514,
      "num_tokens": 28674851.0,
      "step": 12510
    },
    {
      "entropy": 5.331305694580078,
      "epoch": 1.202209414024976,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048629082581965355,
      "loss": 5.1677,
      "mean_token_accuracy": 0.19896040856838226,
      "num_tokens": 28688444.0,
      "step": 12515
    },
    {
      "entropy": 5.399853038787842,
      "epoch": 1.2026897214217098,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048627903925049033,
      "loss": 5.1955,
      "mean_token_accuracy": 0.19806760996580125,
      "num_tokens": 28699374.0,
      "step": 12520
    },
    {
      "entropy": 5.30116982460022,
      "epoch": 1.2031700288184437,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048626724777605507,
      "loss": 5.1658,
      "mean_token_accuracy": 0.20100985169410707,
      "num_tokens": 28711283.0,
      "step": 12525
    },
    {
      "entropy": 5.334057855606079,
      "epoch": 1.2036503362151778,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004862554513966217,
      "loss": 5.1591,
      "mean_token_accuracy": 0.19736357778310776,
      "num_tokens": 28723445.0,
      "step": 12530
    },
    {
      "entropy": 5.301232528686524,
      "epoch": 1.2041306436119117,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048624365011246405,
      "loss": 5.112,
      "mean_token_accuracy": 0.20565079301595687,
      "num_tokens": 28734505.0,
      "step": 12535
    },
    {
      "entropy": 5.223905801773071,
      "epoch": 1.2046109510086456,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004862318439238561,
      "loss": 5.0672,
      "mean_token_accuracy": 0.20763472616672515,
      "num_tokens": 28745476.0,
      "step": 12540
    },
    {
      "entropy": 5.265627241134643,
      "epoch": 1.2050912584053795,
      "grad_norm": 1.921875,
      "learning_rate": 0.000486220032831072,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2083522617816925,
      "num_tokens": 28755645.0,
      "step": 12545
    },
    {
      "entropy": 5.352916812896728,
      "epoch": 1.2055715658021133,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004862082168343859,
      "loss": 5.1525,
      "mean_token_accuracy": 0.20227930843830108,
      "num_tokens": 28767911.0,
      "step": 12550
    },
    {
      "entropy": 5.305215740203858,
      "epoch": 1.2060518731988472,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004861963959340722,
      "loss": 5.1387,
      "mean_token_accuracy": 0.20126512944698333,
      "num_tokens": 28779418.0,
      "step": 12555
    },
    {
      "entropy": 5.324071311950684,
      "epoch": 1.206532180595581,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004861845701304053,
      "loss": 5.2033,
      "mean_token_accuracy": 0.19455412328243255,
      "num_tokens": 28790857.0,
      "step": 12560
    },
    {
      "entropy": 5.406786966323852,
      "epoch": 1.2070124879923152,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048617273942365977,
      "loss": 5.2036,
      "mean_token_accuracy": 0.19559877067804338,
      "num_tokens": 28802975.0,
      "step": 12565
    },
    {
      "entropy": 5.3403526782989506,
      "epoch": 1.207492795389049,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004861609038141103,
      "loss": 5.1426,
      "mean_token_accuracy": 0.20229509323835373,
      "num_tokens": 28813874.0,
      "step": 12570
    },
    {
      "entropy": 5.34351487159729,
      "epoch": 1.207973102785783,
      "grad_norm": 2.125,
      "learning_rate": 0.00048614906330203165,
      "loss": 5.0988,
      "mean_token_accuracy": 0.20050637573003768,
      "num_tokens": 28826278.0,
      "step": 12575
    },
    {
      "entropy": 5.3576273918151855,
      "epoch": 1.2084534101825168,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004861372178876987,
      "loss": 5.1856,
      "mean_token_accuracy": 0.2016836956143379,
      "num_tokens": 28838607.0,
      "step": 12580
    },
    {
      "entropy": 5.289024734497071,
      "epoch": 1.2089337175792507,
      "grad_norm": 2.265625,
      "learning_rate": 0.00048612536757138653,
      "loss": 5.1401,
      "mean_token_accuracy": 0.20274148434400557,
      "num_tokens": 28849801.0,
      "step": 12585
    },
    {
      "entropy": 5.223912858963013,
      "epoch": 1.2094140249759846,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004861135123533702,
      "loss": 4.991,
      "mean_token_accuracy": 0.21283254027366638,
      "num_tokens": 28861312.0,
      "step": 12590
    },
    {
      "entropy": 5.292352342605591,
      "epoch": 1.2098943323727185,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048610165223392503,
      "loss": 5.1561,
      "mean_token_accuracy": 0.19956706017255782,
      "num_tokens": 28872337.0,
      "step": 12595
    },
    {
      "entropy": 5.347715520858765,
      "epoch": 1.2103746397694524,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004860897872133263,
      "loss": 5.0174,
      "mean_token_accuracy": 0.21256937384605407,
      "num_tokens": 28883263.0,
      "step": 12600
    },
    {
      "entropy": 5.32048397064209,
      "epoch": 1.2108549471661862,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004860779172918496,
      "loss": 5.1042,
      "mean_token_accuracy": 0.20354340821504593,
      "num_tokens": 28896166.0,
      "step": 12605
    },
    {
      "entropy": 5.321988916397094,
      "epoch": 1.2113352545629203,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004860660424697704,
      "loss": 5.1581,
      "mean_token_accuracy": 0.2010759100317955,
      "num_tokens": 28906928.0,
      "step": 12610
    },
    {
      "entropy": 5.3950645446777346,
      "epoch": 1.2118155619596542,
      "grad_norm": 2.265625,
      "learning_rate": 0.00048605416274736434,
      "loss": 5.2016,
      "mean_token_accuracy": 0.19412957280874252,
      "num_tokens": 28918328.0,
      "step": 12615
    },
    {
      "entropy": 5.37586088180542,
      "epoch": 1.2122958693563881,
      "grad_norm": 2.25,
      "learning_rate": 0.00048604227812490744,
      "loss": 5.1236,
      "mean_token_accuracy": 0.2029413402080536,
      "num_tokens": 28930327.0,
      "step": 12620
    },
    {
      "entropy": 5.312181425094605,
      "epoch": 1.212776176753122,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048603038860267546,
      "loss": 5.1161,
      "mean_token_accuracy": 0.2062818691134453,
      "num_tokens": 28942166.0,
      "step": 12625
    },
    {
      "entropy": 5.251681280136109,
      "epoch": 1.2132564841498559,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004860184941809445,
      "loss": 4.9994,
      "mean_token_accuracy": 0.21593582183122634,
      "num_tokens": 28953159.0,
      "step": 12630
    },
    {
      "entropy": 5.2858906269073485,
      "epoch": 1.2137367915465898,
      "grad_norm": 2.40625,
      "learning_rate": 0.00048600659485999073,
      "loss": 5.059,
      "mean_token_accuracy": 0.20741520673036576,
      "num_tokens": 28963897.0,
      "step": 12635
    },
    {
      "entropy": 5.254169082641601,
      "epoch": 1.2142170989433236,
      "grad_norm": 1.890625,
      "learning_rate": 0.00048599469064009027,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2046999827027321,
      "num_tokens": 28974452.0,
      "step": 12640
    },
    {
      "entropy": 5.2533927917480465,
      "epoch": 1.2146974063400577,
      "grad_norm": 2.671875,
      "learning_rate": 0.00048598278152151974,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20632339864969254,
      "num_tokens": 28984664.0,
      "step": 12645
    },
    {
      "entropy": 5.337751293182373,
      "epoch": 1.2151777137367916,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004859708675045555,
      "loss": 5.0986,
      "mean_token_accuracy": 0.20948852896690368,
      "num_tokens": 28994992.0,
      "step": 12650
    },
    {
      "entropy": 5.335508823394775,
      "epoch": 1.2156580211335255,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004859589485894741,
      "loss": 5.1247,
      "mean_token_accuracy": 0.19865890890359877,
      "num_tokens": 29006190.0,
      "step": 12655
    },
    {
      "entropy": 5.288554430007935,
      "epoch": 1.2161383285302594,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004859470247765524,
      "loss": 5.0943,
      "mean_token_accuracy": 0.1990597978234291,
      "num_tokens": 29017987.0,
      "step": 12660
    },
    {
      "entropy": 5.26961236000061,
      "epoch": 1.2166186359269933,
      "grad_norm": 1.375,
      "learning_rate": 0.0004859350960660671,
      "loss": 5.1137,
      "mean_token_accuracy": 0.20575728118419648,
      "num_tokens": 29030012.0,
      "step": 12665
    },
    {
      "entropy": 5.403852415084839,
      "epoch": 1.2170989433237271,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004859231624582953,
      "loss": 5.1937,
      "mean_token_accuracy": 0.19742430597543717,
      "num_tokens": 29041684.0,
      "step": 12670
    },
    {
      "entropy": 5.325316905975342,
      "epoch": 1.217579250720461,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048591122395351394,
      "loss": 5.0624,
      "mean_token_accuracy": 0.20583488643169404,
      "num_tokens": 29052525.0,
      "step": 12675
    },
    {
      "entropy": 5.279381513595581,
      "epoch": 1.218059558117195,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004858992805520003,
      "loss": 5.0883,
      "mean_token_accuracy": 0.20701997280120848,
      "num_tokens": 29063471.0,
      "step": 12680
    },
    {
      "entropy": 5.248632431030273,
      "epoch": 1.218539865513929,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048588733225403153,
      "loss": 5.0294,
      "mean_token_accuracy": 0.2102026790380478,
      "num_tokens": 29074616.0,
      "step": 12685
    },
    {
      "entropy": 5.2860675811767575,
      "epoch": 1.219020172910663,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004858753790598851,
      "loss": 5.1577,
      "mean_token_accuracy": 0.20010307133197786,
      "num_tokens": 29086603.0,
      "step": 12690
    },
    {
      "entropy": 5.317056560516358,
      "epoch": 1.2195004803073968,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004858634209698386,
      "loss": 5.1254,
      "mean_token_accuracy": 0.20735129565000535,
      "num_tokens": 29097232.0,
      "step": 12695
    },
    {
      "entropy": 5.439657592773438,
      "epoch": 1.2199807877041307,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048585145798416956,
      "loss": 5.196,
      "mean_token_accuracy": 0.19617790579795838,
      "num_tokens": 29107453.0,
      "step": 12700
    },
    {
      "entropy": 5.262882041931152,
      "epoch": 1.2204610951008645,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004858394901031558,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20033658742904664,
      "num_tokens": 29120255.0,
      "step": 12705
    },
    {
      "entropy": 5.348303318023682,
      "epoch": 1.2209414024975984,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004858275173270751,
      "loss": 5.1484,
      "mean_token_accuracy": 0.1961182415485382,
      "num_tokens": 29130892.0,
      "step": 12710
    },
    {
      "entropy": 5.413280963897705,
      "epoch": 1.2214217098943323,
      "grad_norm": 2.40625,
      "learning_rate": 0.00048581553965620553,
      "loss": 5.1766,
      "mean_token_accuracy": 0.19905781149864196,
      "num_tokens": 29142613.0,
      "step": 12715
    },
    {
      "entropy": 5.336814260482788,
      "epoch": 1.2219020172910664,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00048580355709082506,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20638584345579147,
      "num_tokens": 29153045.0,
      "step": 12720
    },
    {
      "entropy": 5.207999515533447,
      "epoch": 1.2223823246878003,
      "grad_norm": 1.8984375,
      "learning_rate": 0.000485791569631212,
      "loss": 5.0029,
      "mean_token_accuracy": 0.208831225335598,
      "num_tokens": 29163767.0,
      "step": 12725
    },
    {
      "entropy": 5.295343494415283,
      "epoch": 1.2228626320845342,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004857795772776446,
      "loss": 5.1485,
      "mean_token_accuracy": 0.20012069046497344,
      "num_tokens": 29174277.0,
      "step": 12730
    },
    {
      "entropy": 5.368566799163818,
      "epoch": 1.223342939481268,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048576758003040127,
      "loss": 5.1947,
      "mean_token_accuracy": 0.20096989274024962,
      "num_tokens": 29186078.0,
      "step": 12735
    },
    {
      "entropy": 5.293829298019409,
      "epoch": 1.223823246878002,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00048575557788976066,
      "loss": 5.0662,
      "mean_token_accuracy": 0.2092044770717621,
      "num_tokens": 29198056.0,
      "step": 12740
    },
    {
      "entropy": 5.3911127090454105,
      "epoch": 1.2243035542747358,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004857435708560013,
      "loss": 5.2564,
      "mean_token_accuracy": 0.1908797338604927,
      "num_tokens": 29209892.0,
      "step": 12745
    },
    {
      "entropy": 5.419103193283081,
      "epoch": 1.2247838616714697,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048573155892940204,
      "loss": 5.138,
      "mean_token_accuracy": 0.20095694661140442,
      "num_tokens": 29220813.0,
      "step": 12750
    },
    {
      "entropy": 5.353153276443481,
      "epoch": 1.2252641690682036,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00048571954211024164,
      "loss": 5.1284,
      "mean_token_accuracy": 0.19931820183992385,
      "num_tokens": 29232807.0,
      "step": 12755
    },
    {
      "entropy": 5.333857107162475,
      "epoch": 1.2257444764649374,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048570752039879924,
      "loss": 5.1184,
      "mean_token_accuracy": 0.19930154979228973,
      "num_tokens": 29243721.0,
      "step": 12760
    },
    {
      "entropy": 5.285833215713501,
      "epoch": 1.2262247838616716,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004856954937953539,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20339198261499405,
      "num_tokens": 29255783.0,
      "step": 12765
    },
    {
      "entropy": 5.251891422271728,
      "epoch": 1.2267050912584054,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004856834623001848,
      "loss": 5.0497,
      "mean_token_accuracy": 0.20860035270452498,
      "num_tokens": 29266857.0,
      "step": 12770
    },
    {
      "entropy": 5.225630617141723,
      "epoch": 1.2271853986551393,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004856714259135713,
      "loss": 5.0136,
      "mean_token_accuracy": 0.21046051979064942,
      "num_tokens": 29278479.0,
      "step": 12775
    },
    {
      "entropy": 5.420551061630249,
      "epoch": 1.2276657060518732,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004856593846357929,
      "loss": 5.0905,
      "mean_token_accuracy": 0.20781322419643403,
      "num_tokens": 29289369.0,
      "step": 12780
    },
    {
      "entropy": 5.3145161151885985,
      "epoch": 1.228146013448607,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004856473384671291,
      "loss": 5.1056,
      "mean_token_accuracy": 0.1989370748400688,
      "num_tokens": 29301015.0,
      "step": 12785
    },
    {
      "entropy": 5.3360082626342775,
      "epoch": 1.228626320845341,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048563528740785955,
      "loss": 5.1799,
      "mean_token_accuracy": 0.19770618230104448,
      "num_tokens": 29312638.0,
      "step": 12790
    },
    {
      "entropy": 5.33776330947876,
      "epoch": 1.2291066282420748,
      "grad_norm": 2.8125,
      "learning_rate": 0.00048562323145826414,
      "loss": 5.1069,
      "mean_token_accuracy": 0.20949000418186187,
      "num_tokens": 29323540.0,
      "step": 12795
    },
    {
      "entropy": 5.306763172149658,
      "epoch": 1.229586935638809,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004856111706186227,
      "loss": 5.1361,
      "mean_token_accuracy": 0.20181840360164643,
      "num_tokens": 29334287.0,
      "step": 12800
    },
    {
      "entropy": 5.308620977401733,
      "epoch": 1.2300672430355428,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048559910488921534,
      "loss": 5.0914,
      "mean_token_accuracy": 0.19543766975402832,
      "num_tokens": 29345278.0,
      "step": 12805
    },
    {
      "entropy": 5.400620365142823,
      "epoch": 1.2305475504322767,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000485587034270322,
      "loss": 5.173,
      "mean_token_accuracy": 0.1949384719133377,
      "num_tokens": 29357297.0,
      "step": 12810
    },
    {
      "entropy": 5.335856628417969,
      "epoch": 1.2310278578290106,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000485574958762223,
      "loss": 5.0965,
      "mean_token_accuracy": 0.20904208421707154,
      "num_tokens": 29368513.0,
      "step": 12815
    },
    {
      "entropy": 5.309087657928467,
      "epoch": 1.2315081652257445,
      "grad_norm": 2.359375,
      "learning_rate": 0.00048556287836519886,
      "loss": 5.1202,
      "mean_token_accuracy": 0.19918393045663835,
      "num_tokens": 29379227.0,
      "step": 12820
    },
    {
      "entropy": 5.2387528896331785,
      "epoch": 1.2319884726224783,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004855507930795299,
      "loss": 5.1182,
      "mean_token_accuracy": 0.20332754105329515,
      "num_tokens": 29390048.0,
      "step": 12825
    },
    {
      "entropy": 5.348978042602539,
      "epoch": 1.2324687800192122,
      "grad_norm": 2.609375,
      "learning_rate": 0.00048553870290549665,
      "loss": 5.1369,
      "mean_token_accuracy": 0.20016331672668458,
      "num_tokens": 29401252.0,
      "step": 12830
    },
    {
      "entropy": 5.356282901763916,
      "epoch": 1.232949087415946,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048552660784338,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20935385674238205,
      "num_tokens": 29413122.0,
      "step": 12835
    },
    {
      "entropy": 5.397066879272461,
      "epoch": 1.23342939481268,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004855145078934606,
      "loss": 5.191,
      "mean_token_accuracy": 0.19734654426574708,
      "num_tokens": 29423240.0,
      "step": 12840
    },
    {
      "entropy": 5.271270370483398,
      "epoch": 1.233909702209414,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004855024030560195,
      "loss": 5.0401,
      "mean_token_accuracy": 0.21129116266965867,
      "num_tokens": 29433818.0,
      "step": 12845
    },
    {
      "entropy": 5.386311340332031,
      "epoch": 1.234390009606148,
      "grad_norm": 2.9375,
      "learning_rate": 0.0004854902933313376,
      "loss": 5.2119,
      "mean_token_accuracy": 0.19911233633756636,
      "num_tokens": 29446927.0,
      "step": 12850
    },
    {
      "entropy": 5.393089342117309,
      "epoch": 1.2348703170028819,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048547817871969607,
      "loss": 5.1532,
      "mean_token_accuracy": 0.20255052000284196,
      "num_tokens": 29458092.0,
      "step": 12855
    },
    {
      "entropy": 5.352181529998779,
      "epoch": 1.2353506243996157,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048546605922137633,
      "loss": 5.2139,
      "mean_token_accuracy": 0.19526202380657195,
      "num_tokens": 29470395.0,
      "step": 12860
    },
    {
      "entropy": 5.384905242919922,
      "epoch": 1.2358309317963496,
      "grad_norm": 4.65625,
      "learning_rate": 0.0004854539348366596,
      "loss": 5.1778,
      "mean_token_accuracy": 0.2009596824645996,
      "num_tokens": 29481177.0,
      "step": 12865
    },
    {
      "entropy": 5.336521339416504,
      "epoch": 1.2363112391930835,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004854418055658274,
      "loss": 5.065,
      "mean_token_accuracy": 0.2043983444571495,
      "num_tokens": 29492305.0,
      "step": 12870
    },
    {
      "entropy": 5.336456346511841,
      "epoch": 1.2367915465898176,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048542967140916134,
      "loss": 5.1159,
      "mean_token_accuracy": 0.2075643941760063,
      "num_tokens": 29503237.0,
      "step": 12875
    },
    {
      "entropy": 5.2819195747375485,
      "epoch": 1.2372718539865515,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004854175323669432,
      "loss": 5.1161,
      "mean_token_accuracy": 0.2041839689016342,
      "num_tokens": 29514195.0,
      "step": 12880
    },
    {
      "entropy": 5.338493490219117,
      "epoch": 1.2377521613832854,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004854053884394547,
      "loss": 5.2315,
      "mean_token_accuracy": 0.19475714713335038,
      "num_tokens": 29526110.0,
      "step": 12885
    },
    {
      "entropy": 5.308443593978882,
      "epoch": 1.2382324687800192,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048539323962697796,
      "loss": 5.1188,
      "mean_token_accuracy": 0.20441027581691742,
      "num_tokens": 29536912.0,
      "step": 12890
    },
    {
      "entropy": 5.347447919845581,
      "epoch": 1.2387127761767531,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004853810859297949,
      "loss": 5.1232,
      "mean_token_accuracy": 0.20002758055925368,
      "num_tokens": 29548070.0,
      "step": 12895
    },
    {
      "entropy": 5.33713698387146,
      "epoch": 1.239193083573487,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048536892734818773,
      "loss": 5.1774,
      "mean_token_accuracy": 0.1967310607433319,
      "num_tokens": 29560713.0,
      "step": 12900
    },
    {
      "entropy": 5.326811695098877,
      "epoch": 1.239673390970221,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004853567638824387,
      "loss": 5.0633,
      "mean_token_accuracy": 0.21816372573375703,
      "num_tokens": 29571269.0,
      "step": 12905
    },
    {
      "entropy": 5.361053371429444,
      "epoch": 1.2401536983669548,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048534459553283026,
      "loss": 5.1323,
      "mean_token_accuracy": 0.2079157918691635,
      "num_tokens": 29584024.0,
      "step": 12910
    },
    {
      "entropy": 5.210634183883667,
      "epoch": 1.2406340057636887,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004853324222996449,
      "loss": 4.9711,
      "mean_token_accuracy": 0.21665873229503632,
      "num_tokens": 29595679.0,
      "step": 12915
    },
    {
      "entropy": 5.290135622024536,
      "epoch": 1.2411143131604228,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048532024418316525,
      "loss": 5.0937,
      "mean_token_accuracy": 0.205694717168808,
      "num_tokens": 29607316.0,
      "step": 12920
    },
    {
      "entropy": 5.281459093093872,
      "epoch": 1.2415946205571566,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004853080611836741,
      "loss": 5.1019,
      "mean_token_accuracy": 0.20694130808115005,
      "num_tokens": 29618929.0,
      "step": 12925
    },
    {
      "entropy": 5.3690094470977785,
      "epoch": 1.2420749279538905,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048529587330145427,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19915160536766052,
      "num_tokens": 29630253.0,
      "step": 12930
    },
    {
      "entropy": 5.306268548965454,
      "epoch": 1.2425552353506244,
      "grad_norm": 1.6875,
      "learning_rate": 0.00048528368053678863,
      "loss": 5.1215,
      "mean_token_accuracy": 0.20170108675956727,
      "num_tokens": 29641676.0,
      "step": 12935
    },
    {
      "entropy": 5.282699298858643,
      "epoch": 1.2430355427473583,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004852714828899604,
      "loss": 5.1419,
      "mean_token_accuracy": 0.20086293816566467,
      "num_tokens": 29651942.0,
      "step": 12940
    },
    {
      "entropy": 5.3701231479644775,
      "epoch": 1.2435158501440922,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048525928036125264,
      "loss": 5.1956,
      "mean_token_accuracy": 0.196039317548275,
      "num_tokens": 29663094.0,
      "step": 12945
    },
    {
      "entropy": 5.314688444137573,
      "epoch": 1.243996157540826,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048524707295094884,
      "loss": 5.1426,
      "mean_token_accuracy": 0.20486510992050172,
      "num_tokens": 29675412.0,
      "step": 12950
    },
    {
      "entropy": 5.2434163093566895,
      "epoch": 1.2444764649375601,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004852348606593322,
      "loss": 5.0948,
      "mean_token_accuracy": 0.20069140195846558,
      "num_tokens": 29686684.0,
      "step": 12955
    },
    {
      "entropy": 5.297210311889648,
      "epoch": 1.244956772334294,
      "grad_norm": 1.5,
      "learning_rate": 0.00048522264348668646,
      "loss": 5.0645,
      "mean_token_accuracy": 0.2088254988193512,
      "num_tokens": 29697387.0,
      "step": 12960
    },
    {
      "entropy": 5.31263484954834,
      "epoch": 1.245437079731028,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004852104214332951,
      "loss": 5.017,
      "mean_token_accuracy": 0.21754256784915924,
      "num_tokens": 29708677.0,
      "step": 12965
    },
    {
      "entropy": 5.280858087539673,
      "epoch": 1.2459173871277618,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048519819449944205,
      "loss": 5.2034,
      "mean_token_accuracy": 0.19616346508264543,
      "num_tokens": 29720430.0,
      "step": 12970
    },
    {
      "entropy": 5.274242162704468,
      "epoch": 1.2463976945244957,
      "grad_norm": 1.40625,
      "learning_rate": 0.000485185962685411,
      "loss": 5.0444,
      "mean_token_accuracy": 0.21305937618017196,
      "num_tokens": 29731370.0,
      "step": 12975
    },
    {
      "entropy": 5.419309139251709,
      "epoch": 1.2468780019212296,
      "grad_norm": 1.25,
      "learning_rate": 0.000485173725991486,
      "loss": 5.1597,
      "mean_token_accuracy": 0.1919698014855385,
      "num_tokens": 29743399.0,
      "step": 12980
    },
    {
      "entropy": 5.3649486064910885,
      "epoch": 1.2473583093179634,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048516148441795124,
      "loss": 5.1378,
      "mean_token_accuracy": 0.19681939482688904,
      "num_tokens": 29755693.0,
      "step": 12985
    },
    {
      "entropy": 5.299527168273926,
      "epoch": 1.2478386167146973,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004851492379650908,
      "loss": 5.1751,
      "mean_token_accuracy": 0.1986872985959053,
      "num_tokens": 29766576.0,
      "step": 12990
    },
    {
      "entropy": 5.263984537124633,
      "epoch": 1.2483189241114312,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004851369866331891,
      "loss": 5.088,
      "mean_token_accuracy": 0.19691413193941115,
      "num_tokens": 29779255.0,
      "step": 12995
    },
    {
      "entropy": 5.2304089069366455,
      "epoch": 1.2487992315081653,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004851247304225306,
      "loss": 5.0496,
      "mean_token_accuracy": 0.21078928411006928,
      "num_tokens": 29790172.0,
      "step": 13000
    },
    {
      "entropy": 5.302167224884033,
      "epoch": 1.2492795389048992,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004851124693333997,
      "loss": 5.1463,
      "mean_token_accuracy": 0.19496386349201203,
      "num_tokens": 29801206.0,
      "step": 13005
    },
    {
      "entropy": 5.433167219161987,
      "epoch": 1.249759846301633,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004851002033660812,
      "loss": 5.1855,
      "mean_token_accuracy": 0.19225510358810424,
      "num_tokens": 29813917.0,
      "step": 13010
    },
    {
      "entropy": 5.264387321472168,
      "epoch": 1.250240153698367,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048508793252085994,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20590272098779677,
      "num_tokens": 29826036.0,
      "step": 13015
    },
    {
      "entropy": 5.230810213088989,
      "epoch": 1.2507204610951008,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004850756567980206,
      "loss": 4.983,
      "mean_token_accuracy": 0.21147585064172744,
      "num_tokens": 29837901.0,
      "step": 13020
    },
    {
      "entropy": 5.227533197402954,
      "epoch": 1.2512007684918347,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048506337619784836,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20864436626434327,
      "num_tokens": 29850074.0,
      "step": 13025
    },
    {
      "entropy": 5.303568696975708,
      "epoch": 1.2516810758885688,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004850510907206283,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20069130510091782,
      "num_tokens": 29862447.0,
      "step": 13030
    },
    {
      "entropy": 5.295779132843018,
      "epoch": 1.2521613832853027,
      "grad_norm": 1.875,
      "learning_rate": 0.00048503880036664555,
      "loss": 5.115,
      "mean_token_accuracy": 0.20184466242790222,
      "num_tokens": 29873483.0,
      "step": 13035
    },
    {
      "entropy": 5.255629158020019,
      "epoch": 1.2526416906820366,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004850265051361857,
      "loss": 5.0532,
      "mean_token_accuracy": 0.20545682013034822,
      "num_tokens": 29884807.0,
      "step": 13040
    },
    {
      "entropy": 5.299468946456909,
      "epoch": 1.2531219980787704,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004850142050295339,
      "loss": 5.0797,
      "mean_token_accuracy": 0.19742083102464675,
      "num_tokens": 29895994.0,
      "step": 13045
    },
    {
      "entropy": 5.319211483001709,
      "epoch": 1.2536023054755043,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048500190004697595,
      "loss": 5.1466,
      "mean_token_accuracy": 0.20169174373149873,
      "num_tokens": 29907534.0,
      "step": 13050
    },
    {
      "entropy": 5.323121118545532,
      "epoch": 1.2540826128722382,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004849895901887974,
      "loss": 5.1121,
      "mean_token_accuracy": 0.20402978211641312,
      "num_tokens": 29919747.0,
      "step": 13055
    },
    {
      "entropy": 5.287778902053833,
      "epoch": 1.254562920268972,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004849772754552842,
      "loss": 5.1034,
      "mean_token_accuracy": 0.2085968166589737,
      "num_tokens": 29930254.0,
      "step": 13060
    },
    {
      "entropy": 5.401095724105835,
      "epoch": 1.255043227665706,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048496495584672214,
      "loss": 5.2284,
      "mean_token_accuracy": 0.19882743656635285,
      "num_tokens": 29941232.0,
      "step": 13065
    },
    {
      "entropy": 5.33468074798584,
      "epoch": 1.2555235350624399,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048495263136339725,
      "loss": 5.069,
      "mean_token_accuracy": 0.21000776141881944,
      "num_tokens": 29952452.0,
      "step": 13070
    },
    {
      "entropy": 5.313631343841553,
      "epoch": 1.2560038424591737,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004849403020055956,
      "loss": 5.1153,
      "mean_token_accuracy": 0.20833889544010162,
      "num_tokens": 29963583.0,
      "step": 13075
    },
    {
      "entropy": 5.262060928344726,
      "epoch": 1.2564841498559078,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048492796777360373,
      "loss": 5.0948,
      "mean_token_accuracy": 0.20539553910493852,
      "num_tokens": 29975186.0,
      "step": 13080
    },
    {
      "entropy": 5.241076183319092,
      "epoch": 1.2569644572526417,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048491562866770767,
      "loss": 5.067,
      "mean_token_accuracy": 0.20988473892211915,
      "num_tokens": 29987474.0,
      "step": 13085
    },
    {
      "entropy": 5.241166305541992,
      "epoch": 1.2574447646493756,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048490328468819404,
      "loss": 5.0622,
      "mean_token_accuracy": 0.21254001408815384,
      "num_tokens": 29999254.0,
      "step": 13090
    },
    {
      "entropy": 5.312515211105347,
      "epoch": 1.2579250720461095,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048489093583534945,
      "loss": 5.1265,
      "mean_token_accuracy": 0.19973569512367248,
      "num_tokens": 30009901.0,
      "step": 13095
    },
    {
      "entropy": 5.343769598007202,
      "epoch": 1.2584053794428434,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004848785821094606,
      "loss": 5.2148,
      "mean_token_accuracy": 0.20085404217243194,
      "num_tokens": 30022424.0,
      "step": 13100
    },
    {
      "entropy": 5.29136061668396,
      "epoch": 1.2588856868395775,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004848662235108142,
      "loss": 5.1387,
      "mean_token_accuracy": 0.20845301896333696,
      "num_tokens": 30034209.0,
      "step": 13105
    },
    {
      "entropy": 5.2596007823944095,
      "epoch": 1.2593659942363113,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004848538600396973,
      "loss": 5.0475,
      "mean_token_accuracy": 0.2084960699081421,
      "num_tokens": 30044885.0,
      "step": 13110
    },
    {
      "entropy": 5.378102350234985,
      "epoch": 1.2598463016330452,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048484149169639694,
      "loss": 5.1306,
      "mean_token_accuracy": 0.2026064068078995,
      "num_tokens": 30056592.0,
      "step": 13115
    },
    {
      "entropy": 5.426629209518433,
      "epoch": 1.260326609029779,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004848291184812003,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19223733395338058,
      "num_tokens": 30068423.0,
      "step": 13120
    },
    {
      "entropy": 5.325303363800049,
      "epoch": 1.260806916426513,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004848167403943945,
      "loss": 5.1358,
      "mean_token_accuracy": 0.20521422922611238,
      "num_tokens": 30079132.0,
      "step": 13125
    },
    {
      "entropy": 5.220373106002808,
      "epoch": 1.2612872238232469,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048480435743626703,
      "loss": 5.0255,
      "mean_token_accuracy": 0.2128068432211876,
      "num_tokens": 30090335.0,
      "step": 13130
    },
    {
      "entropy": 5.313543272018433,
      "epoch": 1.2617675312199808,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004847919696071054,
      "loss": 5.1355,
      "mean_token_accuracy": 0.20210581868886948,
      "num_tokens": 30101781.0,
      "step": 13135
    },
    {
      "entropy": 5.294147491455078,
      "epoch": 1.2622478386167146,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048477957690719716,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21108312159776688,
      "num_tokens": 30112486.0,
      "step": 13140
    },
    {
      "entropy": 5.299340629577637,
      "epoch": 1.2627281460134485,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004847671793368301,
      "loss": 5.1221,
      "mean_token_accuracy": 0.2062865138053894,
      "num_tokens": 30123664.0,
      "step": 13145
    },
    {
      "entropy": 5.34754433631897,
      "epoch": 1.2632084534101824,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000484754776896292,
      "loss": 5.1742,
      "mean_token_accuracy": 0.1988801673054695,
      "num_tokens": 30134139.0,
      "step": 13150
    },
    {
      "entropy": 5.282856464385986,
      "epoch": 1.2636887608069165,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004847423695858708,
      "loss": 5.1323,
      "mean_token_accuracy": 0.19887967705726622,
      "num_tokens": 30146127.0,
      "step": 13155
    },
    {
      "entropy": 5.381552743911743,
      "epoch": 1.2641690682036504,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048472995740585456,
      "loss": 5.1446,
      "mean_token_accuracy": 0.19968159794807433,
      "num_tokens": 30158177.0,
      "step": 13160
    },
    {
      "entropy": 5.368123483657837,
      "epoch": 1.2646493756003843,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004847175403565316,
      "loss": 5.1213,
      "mean_token_accuracy": 0.2007624939084053,
      "num_tokens": 30168930.0,
      "step": 13165
    },
    {
      "entropy": 5.34933500289917,
      "epoch": 1.2651296829971181,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048470511843818996,
      "loss": 5.1539,
      "mean_token_accuracy": 0.1997435674071312,
      "num_tokens": 30181206.0,
      "step": 13170
    },
    {
      "entropy": 5.339201259613037,
      "epoch": 1.265609990393852,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004846926916511182,
      "loss": 5.1547,
      "mean_token_accuracy": 0.2086367353796959,
      "num_tokens": 30193093.0,
      "step": 13175
    },
    {
      "entropy": 5.295947170257568,
      "epoch": 1.266090297790586,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004846802599956048,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21331311613321305,
      "num_tokens": 30204864.0,
      "step": 13180
    },
    {
      "entropy": 5.340208292007446,
      "epoch": 1.26657060518732,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048466782347193847,
      "loss": 5.1614,
      "mean_token_accuracy": 0.1968665227293968,
      "num_tokens": 30216312.0,
      "step": 13185
    },
    {
      "entropy": 5.345974683761597,
      "epoch": 1.267050912584054,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048465538208040775,
      "loss": 5.1902,
      "mean_token_accuracy": 0.20015709698200226,
      "num_tokens": 30227824.0,
      "step": 13190
    },
    {
      "entropy": 5.367926836013794,
      "epoch": 1.2675312199807878,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048464293582130166,
      "loss": 5.1747,
      "mean_token_accuracy": 0.1965886875987053,
      "num_tokens": 30239065.0,
      "step": 13195
    },
    {
      "entropy": 5.293788576126099,
      "epoch": 1.2680115273775217,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004846304846949091,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20374094247817992,
      "num_tokens": 30250684.0,
      "step": 13200
    },
    {
      "entropy": 5.314840364456177,
      "epoch": 1.2684918347742555,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048461802870151916,
      "loss": 5.1248,
      "mean_token_accuracy": 0.2022750496864319,
      "num_tokens": 30261376.0,
      "step": 13205
    },
    {
      "entropy": 5.309084558486939,
      "epoch": 1.2689721421709894,
      "grad_norm": 1.5,
      "learning_rate": 0.00048460556784142106,
      "loss": 5.0978,
      "mean_token_accuracy": 0.20967828780412673,
      "num_tokens": 30272441.0,
      "step": 13210
    },
    {
      "entropy": 5.249032831192016,
      "epoch": 1.2694524495677233,
      "grad_norm": 1.5,
      "learning_rate": 0.00048459310211490406,
      "loss": 5.0995,
      "mean_token_accuracy": 0.20624102354049684,
      "num_tokens": 30284446.0,
      "step": 13215
    },
    {
      "entropy": 5.361384534835816,
      "epoch": 1.2699327569644572,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004845806315222576,
      "loss": 5.2119,
      "mean_token_accuracy": 0.20122386515140533,
      "num_tokens": 30295666.0,
      "step": 13220
    },
    {
      "entropy": 5.419999027252198,
      "epoch": 1.270413064361191,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004845681560637711,
      "loss": 5.1764,
      "mean_token_accuracy": 0.20234466046094896,
      "num_tokens": 30307663.0,
      "step": 13225
    },
    {
      "entropy": 5.365679693222046,
      "epoch": 1.270893371757925,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004845556757397344,
      "loss": 5.137,
      "mean_token_accuracy": 0.20308271795511246,
      "num_tokens": 30318914.0,
      "step": 13230
    },
    {
      "entropy": 5.283757257461548,
      "epoch": 1.271373679154659,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004845431905504372,
      "loss": 5.1923,
      "mean_token_accuracy": 0.19223287254571914,
      "num_tokens": 30331499.0,
      "step": 13235
    },
    {
      "entropy": 5.405780696868897,
      "epoch": 1.271853986551393,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048453070049616926,
      "loss": 5.1388,
      "mean_token_accuracy": 0.20489322543144226,
      "num_tokens": 30342186.0,
      "step": 13240
    },
    {
      "entropy": 5.3887745380401615,
      "epoch": 1.2723342939481268,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048451820557722064,
      "loss": 5.1412,
      "mean_token_accuracy": 0.19996339678764344,
      "num_tokens": 30353423.0,
      "step": 13245
    },
    {
      "entropy": 5.237060689926148,
      "epoch": 1.2728146013448607,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004845057057938815,
      "loss": 5.0777,
      "mean_token_accuracy": 0.2032301679253578,
      "num_tokens": 30365008.0,
      "step": 13250
    },
    {
      "entropy": 5.2490966796875,
      "epoch": 1.2732949087415946,
      "grad_norm": 1.84375,
      "learning_rate": 0.00048449320114644185,
      "loss": 5.0059,
      "mean_token_accuracy": 0.2147151783108711,
      "num_tokens": 30376589.0,
      "step": 13255
    },
    {
      "entropy": 5.3086690425872805,
      "epoch": 1.2737752161383284,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004844806916351922,
      "loss": 5.099,
      "mean_token_accuracy": 0.2031288132071495,
      "num_tokens": 30388228.0,
      "step": 13260
    },
    {
      "entropy": 5.281635570526123,
      "epoch": 1.2742555235350626,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004844681772604229,
      "loss": 5.1659,
      "mean_token_accuracy": 0.20615471303462982,
      "num_tokens": 30400225.0,
      "step": 13265
    },
    {
      "entropy": 5.338594007492065,
      "epoch": 1.2747358309317964,
      "grad_norm": 3.03125,
      "learning_rate": 0.00048445565802242454,
      "loss": 5.1099,
      "mean_token_accuracy": 0.19828218519687651,
      "num_tokens": 30411913.0,
      "step": 13270
    },
    {
      "entropy": 5.312411785125732,
      "epoch": 1.2752161383285303,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004844431339214878,
      "loss": 5.104,
      "mean_token_accuracy": 0.20377818793058394,
      "num_tokens": 30423891.0,
      "step": 13275
    },
    {
      "entropy": 5.379608011245727,
      "epoch": 1.2756964457252642,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004844306049579034,
      "loss": 5.1196,
      "mean_token_accuracy": 0.20439768582582474,
      "num_tokens": 30434144.0,
      "step": 13280
    },
    {
      "entropy": 5.282620811462403,
      "epoch": 1.276176753121998,
      "grad_norm": 1.5,
      "learning_rate": 0.00048441807113196216,
      "loss": 5.088,
      "mean_token_accuracy": 0.20222626775503158,
      "num_tokens": 30446628.0,
      "step": 13285
    },
    {
      "entropy": 5.372618246078491,
      "epoch": 1.276657060518732,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048440553244395517,
      "loss": 5.0814,
      "mean_token_accuracy": 0.20335557013750077,
      "num_tokens": 30456885.0,
      "step": 13290
    },
    {
      "entropy": 5.378661489486694,
      "epoch": 1.2771373679154658,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048439298889417357,
      "loss": 5.209,
      "mean_token_accuracy": 0.19687862396240235,
      "num_tokens": 30468642.0,
      "step": 13295
    },
    {
      "entropy": 5.3126599311828615,
      "epoch": 1.2776176753121997,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048438044048290847,
      "loss": 5.108,
      "mean_token_accuracy": 0.1910550221800804,
      "num_tokens": 30480393.0,
      "step": 13300
    },
    {
      "entropy": 5.307544469833374,
      "epoch": 1.2780979827089336,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048436788721045135,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20550643503665925,
      "num_tokens": 30491918.0,
      "step": 13305
    },
    {
      "entropy": 5.32834300994873,
      "epoch": 1.2785782901056677,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004843553290770935,
      "loss": 5.0321,
      "mean_token_accuracy": 0.21037620604038237,
      "num_tokens": 30502250.0,
      "step": 13310
    },
    {
      "entropy": 5.267922449111938,
      "epoch": 1.2790585975024016,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004843427660831266,
      "loss": 5.0694,
      "mean_token_accuracy": 0.20209538340568542,
      "num_tokens": 30514499.0,
      "step": 13315
    },
    {
      "entropy": 5.232398891448975,
      "epoch": 1.2795389048991355,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048433019822884235,
      "loss": 5.0976,
      "mean_token_accuracy": 0.20324259549379348,
      "num_tokens": 30525133.0,
      "step": 13320
    },
    {
      "entropy": 5.276057958602905,
      "epoch": 1.2800192122958693,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004843176255145325,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21079215109348298,
      "num_tokens": 30535778.0,
      "step": 13325
    },
    {
      "entropy": 5.291419506072998,
      "epoch": 1.2804995196926032,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004843050479404888,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21634441167116164,
      "num_tokens": 30547711.0,
      "step": 13330
    },
    {
      "entropy": 5.198636817932129,
      "epoch": 1.280979827089337,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048429246550700343,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20685922652482985,
      "num_tokens": 30558719.0,
      "step": 13335
    },
    {
      "entropy": 5.227184772491455,
      "epoch": 1.2814601344860712,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004842798782143686,
      "loss": 5.1097,
      "mean_token_accuracy": 0.20437829196453094,
      "num_tokens": 30570173.0,
      "step": 13340
    },
    {
      "entropy": 5.322902059555053,
      "epoch": 1.281940441882805,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048426728606287627,
      "loss": 5.1132,
      "mean_token_accuracy": 0.2077580064535141,
      "num_tokens": 30580438.0,
      "step": 13345
    },
    {
      "entropy": 5.32567777633667,
      "epoch": 1.282420749279539,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004842546890528191,
      "loss": 5.0794,
      "mean_token_accuracy": 0.2023497551679611,
      "num_tokens": 30592191.0,
      "step": 13350
    },
    {
      "entropy": 5.30286021232605,
      "epoch": 1.2829010566762729,
      "grad_norm": 1.375,
      "learning_rate": 0.0004842420871844893,
      "loss": 5.1569,
      "mean_token_accuracy": 0.20408184230327606,
      "num_tokens": 30604041.0,
      "step": 13355
    },
    {
      "entropy": 5.23775577545166,
      "epoch": 1.2833813640730067,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004842294804581796,
      "loss": 5.0634,
      "mean_token_accuracy": 0.20893789529800416,
      "num_tokens": 30616087.0,
      "step": 13360
    },
    {
      "entropy": 5.267513370513916,
      "epoch": 1.2838616714697406,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048421686887418266,
      "loss": 5.0726,
      "mean_token_accuracy": 0.21052325367927552,
      "num_tokens": 30627150.0,
      "step": 13365
    },
    {
      "entropy": 5.2891679286956785,
      "epoch": 1.2843419788664745,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004842042524327912,
      "loss": 5.1014,
      "mean_token_accuracy": 0.2045335128903389,
      "num_tokens": 30638627.0,
      "step": 13370
    },
    {
      "entropy": 5.333196210861206,
      "epoch": 1.2848222862632084,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004841916311342983,
      "loss": 5.1564,
      "mean_token_accuracy": 0.19884990006685258,
      "num_tokens": 30649257.0,
      "step": 13375
    },
    {
      "entropy": 5.293218564987183,
      "epoch": 1.2853025936599423,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004841790049789969,
      "loss": 5.1295,
      "mean_token_accuracy": 0.2091818407177925,
      "num_tokens": 30659765.0,
      "step": 13380
    },
    {
      "entropy": 5.291576242446899,
      "epoch": 1.2857829010566761,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00048416637396718004,
      "loss": 5.0979,
      "mean_token_accuracy": 0.19830523133277894,
      "num_tokens": 30672921.0,
      "step": 13385
    },
    {
      "entropy": 5.36685266494751,
      "epoch": 1.2862632084534102,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004841537380991411,
      "loss": 5.1186,
      "mean_token_accuracy": 0.2038745790719986,
      "num_tokens": 30684994.0,
      "step": 13390
    },
    {
      "entropy": 5.347366428375244,
      "epoch": 1.2867435158501441,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048414109737517346,
      "loss": 5.1379,
      "mean_token_accuracy": 0.20500135868787767,
      "num_tokens": 30698094.0,
      "step": 13395
    },
    {
      "entropy": 5.279841899871826,
      "epoch": 1.287223823246878,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004841284517955706,
      "loss": 5.1442,
      "mean_token_accuracy": 0.19993296414613723,
      "num_tokens": 30710716.0,
      "step": 13400
    },
    {
      "entropy": 5.284693861007691,
      "epoch": 1.2877041306436119,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000484115801360626,
      "loss": 5.1218,
      "mean_token_accuracy": 0.2087017074227333,
      "num_tokens": 30721910.0,
      "step": 13405
    },
    {
      "entropy": 5.249923467636108,
      "epoch": 1.2881844380403458,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004841031460706335,
      "loss": 5.0911,
      "mean_token_accuracy": 0.2072000578045845,
      "num_tokens": 30734445.0,
      "step": 13410
    },
    {
      "entropy": 5.325583600997925,
      "epoch": 1.2886647454370797,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048409048592588683,
      "loss": 5.0382,
      "mean_token_accuracy": 0.2132545068860054,
      "num_tokens": 30745131.0,
      "step": 13415
    },
    {
      "entropy": 5.384016513824463,
      "epoch": 1.2891450528338138,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004840778209266799,
      "loss": 5.0865,
      "mean_token_accuracy": 0.2051120325922966,
      "num_tokens": 30757714.0,
      "step": 13420
    },
    {
      "entropy": 5.1506112575531,
      "epoch": 1.2896253602305476,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048406515107330685,
      "loss": 5.0646,
      "mean_token_accuracy": 0.21309217810630798,
      "num_tokens": 30768585.0,
      "step": 13425
    },
    {
      "entropy": 5.318702316284179,
      "epoch": 1.2901056676272815,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048405247636606173,
      "loss": 5.1466,
      "mean_token_accuracy": 0.20803515911102294,
      "num_tokens": 30779866.0,
      "step": 13430
    },
    {
      "entropy": 5.31899619102478,
      "epoch": 1.2905859750240154,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048403979680523894,
      "loss": 5.0793,
      "mean_token_accuracy": 0.2046441376209259,
      "num_tokens": 30791252.0,
      "step": 13435
    },
    {
      "entropy": 5.312030839920044,
      "epoch": 1.2910662824207493,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004840271123911328,
      "loss": 5.119,
      "mean_token_accuracy": 0.2141035109758377,
      "num_tokens": 30802938.0,
      "step": 13440
    },
    {
      "entropy": 5.366402816772461,
      "epoch": 1.2915465898174832,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004840144231240377,
      "loss": 5.28,
      "mean_token_accuracy": 0.18895422220230101,
      "num_tokens": 30816382.0,
      "step": 13445
    },
    {
      "entropy": 5.340416765213012,
      "epoch": 1.292026897214217,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004840017290042484,
      "loss": 5.033,
      "mean_token_accuracy": 0.20998343229293823,
      "num_tokens": 30826402.0,
      "step": 13450
    },
    {
      "entropy": 5.294251537322998,
      "epoch": 1.292507204610951,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048398903003205957,
      "loss": 5.0647,
      "mean_token_accuracy": 0.2090434268116951,
      "num_tokens": 30837862.0,
      "step": 13455
    },
    {
      "entropy": 5.240585517883301,
      "epoch": 1.2929875120076848,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048397632620776604,
      "loss": 5.0216,
      "mean_token_accuracy": 0.210759636759758,
      "num_tokens": 30848351.0,
      "step": 13460
    },
    {
      "entropy": 5.260119867324829,
      "epoch": 1.293467819404419,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00048396361753166276,
      "loss": 5.0942,
      "mean_token_accuracy": 0.2062607154250145,
      "num_tokens": 30858854.0,
      "step": 13465
    },
    {
      "entropy": 5.240932559967041,
      "epoch": 1.2939481268011528,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048395090400404466,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20509129166603088,
      "num_tokens": 30869977.0,
      "step": 13470
    },
    {
      "entropy": 5.3587761402130125,
      "epoch": 1.2944284341978867,
      "grad_norm": 1.78125,
      "learning_rate": 0.00048393818562520715,
      "loss": 5.123,
      "mean_token_accuracy": 0.206313656270504,
      "num_tokens": 30881620.0,
      "step": 13475
    },
    {
      "entropy": 5.383749914169312,
      "epoch": 1.2949087415946205,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048392546239544535,
      "loss": 5.1557,
      "mean_token_accuracy": 0.1980206087231636,
      "num_tokens": 30893575.0,
      "step": 13480
    },
    {
      "entropy": 5.280803871154785,
      "epoch": 1.2953890489913544,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004839127343150547,
      "loss": 5.0952,
      "mean_token_accuracy": 0.20638943761587142,
      "num_tokens": 30904489.0,
      "step": 13485
    },
    {
      "entropy": 5.323291206359864,
      "epoch": 1.2958693563880883,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004839000013843307,
      "loss": 5.1841,
      "mean_token_accuracy": 0.1973687380552292,
      "num_tokens": 30915513.0,
      "step": 13490
    },
    {
      "entropy": 5.37322096824646,
      "epoch": 1.2963496637848224,
      "grad_norm": 2.390625,
      "learning_rate": 0.00048388726360356894,
      "loss": 5.1024,
      "mean_token_accuracy": 0.2034199818968773,
      "num_tokens": 30927633.0,
      "step": 13495
    },
    {
      "entropy": 5.405366039276123,
      "epoch": 1.2968299711815563,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004838745209730653,
      "loss": 5.205,
      "mean_token_accuracy": 0.19195090532302855,
      "num_tokens": 30938937.0,
      "step": 13500
    },
    {
      "entropy": 5.297329902648926,
      "epoch": 1.2973102785782902,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048386177349311535,
      "loss": 5.1069,
      "mean_token_accuracy": 0.20656817704439162,
      "num_tokens": 30950847.0,
      "step": 13505
    },
    {
      "entropy": 5.325516700744629,
      "epoch": 1.297790585975024,
      "grad_norm": 1.375,
      "learning_rate": 0.0004838490211640153,
      "loss": 5.046,
      "mean_token_accuracy": 0.20572479218244552,
      "num_tokens": 30962583.0,
      "step": 13510
    },
    {
      "entropy": 5.307793617248535,
      "epoch": 1.298270893371758,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004838362639860611,
      "loss": 5.0867,
      "mean_token_accuracy": 0.20421871691942214,
      "num_tokens": 30973487.0,
      "step": 13515
    },
    {
      "entropy": 5.2079479694366455,
      "epoch": 1.2987512007684918,
      "grad_norm": 1.515625,
      "learning_rate": 0.000483823501959549,
      "loss": 5.0097,
      "mean_token_accuracy": 0.2107786938548088,
      "num_tokens": 30985176.0,
      "step": 13520
    },
    {
      "entropy": 5.330988979339599,
      "epoch": 1.2992315081652257,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048381073508477527,
      "loss": 5.2025,
      "mean_token_accuracy": 0.19658235907554628,
      "num_tokens": 30997239.0,
      "step": 13525
    },
    {
      "entropy": 5.396906137466431,
      "epoch": 1.2997118155619596,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048379796336203625,
      "loss": 5.176,
      "mean_token_accuracy": 0.19557297825813294,
      "num_tokens": 31009381.0,
      "step": 13530
    },
    {
      "entropy": 5.325507593154907,
      "epoch": 1.3001921229586935,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004837851867916286,
      "loss": 5.0309,
      "mean_token_accuracy": 0.21089338064193724,
      "num_tokens": 31020728.0,
      "step": 13535
    },
    {
      "entropy": 5.29722752571106,
      "epoch": 1.3006724303554273,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004837724053738488,
      "loss": 5.1792,
      "mean_token_accuracy": 0.20072329342365264,
      "num_tokens": 31032933.0,
      "step": 13540
    },
    {
      "entropy": 5.349125719070434,
      "epoch": 1.3011527377521614,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048375961910899373,
      "loss": 5.1384,
      "mean_token_accuracy": 0.2023770272731781,
      "num_tokens": 31044919.0,
      "step": 13545
    },
    {
      "entropy": 5.352047491073608,
      "epoch": 1.3016330451488953,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004837468279973602,
      "loss": 5.1376,
      "mean_token_accuracy": 0.20337988585233688,
      "num_tokens": 31057140.0,
      "step": 13550
    },
    {
      "entropy": 5.235630893707276,
      "epoch": 1.3021133525456292,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004837340320392451,
      "loss": 5.0567,
      "mean_token_accuracy": 0.2050999864935875,
      "num_tokens": 31068035.0,
      "step": 13555
    },
    {
      "entropy": 5.31885256767273,
      "epoch": 1.302593659942363,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048372123123494563,
      "loss": 5.1271,
      "mean_token_accuracy": 0.2017589420080185,
      "num_tokens": 31079819.0,
      "step": 13560
    },
    {
      "entropy": 5.241572570800781,
      "epoch": 1.303073967339097,
      "grad_norm": 2.453125,
      "learning_rate": 0.000483708425584759,
      "loss": 4.9999,
      "mean_token_accuracy": 0.21237864047288896,
      "num_tokens": 31091272.0,
      "step": 13565
    },
    {
      "entropy": 5.281390428543091,
      "epoch": 1.3035542747358309,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004836956150889825,
      "loss": 5.1038,
      "mean_token_accuracy": 0.2028380036354065,
      "num_tokens": 31101513.0,
      "step": 13570
    },
    {
      "entropy": 5.351021337509155,
      "epoch": 1.304034582132565,
      "grad_norm": 3.125,
      "learning_rate": 0.0004836827997479134,
      "loss": 5.1782,
      "mean_token_accuracy": 0.1959371328353882,
      "num_tokens": 31112937.0,
      "step": 13575
    },
    {
      "entropy": 5.303903198242187,
      "epoch": 1.3045148895292988,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004836699795618494,
      "loss": 5.1686,
      "mean_token_accuracy": 0.2014366015791893,
      "num_tokens": 31123445.0,
      "step": 13580
    },
    {
      "entropy": 5.332173919677734,
      "epoch": 1.3049951969260327,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004836571545310881,
      "loss": 5.1148,
      "mean_token_accuracy": 0.20715692341327668,
      "num_tokens": 31135016.0,
      "step": 13585
    },
    {
      "entropy": 5.258981466293335,
      "epoch": 1.3054755043227666,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048364432465592723,
      "loss": 5.0632,
      "mean_token_accuracy": 0.20391278266906737,
      "num_tokens": 31146772.0,
      "step": 13590
    },
    {
      "entropy": 5.206101179122925,
      "epoch": 1.3059558117195005,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004836314899366647,
      "loss": 5.043,
      "mean_token_accuracy": 0.21126957684755326,
      "num_tokens": 31157977.0,
      "step": 13595
    },
    {
      "entropy": 5.292428112030029,
      "epoch": 1.3064361191162344,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048361865037359846,
      "loss": 5.0587,
      "mean_token_accuracy": 0.20924792289733887,
      "num_tokens": 31170693.0,
      "step": 13600
    },
    {
      "entropy": 5.298367023468018,
      "epoch": 1.3069164265129682,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048360580596702664,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20916780531406404,
      "num_tokens": 31181276.0,
      "step": 13605
    },
    {
      "entropy": 5.376260423660279,
      "epoch": 1.3073967339097021,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048359295671724744,
      "loss": 5.2109,
      "mean_token_accuracy": 0.1960655778646469,
      "num_tokens": 31192740.0,
      "step": 13610
    },
    {
      "entropy": 5.336430788040161,
      "epoch": 1.307877041306436,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004835801026245592,
      "loss": 5.0904,
      "mean_token_accuracy": 0.19876156449317933,
      "num_tokens": 31204075.0,
      "step": 13615
    },
    {
      "entropy": 5.340826892852784,
      "epoch": 1.30835734870317,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048356724368926035,
      "loss": 5.1872,
      "mean_token_accuracy": 0.1933182045817375,
      "num_tokens": 31215590.0,
      "step": 13620
    },
    {
      "entropy": 5.354741430282592,
      "epoch": 1.308837656099904,
      "grad_norm": 1.625,
      "learning_rate": 0.00048355437991164937,
      "loss": 5.1771,
      "mean_token_accuracy": 0.19488224536180496,
      "num_tokens": 31228427.0,
      "step": 13625
    },
    {
      "entropy": 5.3623175621032715,
      "epoch": 1.3093179634966379,
      "grad_norm": 1.53125,
      "learning_rate": 0.000483541511292025,
      "loss": 5.1945,
      "mean_token_accuracy": 0.20818801522254943,
      "num_tokens": 31240348.0,
      "step": 13630
    },
    {
      "entropy": 5.371156549453735,
      "epoch": 1.3097982708933718,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048352863783068594,
      "loss": 5.2023,
      "mean_token_accuracy": 0.2043309584259987,
      "num_tokens": 31253028.0,
      "step": 13635
    },
    {
      "entropy": 5.391271924972534,
      "epoch": 1.3102785782901056,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048351575952793117,
      "loss": 5.1725,
      "mean_token_accuracy": 0.20349130779504776,
      "num_tokens": 31265242.0,
      "step": 13640
    },
    {
      "entropy": 5.355654907226563,
      "epoch": 1.3107588856868395,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004835028763840595,
      "loss": 5.1255,
      "mean_token_accuracy": 0.20129722058773042,
      "num_tokens": 31277248.0,
      "step": 13645
    },
    {
      "entropy": 5.304892539978027,
      "epoch": 1.3112391930835736,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004834899883993703,
      "loss": 5.1859,
      "mean_token_accuracy": 0.1924301415681839,
      "num_tokens": 31289631.0,
      "step": 13650
    },
    {
      "entropy": 5.237356948852539,
      "epoch": 1.3117195004803075,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048347709557416263,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21527002453804017,
      "num_tokens": 31301006.0,
      "step": 13655
    },
    {
      "entropy": 5.334367513656616,
      "epoch": 1.3121998078770414,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004834641979087359,
      "loss": 5.1916,
      "mean_token_accuracy": 0.1964273989200592,
      "num_tokens": 31312621.0,
      "step": 13660
    },
    {
      "entropy": 5.392411804199218,
      "epoch": 1.3126801152737753,
      "grad_norm": 2.125,
      "learning_rate": 0.0004834512954033894,
      "loss": 5.1718,
      "mean_token_accuracy": 0.2025168389081955,
      "num_tokens": 31324173.0,
      "step": 13665
    },
    {
      "entropy": 5.32674617767334,
      "epoch": 1.3131604226705091,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048343838805842284,
      "loss": 5.1242,
      "mean_token_accuracy": 0.20186160802841185,
      "num_tokens": 31336872.0,
      "step": 13670
    },
    {
      "entropy": 5.332505750656128,
      "epoch": 1.313640730067243,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048342547587413583,
      "loss": 5.1484,
      "mean_token_accuracy": 0.20343157649040222,
      "num_tokens": 31348760.0,
      "step": 13675
    },
    {
      "entropy": 5.3718681812286375,
      "epoch": 1.314121037463977,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004834125588508282,
      "loss": 5.0944,
      "mean_token_accuracy": 0.2050413891673088,
      "num_tokens": 31360591.0,
      "step": 13680
    },
    {
      "entropy": 5.367049646377564,
      "epoch": 1.3146013448607108,
      "grad_norm": 1.5,
      "learning_rate": 0.0004833996369887998,
      "loss": 5.2932,
      "mean_token_accuracy": 0.19564501345157623,
      "num_tokens": 31372273.0,
      "step": 13685
    },
    {
      "entropy": 5.281128835678101,
      "epoch": 1.3150816522574447,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048338671028835063,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20583412945270538,
      "num_tokens": 31382569.0,
      "step": 13690
    },
    {
      "entropy": 5.376251840591431,
      "epoch": 1.3155619596541785,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048337377874978086,
      "loss": 5.1875,
      "mean_token_accuracy": 0.1941176176071167,
      "num_tokens": 31393913.0,
      "step": 13695
    },
    {
      "entropy": 5.277043437957763,
      "epoch": 1.3160422670509127,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048336084237339067,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20866882503032685,
      "num_tokens": 31405645.0,
      "step": 13700
    },
    {
      "entropy": 5.213544178009033,
      "epoch": 1.3165225744476465,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004833479011594804,
      "loss": 5.0445,
      "mean_token_accuracy": 0.21281868368387222,
      "num_tokens": 31418728.0,
      "step": 13705
    },
    {
      "entropy": 5.271778917312622,
      "epoch": 1.3170028818443804,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048333495510835057,
      "loss": 5.056,
      "mean_token_accuracy": 0.2108875885605812,
      "num_tokens": 31430226.0,
      "step": 13710
    },
    {
      "entropy": 5.305438423156739,
      "epoch": 1.3174831892411143,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048332200422030163,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2087791696190834,
      "num_tokens": 31441751.0,
      "step": 13715
    },
    {
      "entropy": 5.228700971603393,
      "epoch": 1.3179634966378482,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004833090484956345,
      "loss": 5.1351,
      "mean_token_accuracy": 0.20509380847215652,
      "num_tokens": 31455413.0,
      "step": 13720
    },
    {
      "entropy": 5.303403759002686,
      "epoch": 1.318443804034582,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048329608793464966,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21984123289585114,
      "num_tokens": 31466486.0,
      "step": 13725
    },
    {
      "entropy": 5.29347972869873,
      "epoch": 1.3189241114313162,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004832831225376482,
      "loss": 5.1211,
      "mean_token_accuracy": 0.20623160451650618,
      "num_tokens": 31477048.0,
      "step": 13730
    },
    {
      "entropy": 5.307528400421143,
      "epoch": 1.31940441882805,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004832701523049312,
      "loss": 5.1516,
      "mean_token_accuracy": 0.196910260617733,
      "num_tokens": 31488015.0,
      "step": 13735
    },
    {
      "entropy": 5.260188913345337,
      "epoch": 1.319884726224784,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004832571772367997,
      "loss": 5.0553,
      "mean_token_accuracy": 0.2088855490088463,
      "num_tokens": 31499735.0,
      "step": 13740
    },
    {
      "entropy": 5.168818807601928,
      "epoch": 1.3203650336215178,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048324419733355485,
      "loss": 4.904,
      "mean_token_accuracy": 0.21656289100646972,
      "num_tokens": 31511970.0,
      "step": 13745
    },
    {
      "entropy": 5.32693395614624,
      "epoch": 1.3208453410182517,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048323121259549805,
      "loss": 5.141,
      "mean_token_accuracy": 0.20233110189437867,
      "num_tokens": 31522780.0,
      "step": 13750
    },
    {
      "entropy": 5.322263526916504,
      "epoch": 1.3213256484149856,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048321822302293095,
      "loss": 5.1466,
      "mean_token_accuracy": 0.2005403757095337,
      "num_tokens": 31535578.0,
      "step": 13755
    },
    {
      "entropy": 5.328307819366455,
      "epoch": 1.3218059558117194,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004832052286161549,
      "loss": 5.2202,
      "mean_token_accuracy": 0.1936349853873253,
      "num_tokens": 31546623.0,
      "step": 13760
    },
    {
      "entropy": 5.384298086166382,
      "epoch": 1.3222862632084533,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048319222937547176,
      "loss": 5.2774,
      "mean_token_accuracy": 0.19312852919101714,
      "num_tokens": 31559334.0,
      "step": 13765
    },
    {
      "entropy": 5.353963470458984,
      "epoch": 1.3227665706051872,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048317922530118323,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20730090588331224,
      "num_tokens": 31570431.0,
      "step": 13770
    },
    {
      "entropy": 5.217178678512573,
      "epoch": 1.3232468780019213,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004831662163935912,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2101953685283661,
      "num_tokens": 31580314.0,
      "step": 13775
    },
    {
      "entropy": 5.263674926757813,
      "epoch": 1.3237271853986552,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048315320265299784,
      "loss": 5.0464,
      "mean_token_accuracy": 0.20749808400869368,
      "num_tokens": 31592071.0,
      "step": 13780
    },
    {
      "entropy": 5.3606713771820065,
      "epoch": 1.324207492795389,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048314018407970516,
      "loss": 5.2203,
      "mean_token_accuracy": 0.1925863653421402,
      "num_tokens": 31603143.0,
      "step": 13785
    },
    {
      "entropy": 5.36862735748291,
      "epoch": 1.324687800192123,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048312716067401535,
      "loss": 5.2036,
      "mean_token_accuracy": 0.20325835943222045,
      "num_tokens": 31613979.0,
      "step": 13790
    },
    {
      "entropy": 5.275058031082153,
      "epoch": 1.3251681075888568,
      "grad_norm": 1.515625,
      "learning_rate": 0.000483114132436231,
      "loss": 5.0965,
      "mean_token_accuracy": 0.20466650277376175,
      "num_tokens": 31625206.0,
      "step": 13795
    },
    {
      "entropy": 5.369866752624512,
      "epoch": 1.3256484149855907,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004831010993666545,
      "loss": 5.2099,
      "mean_token_accuracy": 0.2073507934808731,
      "num_tokens": 31637709.0,
      "step": 13800
    },
    {
      "entropy": 5.3535229682922365,
      "epoch": 1.3261287223823248,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004830880614655884,
      "loss": 5.138,
      "mean_token_accuracy": 0.20055382698774338,
      "num_tokens": 31648754.0,
      "step": 13805
    },
    {
      "entropy": 5.23485803604126,
      "epoch": 1.3266090297790587,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048307501873333527,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21492660343647002,
      "num_tokens": 31660177.0,
      "step": 13810
    },
    {
      "entropy": 5.270106410980224,
      "epoch": 1.3270893371757926,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004830619711701982,
      "loss": 5.1633,
      "mean_token_accuracy": 0.19692057073116304,
      "num_tokens": 31671467.0,
      "step": 13815
    },
    {
      "entropy": 5.3675154685974125,
      "epoch": 1.3275696445725265,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004830489187764799,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20794325619935988,
      "num_tokens": 31683073.0,
      "step": 13820
    },
    {
      "entropy": 5.244682502746582,
      "epoch": 1.3280499519692603,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004830358615524835,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20769217908382415,
      "num_tokens": 31694661.0,
      "step": 13825
    },
    {
      "entropy": 5.197188472747802,
      "epoch": 1.3285302593659942,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048302279949851215,
      "loss": 5.1018,
      "mean_token_accuracy": 0.205169877409935,
      "num_tokens": 31706904.0,
      "step": 13830
    },
    {
      "entropy": 5.317732524871826,
      "epoch": 1.329010566762728,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048300973261486906,
      "loss": 5.0563,
      "mean_token_accuracy": 0.20845714807510377,
      "num_tokens": 31718087.0,
      "step": 13835
    },
    {
      "entropy": 5.238240909576416,
      "epoch": 1.329490874159462,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004829966609018577,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2071780502796173,
      "num_tokens": 31728960.0,
      "step": 13840
    },
    {
      "entropy": 5.2936629295349125,
      "epoch": 1.3299711815561959,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048298358435978146,
      "loss": 5.142,
      "mean_token_accuracy": 0.20160164982080458,
      "num_tokens": 31741461.0,
      "step": 13845
    },
    {
      "entropy": 5.221812629699707,
      "epoch": 1.3304514889529298,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048297050298894394,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20074637681245805,
      "num_tokens": 31753582.0,
      "step": 13850
    },
    {
      "entropy": 5.246649074554443,
      "epoch": 1.3309317963496639,
      "grad_norm": 1.375,
      "learning_rate": 0.0004829574167896489,
      "loss": 5.0182,
      "mean_token_accuracy": 0.2078266829252243,
      "num_tokens": 31764354.0,
      "step": 13855
    },
    {
      "entropy": 5.248791217803955,
      "epoch": 1.3314121037463977,
      "grad_norm": 1.375,
      "learning_rate": 0.00048294432576220027,
      "loss": 5.0821,
      "mean_token_accuracy": 0.1995893344283104,
      "num_tokens": 31775860.0,
      "step": 13860
    },
    {
      "entropy": 5.325626468658447,
      "epoch": 1.3318924111431316,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004829312299069017,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20675346851348878,
      "num_tokens": 31787606.0,
      "step": 13865
    },
    {
      "entropy": 5.282469892501831,
      "epoch": 1.3323727185398655,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048291812922405755,
      "loss": 5.0823,
      "mean_token_accuracy": 0.2020823448896408,
      "num_tokens": 31798601.0,
      "step": 13870
    },
    {
      "entropy": 5.247089767456055,
      "epoch": 1.3328530259365994,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004829050237139717,
      "loss": 5.04,
      "mean_token_accuracy": 0.20872641503810882,
      "num_tokens": 31809843.0,
      "step": 13875
    },
    {
      "entropy": 5.3048521995544435,
      "epoch": 1.3333333333333333,
      "grad_norm": 1.375,
      "learning_rate": 0.0004828919133769486,
      "loss": 5.1526,
      "mean_token_accuracy": 0.19950384199619292,
      "num_tokens": 31821309.0,
      "step": 13880
    },
    {
      "entropy": 5.338651847839356,
      "epoch": 1.3338136407300674,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004828787982132926,
      "loss": 5.0685,
      "mean_token_accuracy": 0.20668427050113677,
      "num_tokens": 31832215.0,
      "step": 13885
    },
    {
      "entropy": 5.257757711410522,
      "epoch": 1.3342939481268012,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048286567822330815,
      "loss": 5.0376,
      "mean_token_accuracy": 0.21179052591323852,
      "num_tokens": 31843254.0,
      "step": 13890
    },
    {
      "entropy": 5.210302400588989,
      "epoch": 1.3347742555235351,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004828525534072999,
      "loss": 5.0472,
      "mean_token_accuracy": 0.20214553773403168,
      "num_tokens": 31855678.0,
      "step": 13895
    },
    {
      "entropy": 5.343670511245728,
      "epoch": 1.335254562920269,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048283942376557254,
      "loss": 5.1389,
      "mean_token_accuracy": 0.20735029131174088,
      "num_tokens": 31867264.0,
      "step": 13900
    },
    {
      "entropy": 5.217453336715698,
      "epoch": 1.3357348703170029,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00048282628929843097,
      "loss": 5.003,
      "mean_token_accuracy": 0.21520948261022568,
      "num_tokens": 31878908.0,
      "step": 13905
    },
    {
      "entropy": 5.265676069259643,
      "epoch": 1.3362151777137368,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048281315000617996,
      "loss": 5.111,
      "mean_token_accuracy": 0.2045470267534256,
      "num_tokens": 31890299.0,
      "step": 13910
    },
    {
      "entropy": 5.25949125289917,
      "epoch": 1.3366954851104706,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004828000058891248,
      "loss": 5.0546,
      "mean_token_accuracy": 0.20858609527349473,
      "num_tokens": 31901279.0,
      "step": 13915
    },
    {
      "entropy": 5.302527999877929,
      "epoch": 1.3371757925072045,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004827868569475706,
      "loss": 5.0986,
      "mean_token_accuracy": 0.20815426409244536,
      "num_tokens": 31912851.0,
      "step": 13920
    },
    {
      "entropy": 5.2512401103973385,
      "epoch": 1.3376560999039384,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048277370318182243,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20999156981706618,
      "num_tokens": 31923479.0,
      "step": 13925
    },
    {
      "entropy": 5.26459321975708,
      "epoch": 1.3381364073006723,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048276054459218596,
      "loss": 5.1751,
      "mean_token_accuracy": 0.20384859442710876,
      "num_tokens": 31935035.0,
      "step": 13930
    },
    {
      "entropy": 5.408431625366211,
      "epoch": 1.3386167146974064,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048274738117896643,
      "loss": 5.1536,
      "mean_token_accuracy": 0.19912002980709076,
      "num_tokens": 31946140.0,
      "step": 13935
    },
    {
      "entropy": 5.293941688537598,
      "epoch": 1.3390970220941403,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048273421294246966,
      "loss": 5.0742,
      "mean_token_accuracy": 0.2008925125002861,
      "num_tokens": 31957315.0,
      "step": 13940
    },
    {
      "entropy": 5.217030382156372,
      "epoch": 1.3395773294908742,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048272103988300134,
      "loss": 5.0455,
      "mean_token_accuracy": 0.20936957746744156,
      "num_tokens": 31967773.0,
      "step": 13945
    },
    {
      "entropy": 5.363483142852783,
      "epoch": 1.340057636887608,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004827078620008672,
      "loss": 5.2091,
      "mean_token_accuracy": 0.19134783148765563,
      "num_tokens": 31980112.0,
      "step": 13950
    },
    {
      "entropy": 5.29919319152832,
      "epoch": 1.340537944284342,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048269467929637337,
      "loss": 4.9994,
      "mean_token_accuracy": 0.2154555231332779,
      "num_tokens": 31991655.0,
      "step": 13955
    },
    {
      "entropy": 5.235597467422485,
      "epoch": 1.341018251681076,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048268149176982576,
      "loss": 5.048,
      "mean_token_accuracy": 0.21245056241750718,
      "num_tokens": 32003551.0,
      "step": 13960
    },
    {
      "entropy": 5.208331441879272,
      "epoch": 1.34149855907781,
      "grad_norm": 1.375,
      "learning_rate": 0.00048266829942153055,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2112207904458046,
      "num_tokens": 32015096.0,
      "step": 13965
    },
    {
      "entropy": 5.286658811569214,
      "epoch": 1.3419788664745438,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048265510225179413,
      "loss": 5.1295,
      "mean_token_accuracy": 0.20037444680929184,
      "num_tokens": 32026219.0,
      "step": 13970
    },
    {
      "entropy": 5.307773017883301,
      "epoch": 1.3424591738712777,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004826419002609229,
      "loss": 5.1294,
      "mean_token_accuracy": 0.20573511123657226,
      "num_tokens": 32037301.0,
      "step": 13975
    },
    {
      "entropy": 5.2283487796783445,
      "epoch": 1.3429394812680115,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048262869344922326,
      "loss": 5.0143,
      "mean_token_accuracy": 0.2119107574224472,
      "num_tokens": 32048707.0,
      "step": 13980
    },
    {
      "entropy": 5.370496368408203,
      "epoch": 1.3434197886647454,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048261548181700186,
      "loss": 5.1731,
      "mean_token_accuracy": 0.1947040304541588,
      "num_tokens": 32061008.0,
      "step": 13985
    },
    {
      "entropy": 5.251113700866699,
      "epoch": 1.3439000960614793,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004826022653645655,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2203355088829994,
      "num_tokens": 32071360.0,
      "step": 13990
    },
    {
      "entropy": 5.183132743835449,
      "epoch": 1.3443804034582132,
      "grad_norm": 1.265625,
      "learning_rate": 0.000482589044092221,
      "loss": 5.0543,
      "mean_token_accuracy": 0.20888658314943315,
      "num_tokens": 32083441.0,
      "step": 13995
    },
    {
      "entropy": 5.281044244766235,
      "epoch": 1.344860710854947,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048257581800027527,
      "loss": 5.09,
      "mean_token_accuracy": 0.2126864731311798,
      "num_tokens": 32094773.0,
      "step": 14000
    },
    {
      "entropy": 5.260300922393799,
      "epoch": 1.345341018251681,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004825625870890354,
      "loss": 5.1021,
      "mean_token_accuracy": 0.20451863706111909,
      "num_tokens": 32107076.0,
      "step": 14005
    },
    {
      "entropy": 5.416861438751221,
      "epoch": 1.345821325648415,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004825493513588086,
      "loss": 5.2184,
      "mean_token_accuracy": 0.18897178769111633,
      "num_tokens": 32118781.0,
      "step": 14010
    },
    {
      "entropy": 5.248859071731568,
      "epoch": 1.346301633045149,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048253611080990226,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2103603959083557,
      "num_tokens": 32130181.0,
      "step": 14015
    },
    {
      "entropy": 5.308266496658325,
      "epoch": 1.3467819404418828,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004825228654426236,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20534055531024933,
      "num_tokens": 32142576.0,
      "step": 14020
    },
    {
      "entropy": 5.282379865646362,
      "epoch": 1.3472622478386167,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004825096152572803,
      "loss": 5.0761,
      "mean_token_accuracy": 0.21158723086118697,
      "num_tokens": 32152413.0,
      "step": 14025
    },
    {
      "entropy": 5.24849681854248,
      "epoch": 1.3477425552353506,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048249636025417974,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20601485669612885,
      "num_tokens": 32163539.0,
      "step": 14030
    },
    {
      "entropy": 5.3024149417877195,
      "epoch": 1.3482228626320845,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048248310043362997,
      "loss": 5.0587,
      "mean_token_accuracy": 0.20206883698701858,
      "num_tokens": 32174662.0,
      "step": 14035
    },
    {
      "entropy": 5.263035774230957,
      "epoch": 1.3487031700288186,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004824698357959386,
      "loss": 5.0693,
      "mean_token_accuracy": 0.20987719297409058,
      "num_tokens": 32185950.0,
      "step": 14040
    },
    {
      "entropy": 5.3468786716461185,
      "epoch": 1.3491834774255524,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048245656634141385,
      "loss": 5.1286,
      "mean_token_accuracy": 0.20336291640996934,
      "num_tokens": 32198076.0,
      "step": 14045
    },
    {
      "entropy": 5.333752918243408,
      "epoch": 1.3496637848222863,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048244329207036354,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20356308221817015,
      "num_tokens": 32208739.0,
      "step": 14050
    },
    {
      "entropy": 5.331406116485596,
      "epoch": 1.3501440922190202,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048243001298309604,
      "loss": 5.2761,
      "mean_token_accuracy": 0.1918325513601303,
      "num_tokens": 32221494.0,
      "step": 14055
    },
    {
      "entropy": 5.2707366943359375,
      "epoch": 1.350624399615754,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048241672907991954,
      "loss": 5.0438,
      "mean_token_accuracy": 0.20502636432647706,
      "num_tokens": 32232991.0,
      "step": 14060
    },
    {
      "entropy": 5.344660425186158,
      "epoch": 1.351104707012488,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004824034403611424,
      "loss": 5.116,
      "mean_token_accuracy": 0.2053519830107689,
      "num_tokens": 32244022.0,
      "step": 14065
    },
    {
      "entropy": 5.288215303421021,
      "epoch": 1.3515850144092219,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004823901468270733,
      "loss": 5.077,
      "mean_token_accuracy": 0.2022151008248329,
      "num_tokens": 32255779.0,
      "step": 14070
    },
    {
      "entropy": 5.2313391208648685,
      "epoch": 1.3520653218059557,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004823768484780209,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21022157669067382,
      "num_tokens": 32267505.0,
      "step": 14075
    },
    {
      "entropy": 5.255670833587646,
      "epoch": 1.3525456292026896,
      "grad_norm": 1.375,
      "learning_rate": 0.00048236354531429375,
      "loss": 5.1134,
      "mean_token_accuracy": 0.20810193866491317,
      "num_tokens": 32278675.0,
      "step": 14080
    },
    {
      "entropy": 5.257540321350097,
      "epoch": 1.3530259365994235,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004823502373362009,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20960791409015656,
      "num_tokens": 32289372.0,
      "step": 14085
    },
    {
      "entropy": 5.284888172149659,
      "epoch": 1.3535062439961576,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004823369245440512,
      "loss": 5.0833,
      "mean_token_accuracy": 0.20524824559688568,
      "num_tokens": 32301281.0,
      "step": 14090
    },
    {
      "entropy": 5.370954322814941,
      "epoch": 1.3539865513928915,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048232360693815387,
      "loss": 5.1005,
      "mean_token_accuracy": 0.2071798175573349,
      "num_tokens": 32312204.0,
      "step": 14095
    },
    {
      "entropy": 5.310915327072143,
      "epoch": 1.3544668587896254,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048231028451881786,
      "loss": 5.0578,
      "mean_token_accuracy": 0.20960606783628463,
      "num_tokens": 32323607.0,
      "step": 14100
    },
    {
      "entropy": 5.302578401565552,
      "epoch": 1.3549471661863592,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004822969572863527,
      "loss": 5.0909,
      "mean_token_accuracy": 0.21421051621437073,
      "num_tokens": 32334864.0,
      "step": 14105
    },
    {
      "entropy": 5.200412034988403,
      "epoch": 1.3554274735830931,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048228362524106776,
      "loss": 5.1178,
      "mean_token_accuracy": 0.19859998673200607,
      "num_tokens": 32345664.0,
      "step": 14110
    },
    {
      "entropy": 5.286539840698242,
      "epoch": 1.3559077809798272,
      "grad_norm": 1.25,
      "learning_rate": 0.00048227028838327253,
      "loss": 5.0606,
      "mean_token_accuracy": 0.21621714830398558,
      "num_tokens": 32355727.0,
      "step": 14115
    },
    {
      "entropy": 5.295228767395019,
      "epoch": 1.356388088376561,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048225694671327665,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2127278372645378,
      "num_tokens": 32367132.0,
      "step": 14120
    },
    {
      "entropy": 5.2153490543365475,
      "epoch": 1.356868395773295,
      "grad_norm": 1.5,
      "learning_rate": 0.0004822436002313899,
      "loss": 4.9713,
      "mean_token_accuracy": 0.21138667613267897,
      "num_tokens": 32377599.0,
      "step": 14125
    },
    {
      "entropy": 5.301166296005249,
      "epoch": 1.3573487031700289,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004822302489379222,
      "loss": 5.1542,
      "mean_token_accuracy": 0.1979380652308464,
      "num_tokens": 32389659.0,
      "step": 14130
    },
    {
      "entropy": 5.2974278926849365,
      "epoch": 1.3578290105667628,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048221689283318335,
      "loss": 5.0561,
      "mean_token_accuracy": 0.2098655954003334,
      "num_tokens": 32400573.0,
      "step": 14135
    },
    {
      "entropy": 5.324223566055298,
      "epoch": 1.3583093179634966,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004822035319174837,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20765417963266372,
      "num_tokens": 32411009.0,
      "step": 14140
    },
    {
      "entropy": 5.198302268981934,
      "epoch": 1.3587896253602305,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004821901661911332,
      "loss": 5.0965,
      "mean_token_accuracy": 0.21001513600349425,
      "num_tokens": 32422489.0,
      "step": 14145
    },
    {
      "entropy": 5.207678318023682,
      "epoch": 1.3592699327569644,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004821767956544423,
      "loss": 4.9557,
      "mean_token_accuracy": 0.20899262130260468,
      "num_tokens": 32432613.0,
      "step": 14150
    },
    {
      "entropy": 5.270501756668091,
      "epoch": 1.3597502401536983,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004821634203077214,
      "loss": 5.0769,
      "mean_token_accuracy": 0.20832620114088057,
      "num_tokens": 32444556.0,
      "step": 14155
    },
    {
      "entropy": 5.309086799621582,
      "epoch": 1.3602305475504322,
      "grad_norm": 2.484375,
      "learning_rate": 0.000482150040151281,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20718418061733246,
      "num_tokens": 32455143.0,
      "step": 14160
    },
    {
      "entropy": 5.202695655822754,
      "epoch": 1.3607108549471663,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004821366551854318,
      "loss": 5.088,
      "mean_token_accuracy": 0.21282510608434677,
      "num_tokens": 32466966.0,
      "step": 14165
    },
    {
      "entropy": 5.200383615493775,
      "epoch": 1.3611911623439001,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004821232654104845,
      "loss": 5.0002,
      "mean_token_accuracy": 0.21303300112485885,
      "num_tokens": 32479435.0,
      "step": 14170
    },
    {
      "entropy": 5.2383123397827145,
      "epoch": 1.361671469740634,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048210987082675005,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20764392465353013,
      "num_tokens": 32490923.0,
      "step": 14175
    },
    {
      "entropy": 5.351259088516235,
      "epoch": 1.362151777137368,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048209647143453946,
      "loss": 5.1239,
      "mean_token_accuracy": 0.2096369355916977,
      "num_tokens": 32500939.0,
      "step": 14180
    },
    {
      "entropy": 5.215003108978271,
      "epoch": 1.3626320845341018,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048208306723416356,
      "loss": 5.0139,
      "mean_token_accuracy": 0.21083469539880753,
      "num_tokens": 32513260.0,
      "step": 14185
    },
    {
      "entropy": 5.332503747940064,
      "epoch": 1.3631123919308357,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004820696582259339,
      "loss": 5.1705,
      "mean_token_accuracy": 0.2019680455327034,
      "num_tokens": 32524813.0,
      "step": 14190
    },
    {
      "entropy": 5.276840496063232,
      "epoch": 1.3635926993275698,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004820562444101616,
      "loss": 5.0952,
      "mean_token_accuracy": 0.2037371203303337,
      "num_tokens": 32536904.0,
      "step": 14195
    },
    {
      "entropy": 5.207037162780762,
      "epoch": 1.3640730067243036,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004820428257871581,
      "loss": 4.9876,
      "mean_token_accuracy": 0.21034082174301147,
      "num_tokens": 32547419.0,
      "step": 14200
    },
    {
      "entropy": 5.194160509109497,
      "epoch": 1.3645533141210375,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004820294023572351,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22080369144678116,
      "num_tokens": 32558644.0,
      "step": 14205
    },
    {
      "entropy": 5.251490497589112,
      "epoch": 1.3650336215177714,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000482015974120704,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21057114154100418,
      "num_tokens": 32569629.0,
      "step": 14210
    },
    {
      "entropy": 5.220726060867309,
      "epoch": 1.3655139289145053,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048200254107787677,
      "loss": 5.0889,
      "mean_token_accuracy": 0.20386557579040526,
      "num_tokens": 32581850.0,
      "step": 14215
    },
    {
      "entropy": 5.369290161132812,
      "epoch": 1.3659942363112392,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048198910322906516,
      "loss": 5.1881,
      "mean_token_accuracy": 0.1962924614548683,
      "num_tokens": 32592478.0,
      "step": 14220
    },
    {
      "entropy": 5.40494294166565,
      "epoch": 1.366474543707973,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048197566057458125,
      "loss": 5.2635,
      "mean_token_accuracy": 0.19407038986682892,
      "num_tokens": 32603099.0,
      "step": 14225
    },
    {
      "entropy": 5.335611295700073,
      "epoch": 1.366954851104707,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000481962213114737,
      "loss": 5.1533,
      "mean_token_accuracy": 0.19999993294477464,
      "num_tokens": 32613415.0,
      "step": 14230
    },
    {
      "entropy": 5.26595458984375,
      "epoch": 1.3674351585014408,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004819487608498448,
      "loss": 5.1238,
      "mean_token_accuracy": 0.2075204908847809,
      "num_tokens": 32625348.0,
      "step": 14235
    },
    {
      "entropy": 5.289305210113525,
      "epoch": 1.3679154658981747,
      "grad_norm": 1.625,
      "learning_rate": 0.00048193530378021687,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20473739951848985,
      "num_tokens": 32636878.0,
      "step": 14240
    },
    {
      "entropy": 5.320534992218017,
      "epoch": 1.3683957732949088,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048192184190616567,
      "loss": 5.1925,
      "mean_token_accuracy": 0.20035700649023055,
      "num_tokens": 32647808.0,
      "step": 14245
    },
    {
      "entropy": 5.5228376388549805,
      "epoch": 1.3688760806916427,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004819083752280037,
      "loss": 5.2635,
      "mean_token_accuracy": 0.18897124677896499,
      "num_tokens": 32659425.0,
      "step": 14250
    },
    {
      "entropy": 5.342432928085327,
      "epoch": 1.3693563880883766,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048189490374604373,
      "loss": 5.1212,
      "mean_token_accuracy": 0.20096231251955032,
      "num_tokens": 32671797.0,
      "step": 14255
    },
    {
      "entropy": 5.228985786437988,
      "epoch": 1.3698366954851104,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004818814274605983,
      "loss": 5.0685,
      "mean_token_accuracy": 0.2058327615261078,
      "num_tokens": 32683506.0,
      "step": 14260
    },
    {
      "entropy": 5.3756920337677006,
      "epoch": 1.3703170028818443,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004818679463719805,
      "loss": 5.2078,
      "mean_token_accuracy": 0.18928752690553666,
      "num_tokens": 32695047.0,
      "step": 14265
    },
    {
      "entropy": 5.284833669662476,
      "epoch": 1.3707973102785782,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004818544604805033,
      "loss": 5.031,
      "mean_token_accuracy": 0.20956654995679855,
      "num_tokens": 32705433.0,
      "step": 14270
    },
    {
      "entropy": 5.277481365203857,
      "epoch": 1.3712776176753123,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004818409697864798,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19748990386724471,
      "num_tokens": 32717315.0,
      "step": 14275
    },
    {
      "entropy": 5.2386373519897464,
      "epoch": 1.3717579250720462,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048182747429022303,
      "loss": 5.0309,
      "mean_token_accuracy": 0.21385642141103745,
      "num_tokens": 32728817.0,
      "step": 14280
    },
    {
      "entropy": 5.365106439590454,
      "epoch": 1.37223823246878,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004818139739920465,
      "loss": 5.1841,
      "mean_token_accuracy": 0.20418056845664978,
      "num_tokens": 32741030.0,
      "step": 14285
    },
    {
      "entropy": 5.364362668991089,
      "epoch": 1.372718539865514,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004818004688922637,
      "loss": 5.2133,
      "mean_token_accuracy": 0.19529918432235718,
      "num_tokens": 32752212.0,
      "step": 14290
    },
    {
      "entropy": 5.357215785980225,
      "epoch": 1.3731988472622478,
      "grad_norm": 1.984375,
      "learning_rate": 0.000481786958991188,
      "loss": 5.1077,
      "mean_token_accuracy": 0.20349634289741517,
      "num_tokens": 32764581.0,
      "step": 14295
    },
    {
      "entropy": 5.307335090637207,
      "epoch": 1.3736791546589817,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048177344428913316,
      "loss": 5.0922,
      "mean_token_accuracy": 0.20846836268901825,
      "num_tokens": 32775992.0,
      "step": 14300
    },
    {
      "entropy": 5.2759510517120365,
      "epoch": 1.3741594620557156,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048175992478641293,
      "loss": 5.1463,
      "mean_token_accuracy": 0.20535604804754257,
      "num_tokens": 32787164.0,
      "step": 14305
    },
    {
      "entropy": 5.219994354248047,
      "epoch": 1.3746397694524495,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004817464004833412,
      "loss": 5.038,
      "mean_token_accuracy": 0.21287645995616913,
      "num_tokens": 32798190.0,
      "step": 14310
    },
    {
      "entropy": 5.257849788665771,
      "epoch": 1.3751200768491834,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048173287138023204,
      "loss": 5.012,
      "mean_token_accuracy": 0.21279800236225127,
      "num_tokens": 32809559.0,
      "step": 14315
    },
    {
      "entropy": 5.340298414230347,
      "epoch": 1.3756003842459175,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004817193374773993,
      "loss": 5.13,
      "mean_token_accuracy": 0.20079873204231263,
      "num_tokens": 32820799.0,
      "step": 14320
    },
    {
      "entropy": 5.29724235534668,
      "epoch": 1.3760806916426513,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048170579877515753,
      "loss": 5.099,
      "mean_token_accuracy": 0.20207027345895767,
      "num_tokens": 32833971.0,
      "step": 14325
    },
    {
      "entropy": 5.21874508857727,
      "epoch": 1.3765609990393852,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004816922552738209,
      "loss": 5.01,
      "mean_token_accuracy": 0.21199668794870377,
      "num_tokens": 32845237.0,
      "step": 14330
    },
    {
      "entropy": 5.344145774841309,
      "epoch": 1.377041306436119,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048167870697370373,
      "loss": 5.1433,
      "mean_token_accuracy": 0.1966767981648445,
      "num_tokens": 32857960.0,
      "step": 14335
    },
    {
      "entropy": 5.160149145126343,
      "epoch": 1.377521613832853,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004816651538751207,
      "loss": 5.0117,
      "mean_token_accuracy": 0.21387763619422911,
      "num_tokens": 32868801.0,
      "step": 14340
    },
    {
      "entropy": 5.245753955841065,
      "epoch": 1.3780019212295869,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048165159597838664,
      "loss": 5.1071,
      "mean_token_accuracy": 0.202817203104496,
      "num_tokens": 32880544.0,
      "step": 14345
    },
    {
      "entropy": 5.446790170669556,
      "epoch": 1.378482228626321,
      "grad_norm": 1.671875,
      "learning_rate": 0.000481638033283816,
      "loss": 5.1552,
      "mean_token_accuracy": 0.19531744867563247,
      "num_tokens": 32892179.0,
      "step": 14350
    },
    {
      "entropy": 5.436510133743286,
      "epoch": 1.3789625360230549,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048162446579172387,
      "loss": 5.152,
      "mean_token_accuracy": 0.20798097401857377,
      "num_tokens": 32903854.0,
      "step": 14355
    },
    {
      "entropy": 5.302389097213745,
      "epoch": 1.3794428434197887,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004816108935024252,
      "loss": 5.1088,
      "mean_token_accuracy": 0.21108718067407609,
      "num_tokens": 32914731.0,
      "step": 14360
    },
    {
      "entropy": 5.186767482757569,
      "epoch": 1.3799231508165226,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048159731641623507,
      "loss": 4.9998,
      "mean_token_accuracy": 0.2053273379802704,
      "num_tokens": 32925894.0,
      "step": 14365
    },
    {
      "entropy": 5.235160207748413,
      "epoch": 1.3804034582132565,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004815837345334687,
      "loss": 5.0842,
      "mean_token_accuracy": 0.21081809997558593,
      "num_tokens": 32937463.0,
      "step": 14370
    },
    {
      "entropy": 5.439518117904663,
      "epoch": 1.3808837656099904,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004815701478544415,
      "loss": 5.1522,
      "mean_token_accuracy": 0.19911275655031205,
      "num_tokens": 32949635.0,
      "step": 14375
    },
    {
      "entropy": 5.310077667236328,
      "epoch": 1.3813640730067243,
      "grad_norm": 2.46875,
      "learning_rate": 0.00048155655637946876,
      "loss": 5.129,
      "mean_token_accuracy": 0.2048836588859558,
      "num_tokens": 32961043.0,
      "step": 14380
    },
    {
      "entropy": 5.274164342880249,
      "epoch": 1.3818443804034581,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004815429601088662,
      "loss": 5.1693,
      "mean_token_accuracy": 0.20170051604509354,
      "num_tokens": 32972218.0,
      "step": 14385
    },
    {
      "entropy": 5.284476566314697,
      "epoch": 1.382324687800192,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004815293590429494,
      "loss": 5.0045,
      "mean_token_accuracy": 0.21088132709264756,
      "num_tokens": 32983157.0,
      "step": 14390
    },
    {
      "entropy": 5.3252636909484865,
      "epoch": 1.382804995196926,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048151575318203417,
      "loss": 5.161,
      "mean_token_accuracy": 0.2095392644405365,
      "num_tokens": 32994524.0,
      "step": 14395
    },
    {
      "entropy": 5.292595386505127,
      "epoch": 1.38328530259366,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048150214252643637,
      "loss": 5.1635,
      "mean_token_accuracy": 0.19738970547914506,
      "num_tokens": 33006477.0,
      "step": 14400
    },
    {
      "entropy": 5.315766477584839,
      "epoch": 1.3837656099903939,
      "grad_norm": 1.453125,
      "learning_rate": 0.000481488527076472,
      "loss": 5.1075,
      "mean_token_accuracy": 0.20472068190574647,
      "num_tokens": 33018067.0,
      "step": 14405
    },
    {
      "entropy": 5.354547357559204,
      "epoch": 1.3842459173871278,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004814749068324571,
      "loss": 5.1594,
      "mean_token_accuracy": 0.20357175916433334,
      "num_tokens": 33031927.0,
      "step": 14410
    },
    {
      "entropy": 5.339977359771728,
      "epoch": 1.3847262247838616,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048146128179470804,
      "loss": 5.1469,
      "mean_token_accuracy": 0.20386642664670945,
      "num_tokens": 33042516.0,
      "step": 14415
    },
    {
      "entropy": 5.287857532501221,
      "epoch": 1.3852065321805955,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004814476519635411,
      "loss": 5.0272,
      "mean_token_accuracy": 0.20934412479400635,
      "num_tokens": 33054931.0,
      "step": 14420
    },
    {
      "entropy": 5.3451094150543215,
      "epoch": 1.3856868395773294,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048143401733927274,
      "loss": 5.1925,
      "mean_token_accuracy": 0.19194227904081346,
      "num_tokens": 33067333.0,
      "step": 14425
    },
    {
      "entropy": 5.317368030548096,
      "epoch": 1.3861671469740635,
      "grad_norm": 1.953125,
      "learning_rate": 0.00048142037792221943,
      "loss": 5.0693,
      "mean_token_accuracy": 0.20829637050628663,
      "num_tokens": 33078469.0,
      "step": 14430
    },
    {
      "entropy": 5.3112382888793945,
      "epoch": 1.3866474543707974,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004814067337126978,
      "loss": 5.0733,
      "mean_token_accuracy": 0.2082767590880394,
      "num_tokens": 33089993.0,
      "step": 14435
    },
    {
      "entropy": 5.3364949226379395,
      "epoch": 1.3871277617675313,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004813930847110248,
      "loss": 5.1418,
      "mean_token_accuracy": 0.2002052366733551,
      "num_tokens": 33101837.0,
      "step": 14440
    },
    {
      "entropy": 5.3212151527404785,
      "epoch": 1.3876080691642652,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004813794309175173,
      "loss": 5.0923,
      "mean_token_accuracy": 0.2042114958167076,
      "num_tokens": 33113036.0,
      "step": 14445
    },
    {
      "entropy": 5.231478500366211,
      "epoch": 1.388088376560999,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048136577233249205,
      "loss": 5.0457,
      "mean_token_accuracy": 0.20771461725234985,
      "num_tokens": 33124826.0,
      "step": 14450
    },
    {
      "entropy": 5.250487041473389,
      "epoch": 1.388568683957733,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004813521089562666,
      "loss": 4.9153,
      "mean_token_accuracy": 0.21287890821695327,
      "num_tokens": 33135878.0,
      "step": 14455
    },
    {
      "entropy": 5.294827795028686,
      "epoch": 1.3890489913544668,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004813384407891577,
      "loss": 5.1022,
      "mean_token_accuracy": 0.211035817861557,
      "num_tokens": 33146938.0,
      "step": 14460
    },
    {
      "entropy": 5.257104444503784,
      "epoch": 1.3895292987512007,
      "grad_norm": 1.203125,
      "learning_rate": 0.000481324767831483,
      "loss": 5.0665,
      "mean_token_accuracy": 0.2082359731197357,
      "num_tokens": 33159223.0,
      "step": 14465
    },
    {
      "entropy": 5.28316478729248,
      "epoch": 1.3900096061479346,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004813110900835598,
      "loss": 5.2026,
      "mean_token_accuracy": 0.19833183139562607,
      "num_tokens": 33170235.0,
      "step": 14470
    },
    {
      "entropy": 5.2872239112854,
      "epoch": 1.3904899135446687,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004812974075457058,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21450301706790925,
      "num_tokens": 33181193.0,
      "step": 14475
    },
    {
      "entropy": 5.309751796722412,
      "epoch": 1.3909702209414025,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048128372021823845,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20322603583335877,
      "num_tokens": 33192505.0,
      "step": 14480
    },
    {
      "entropy": 5.282972145080566,
      "epoch": 1.3914505283381364,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048127002810147574,
      "loss": 5.0582,
      "mean_token_accuracy": 0.2049427106976509,
      "num_tokens": 33204113.0,
      "step": 14485
    },
    {
      "entropy": 5.427578163146973,
      "epoch": 1.3919308357348703,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004812563311957355,
      "loss": 5.2715,
      "mean_token_accuracy": 0.19544494450092315,
      "num_tokens": 33216175.0,
      "step": 14490
    },
    {
      "entropy": 5.356990623474121,
      "epoch": 1.3924111431316042,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004812426295013356,
      "loss": 5.1055,
      "mean_token_accuracy": 0.2050111100077629,
      "num_tokens": 33226916.0,
      "step": 14495
    },
    {
      "entropy": 5.280545711517334,
      "epoch": 1.392891450528338,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048122892301859433,
      "loss": 5.1328,
      "mean_token_accuracy": 0.20771569460630418,
      "num_tokens": 33238020.0,
      "step": 14500
    },
    {
      "entropy": 5.261943292617798,
      "epoch": 1.3933717579250722,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048121521174782983,
      "loss": 4.9845,
      "mean_token_accuracy": 0.21012151539325713,
      "num_tokens": 33248343.0,
      "step": 14505
    },
    {
      "entropy": 5.401256895065307,
      "epoch": 1.393852065321806,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048120149568936044,
      "loss": 5.1378,
      "mean_token_accuracy": 0.19927033185958862,
      "num_tokens": 33259206.0,
      "step": 14510
    },
    {
      "entropy": 5.310678243637085,
      "epoch": 1.39433237271854,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004811877748435046,
      "loss": 5.1687,
      "mean_token_accuracy": 0.1969733402132988,
      "num_tokens": 33271020.0,
      "step": 14515
    },
    {
      "entropy": 5.241995811462402,
      "epoch": 1.3948126801152738,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004811740492105809,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21393192708492278,
      "num_tokens": 33282603.0,
      "step": 14520
    },
    {
      "entropy": 5.355679559707641,
      "epoch": 1.3952929875120077,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000481160318790908,
      "loss": 5.1212,
      "mean_token_accuracy": 0.2077313631772995,
      "num_tokens": 33292877.0,
      "step": 14525
    },
    {
      "entropy": 5.370552206039429,
      "epoch": 1.3957732949087416,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048114658358480467,
      "loss": 5.3203,
      "mean_token_accuracy": 0.2016620844602585,
      "num_tokens": 33304835.0,
      "step": 14530
    },
    {
      "entropy": 5.34936375617981,
      "epoch": 1.3962536023054755,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048113284359258977,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20345089733600616,
      "num_tokens": 33315028.0,
      "step": 14535
    },
    {
      "entropy": 5.295795726776123,
      "epoch": 1.3967339097022093,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048111909881458234,
      "loss": 5.0632,
      "mean_token_accuracy": 0.20692850053310394,
      "num_tokens": 33325321.0,
      "step": 14540
    },
    {
      "entropy": 5.305032634735108,
      "epoch": 1.3972142170989432,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048110534925110146,
      "loss": 5.1562,
      "mean_token_accuracy": 0.2057032823562622,
      "num_tokens": 33337677.0,
      "step": 14545
    },
    {
      "entropy": 5.25224118232727,
      "epoch": 1.397694524495677,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004810915949024664,
      "loss": 5.0636,
      "mean_token_accuracy": 0.21026079654693602,
      "num_tokens": 33347613.0,
      "step": 14550
    },
    {
      "entropy": 5.305679607391357,
      "epoch": 1.3981748318924112,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004810778357689965,
      "loss": 5.0633,
      "mean_token_accuracy": 0.2077339306473732,
      "num_tokens": 33358912.0,
      "step": 14555
    },
    {
      "entropy": 5.392436599731445,
      "epoch": 1.398655139289145,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048106407185101116,
      "loss": 5.1922,
      "mean_token_accuracy": 0.2014326497912407,
      "num_tokens": 33371635.0,
      "step": 14560
    },
    {
      "entropy": 5.310579013824463,
      "epoch": 1.399135446685879,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048105030314883,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21497265845537186,
      "num_tokens": 33383521.0,
      "step": 14565
    },
    {
      "entropy": 5.2026196956634525,
      "epoch": 1.3996157540826129,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004810365296627725,
      "loss": 5.0754,
      "mean_token_accuracy": 0.20365598499774934,
      "num_tokens": 33395297.0,
      "step": 14570
    },
    {
      "entropy": 5.220655870437622,
      "epoch": 1.4000960614793467,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004810227513931587,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22229982763528824,
      "num_tokens": 33407168.0,
      "step": 14575
    },
    {
      "entropy": 5.2651430606842045,
      "epoch": 1.4005763688760806,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004810089683403084,
      "loss": 4.9913,
      "mean_token_accuracy": 0.20640457719564437,
      "num_tokens": 33418568.0,
      "step": 14580
    },
    {
      "entropy": 5.251408386230469,
      "epoch": 1.4010566762728147,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004809951805045415,
      "loss": 5.102,
      "mean_token_accuracy": 0.20217150449752808,
      "num_tokens": 33428822.0,
      "step": 14585
    },
    {
      "entropy": 5.264792251586914,
      "epoch": 1.4015369836695486,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048098138788617815,
      "loss": 5.058,
      "mean_token_accuracy": 0.21319797784090042,
      "num_tokens": 33441547.0,
      "step": 14590
    },
    {
      "entropy": 5.334709692001343,
      "epoch": 1.4020172910662825,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004809675904855387,
      "loss": 5.0897,
      "mean_token_accuracy": 0.2100147560238838,
      "num_tokens": 33452445.0,
      "step": 14595
    },
    {
      "entropy": 5.213797187805175,
      "epoch": 1.4024975984630164,
      "grad_norm": 2.5,
      "learning_rate": 0.00048095378830294343,
      "loss": 5.0842,
      "mean_token_accuracy": 0.21043241918087005,
      "num_tokens": 33463845.0,
      "step": 14600
    },
    {
      "entropy": 5.291446876525879,
      "epoch": 1.4029779058597502,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048093998133871276,
      "loss": 5.0599,
      "mean_token_accuracy": 0.20891935378313065,
      "num_tokens": 33474950.0,
      "step": 14605
    },
    {
      "entropy": 5.240527963638305,
      "epoch": 1.4034582132564841,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004809261695931671,
      "loss": 4.9958,
      "mean_token_accuracy": 0.21098326295614242,
      "num_tokens": 33486225.0,
      "step": 14610
    },
    {
      "entropy": 5.294539451599121,
      "epoch": 1.403938520653218,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004809123530666273,
      "loss": 5.0874,
      "mean_token_accuracy": 0.20585947334766388,
      "num_tokens": 33498125.0,
      "step": 14615
    },
    {
      "entropy": 5.246880435943604,
      "epoch": 1.4044188280499519,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004808985317594142,
      "loss": 4.9572,
      "mean_token_accuracy": 0.21119745671749116,
      "num_tokens": 33508731.0,
      "step": 14620
    },
    {
      "entropy": 5.233324098587036,
      "epoch": 1.4048991354466858,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048088470567184854,
      "loss": 5.001,
      "mean_token_accuracy": 0.2164063036441803,
      "num_tokens": 33520356.0,
      "step": 14625
    },
    {
      "entropy": 5.294951152801514,
      "epoch": 1.4053794428434199,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048087087480425133,
      "loss": 5.1489,
      "mean_token_accuracy": 0.2062847539782524,
      "num_tokens": 33533179.0,
      "step": 14630
    },
    {
      "entropy": 5.182510948181152,
      "epoch": 1.4058597502401537,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004808570391569437,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21196479350328445,
      "num_tokens": 33544700.0,
      "step": 14635
    },
    {
      "entropy": 5.31766676902771,
      "epoch": 1.4063400576368876,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048084319873024694,
      "loss": 5.2398,
      "mean_token_accuracy": 0.19813834130764008,
      "num_tokens": 33558781.0,
      "step": 14640
    },
    {
      "entropy": 5.292840099334716,
      "epoch": 1.4068203650336215,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004808293535244823,
      "loss": 5.0601,
      "mean_token_accuracy": 0.2082328036427498,
      "num_tokens": 33569508.0,
      "step": 14645
    },
    {
      "entropy": 5.302446460723877,
      "epoch": 1.4073006724303554,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004808155035399712,
      "loss": 5.0373,
      "mean_token_accuracy": 0.20495803505182267,
      "num_tokens": 33580977.0,
      "step": 14650
    },
    {
      "entropy": 5.26583571434021,
      "epoch": 1.4077809798270893,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004808016487770354,
      "loss": 5.1347,
      "mean_token_accuracy": 0.2053412154316902,
      "num_tokens": 33593757.0,
      "step": 14655
    },
    {
      "entropy": 5.363867092132568,
      "epoch": 1.4082612872238234,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048078778923599637,
      "loss": 5.2005,
      "mean_token_accuracy": 0.19826872050762176,
      "num_tokens": 33604903.0,
      "step": 14660
    },
    {
      "entropy": 5.29718713760376,
      "epoch": 1.4087415946205573,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048077392491717593,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21284384727478028,
      "num_tokens": 33616976.0,
      "step": 14665
    },
    {
      "entropy": 5.29964919090271,
      "epoch": 1.4092219020172911,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048076005582089597,
      "loss": 5.082,
      "mean_token_accuracy": 0.21178105473518372,
      "num_tokens": 33627654.0,
      "step": 14670
    },
    {
      "entropy": 5.241257858276367,
      "epoch": 1.409702209414025,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048074618194747845,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21175001859664916,
      "num_tokens": 33638377.0,
      "step": 14675
    },
    {
      "entropy": 5.295359992980957,
      "epoch": 1.410182516810759,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004807323032972456,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20842555314302444,
      "num_tokens": 33650959.0,
      "step": 14680
    },
    {
      "entropy": 5.377836656570435,
      "epoch": 1.4106628242074928,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004807184198705196,
      "loss": 5.1856,
      "mean_token_accuracy": 0.1969420462846756,
      "num_tokens": 33663069.0,
      "step": 14685
    },
    {
      "entropy": 5.271876287460327,
      "epoch": 1.4111431316042267,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004807045316676226,
      "loss": 5.1179,
      "mean_token_accuracy": 0.2064431831240654,
      "num_tokens": 33674868.0,
      "step": 14690
    },
    {
      "entropy": 5.258528041839599,
      "epoch": 1.4116234390009605,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004806906386888773,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20982298105955124,
      "num_tokens": 33685482.0,
      "step": 14695
    },
    {
      "entropy": 5.2641582012176515,
      "epoch": 1.4121037463976944,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048067674093460607,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20705909579992293,
      "num_tokens": 33697518.0,
      "step": 14700
    },
    {
      "entropy": 5.286493110656738,
      "epoch": 1.4125840537944283,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048066283840513175,
      "loss": 5.0697,
      "mean_token_accuracy": 0.20300565659999847,
      "num_tokens": 33708827.0,
      "step": 14705
    },
    {
      "entropy": 5.291349744796753,
      "epoch": 1.4130643611911624,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004806489311007769,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20813338160514833,
      "num_tokens": 33719452.0,
      "step": 14710
    },
    {
      "entropy": 5.311522483825684,
      "epoch": 1.4135446685878963,
      "grad_norm": 1.375,
      "learning_rate": 0.00048063501902186463,
      "loss": 5.1663,
      "mean_token_accuracy": 0.20635982602834702,
      "num_tokens": 33730630.0,
      "step": 14715
    },
    {
      "entropy": 5.230512094497681,
      "epoch": 1.4140249759846302,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048062110216871775,
      "loss": 5.0349,
      "mean_token_accuracy": 0.21062091886997222,
      "num_tokens": 33742696.0,
      "step": 14720
    },
    {
      "entropy": 5.301029682159424,
      "epoch": 1.414505283381364,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048060718054165945,
      "loss": 5.0916,
      "mean_token_accuracy": 0.20818879902362825,
      "num_tokens": 33753818.0,
      "step": 14725
    },
    {
      "entropy": 5.438642358779907,
      "epoch": 1.414985590778098,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000480593254141013,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1957718938589096,
      "num_tokens": 33766847.0,
      "step": 14730
    },
    {
      "entropy": 5.335894536972046,
      "epoch": 1.4154658981748318,
      "grad_norm": 1.5,
      "learning_rate": 0.00048057932296710165,
      "loss": 5.0373,
      "mean_token_accuracy": 0.20812246650457383,
      "num_tokens": 33777070.0,
      "step": 14735
    },
    {
      "entropy": 5.307952928543091,
      "epoch": 1.415946205571566,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004805653870202489,
      "loss": 5.0675,
      "mean_token_accuracy": 0.2092259332537651,
      "num_tokens": 33788826.0,
      "step": 14740
    },
    {
      "entropy": 5.190117406845093,
      "epoch": 1.4164265129682998,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048055144630077825,
      "loss": 5.0388,
      "mean_token_accuracy": 0.21261132508516312,
      "num_tokens": 33801564.0,
      "step": 14745
    },
    {
      "entropy": 5.256279182434082,
      "epoch": 1.4169068203650337,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048053750080901336,
      "loss": 5.0822,
      "mean_token_accuracy": 0.20800948441028594,
      "num_tokens": 33813389.0,
      "step": 14750
    },
    {
      "entropy": 5.264689922332764,
      "epoch": 1.4173871277617676,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048052355054527794,
      "loss": 5.0754,
      "mean_token_accuracy": 0.20501003563404083,
      "num_tokens": 33824585.0,
      "step": 14755
    },
    {
      "entropy": 5.323046588897705,
      "epoch": 1.4178674351585014,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048050959550989606,
      "loss": 5.1577,
      "mean_token_accuracy": 0.2044488474726677,
      "num_tokens": 33837995.0,
      "step": 14760
    },
    {
      "entropy": 5.27868618965149,
      "epoch": 1.4183477425552353,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004804956357031916,
      "loss": 5.0543,
      "mean_token_accuracy": 0.2140570968389511,
      "num_tokens": 33848849.0,
      "step": 14765
    },
    {
      "entropy": 5.295002746582031,
      "epoch": 1.4188280499519692,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048048167112548873,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21740321218967437,
      "num_tokens": 33859711.0,
      "step": 14770
    },
    {
      "entropy": 5.303115367889404,
      "epoch": 1.419308357348703,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048046770177711157,
      "loss": 5.1447,
      "mean_token_accuracy": 0.2007286235690117,
      "num_tokens": 33872613.0,
      "step": 14775
    },
    {
      "entropy": 5.167349433898925,
      "epoch": 1.419788664745437,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004804537276583844,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21260344833135605,
      "num_tokens": 33883759.0,
      "step": 14780
    },
    {
      "entropy": 5.335402393341065,
      "epoch": 1.420268972142171,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004804397487696319,
      "loss": 5.1326,
      "mean_token_accuracy": 0.2079780176281929,
      "num_tokens": 33894448.0,
      "step": 14785
    },
    {
      "entropy": 5.3530323028564455,
      "epoch": 1.420749279538905,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004804257651111783,
      "loss": 5.1715,
      "mean_token_accuracy": 0.20156182050704957,
      "num_tokens": 33906976.0,
      "step": 14790
    },
    {
      "entropy": 5.268605661392212,
      "epoch": 1.4212295869356388,
      "grad_norm": 2.125,
      "learning_rate": 0.00048041177668334853,
      "loss": 5.1083,
      "mean_token_accuracy": 0.2050240606069565,
      "num_tokens": 33919384.0,
      "step": 14795
    },
    {
      "entropy": 5.230211591720581,
      "epoch": 1.4217098943323727,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004803977834864672,
      "loss": 4.9949,
      "mean_token_accuracy": 0.20880757868289948,
      "num_tokens": 33931080.0,
      "step": 14800
    },
    {
      "entropy": 5.198806524276733,
      "epoch": 1.4221902017291066,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048038378552085927,
      "loss": 5.051,
      "mean_token_accuracy": 0.2100389301776886,
      "num_tokens": 33943021.0,
      "step": 14805
    },
    {
      "entropy": 5.325034332275391,
      "epoch": 1.4226705091258405,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048036978278684974,
      "loss": 5.0603,
      "mean_token_accuracy": 0.2079825758934021,
      "num_tokens": 33953609.0,
      "step": 14810
    },
    {
      "entropy": 5.298333024978637,
      "epoch": 1.4231508165225746,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004803557752847636,
      "loss": 5.1366,
      "mean_token_accuracy": 0.2016480028629303,
      "num_tokens": 33966985.0,
      "step": 14815
    },
    {
      "entropy": 5.2601783752441404,
      "epoch": 1.4236311239193085,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048034176301492616,
      "loss": 5.0298,
      "mean_token_accuracy": 0.2099802553653717,
      "num_tokens": 33976536.0,
      "step": 14820
    },
    {
      "entropy": 5.255039405822754,
      "epoch": 1.4241114313160423,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004803277459776628,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20640670508146286,
      "num_tokens": 33988000.0,
      "step": 14825
    },
    {
      "entropy": 5.316941165924073,
      "epoch": 1.4245917387127762,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048031372417329875,
      "loss": 5.1615,
      "mean_token_accuracy": 0.20761444866657258,
      "num_tokens": 33998944.0,
      "step": 14830
    },
    {
      "entropy": 5.32610239982605,
      "epoch": 1.42507204610951,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004802996976021598,
      "loss": 5.1665,
      "mean_token_accuracy": 0.20465503334999086,
      "num_tokens": 34010440.0,
      "step": 14835
    },
    {
      "entropy": 5.299844217300415,
      "epoch": 1.425552353506244,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048028566626457145,
      "loss": 5.1362,
      "mean_token_accuracy": 0.204570235311985,
      "num_tokens": 34023222.0,
      "step": 14840
    },
    {
      "entropy": 5.246737480163574,
      "epoch": 1.4260326609029779,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048027163016085947,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21307268142700195,
      "num_tokens": 34035413.0,
      "step": 14845
    },
    {
      "entropy": 5.236249780654907,
      "epoch": 1.4265129682997117,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048025758929134976,
      "loss": 5.0817,
      "mean_token_accuracy": 0.20999034792184829,
      "num_tokens": 34046700.0,
      "step": 14850
    },
    {
      "entropy": 5.2617433071136475,
      "epoch": 1.4269932756964456,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004802435436563684,
      "loss": 5.1013,
      "mean_token_accuracy": 0.20697735249996185,
      "num_tokens": 34058039.0,
      "step": 14855
    },
    {
      "entropy": 5.276310586929322,
      "epoch": 1.4274735830931795,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048022949325624134,
      "loss": 5.1013,
      "mean_token_accuracy": 0.20503979176282883,
      "num_tokens": 34069602.0,
      "step": 14860
    },
    {
      "entropy": 5.349841928482055,
      "epoch": 1.4279538904899136,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048021543809129483,
      "loss": 5.1198,
      "mean_token_accuracy": 0.2010275110602379,
      "num_tokens": 34080514.0,
      "step": 14865
    },
    {
      "entropy": 5.243289279937744,
      "epoch": 1.4284341978866475,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004802013781618552,
      "loss": 5.028,
      "mean_token_accuracy": 0.2115473821759224,
      "num_tokens": 34091416.0,
      "step": 14870
    },
    {
      "entropy": 5.301174354553223,
      "epoch": 1.4289145052833814,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048018731346824895,
      "loss": 5.0692,
      "mean_token_accuracy": 0.20230100452899932,
      "num_tokens": 34102153.0,
      "step": 14875
    },
    {
      "entropy": 5.326796913146973,
      "epoch": 1.4293948126801153,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004801732440108026,
      "loss": 5.1364,
      "mean_token_accuracy": 0.2078568547964096,
      "num_tokens": 34111917.0,
      "step": 14880
    },
    {
      "entropy": 5.298147296905517,
      "epoch": 1.4298751200768491,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004801591697898427,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20037685930728913,
      "num_tokens": 34124038.0,
      "step": 14885
    },
    {
      "entropy": 5.318940591812134,
      "epoch": 1.430355427473583,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004801450908056961,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20355573743581773,
      "num_tokens": 34134503.0,
      "step": 14890
    },
    {
      "entropy": 5.201473331451416,
      "epoch": 1.4308357348703171,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004801310070586896,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2074144408106804,
      "num_tokens": 34146145.0,
      "step": 14895
    },
    {
      "entropy": 5.2408805847167965,
      "epoch": 1.431316042267051,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004801169185491503,
      "loss": 5.0571,
      "mean_token_accuracy": 0.20632531344890595,
      "num_tokens": 34156672.0,
      "step": 14900
    },
    {
      "entropy": 5.3370726108551025,
      "epoch": 1.4317963496637849,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048010282527740516,
      "loss": 5.2245,
      "mean_token_accuracy": 0.19375379383563995,
      "num_tokens": 34168817.0,
      "step": 14905
    },
    {
      "entropy": 5.275558996200561,
      "epoch": 1.4322766570605188,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048008872724378146,
      "loss": 5.0594,
      "mean_token_accuracy": 0.21218208968639374,
      "num_tokens": 34179804.0,
      "step": 14910
    },
    {
      "entropy": 5.4179401874542235,
      "epoch": 1.4327569644572526,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004800746244486065,
      "loss": 5.2985,
      "mean_token_accuracy": 0.19245485216379166,
      "num_tokens": 34192293.0,
      "step": 14915
    },
    {
      "entropy": 5.315548944473266,
      "epoch": 1.4332372718539865,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004800605168922077,
      "loss": 5.1315,
      "mean_token_accuracy": 0.2052086651325226,
      "num_tokens": 34203331.0,
      "step": 14920
    },
    {
      "entropy": 5.319584560394287,
      "epoch": 1.4337175792507204,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048004640457491267,
      "loss": 5.0552,
      "mean_token_accuracy": 0.20372386276721954,
      "num_tokens": 34213554.0,
      "step": 14925
    },
    {
      "entropy": 5.254944849014282,
      "epoch": 1.4341978866474543,
      "grad_norm": 1.484375,
      "learning_rate": 0.000480032287497049,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20924965888261796,
      "num_tokens": 34225824.0,
      "step": 14930
    },
    {
      "entropy": 5.251681661605835,
      "epoch": 1.4346781940441882,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048001816565894427,
      "loss": 5.1,
      "mean_token_accuracy": 0.2048413172364235,
      "num_tokens": 34237479.0,
      "step": 14935
    },
    {
      "entropy": 5.289897823333741,
      "epoch": 1.435158501440922,
      "grad_norm": 1.125,
      "learning_rate": 0.0004800040390609267,
      "loss": 5.1556,
      "mean_token_accuracy": 0.19974869936704637,
      "num_tokens": 34248706.0,
      "step": 14940
    },
    {
      "entropy": 5.356386137008667,
      "epoch": 1.4356388088376562,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047998990770332396,
      "loss": 5.1711,
      "mean_token_accuracy": 0.20292578488588334,
      "num_tokens": 34258957.0,
      "step": 14945
    },
    {
      "entropy": 5.290422105789185,
      "epoch": 1.43611911623439,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004799757715864643,
      "loss": 5.111,
      "mean_token_accuracy": 0.20319284796714782,
      "num_tokens": 34270635.0,
      "step": 14950
    },
    {
      "entropy": 5.35513596534729,
      "epoch": 1.436599423631124,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004799616307106759,
      "loss": 5.0798,
      "mean_token_accuracy": 0.21123429387807846,
      "num_tokens": 34282491.0,
      "step": 14955
    },
    {
      "entropy": 5.27608585357666,
      "epoch": 1.4370797310278578,
      "grad_norm": 1.25,
      "learning_rate": 0.000479947485076287,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21328052282333373,
      "num_tokens": 34294666.0,
      "step": 14960
    },
    {
      "entropy": 5.205697727203369,
      "epoch": 1.4375600384245917,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047993333468362607,
      "loss": 4.9925,
      "mean_token_accuracy": 0.20909564048051835,
      "num_tokens": 34305065.0,
      "step": 14965
    },
    {
      "entropy": 5.309795618057251,
      "epoch": 1.4380403458213258,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047991917953302173,
      "loss": 5.1376,
      "mean_token_accuracy": 0.20099881887435914,
      "num_tokens": 34315955.0,
      "step": 14970
    },
    {
      "entropy": 5.238296842575073,
      "epoch": 1.4385206532180597,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047990501962480236,
      "loss": 5.0006,
      "mean_token_accuracy": 0.21361134350299835,
      "num_tokens": 34327011.0,
      "step": 14975
    },
    {
      "entropy": 5.234391927719116,
      "epoch": 1.4390009606147935,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000479890854959297,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20990480184555055,
      "num_tokens": 34338011.0,
      "step": 14980
    },
    {
      "entropy": 5.204637479782105,
      "epoch": 1.4394812680115274,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004798766855368344,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21445910334587098,
      "num_tokens": 34348957.0,
      "step": 14985
    },
    {
      "entropy": 5.265935277938842,
      "epoch": 1.4399615754082613,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047986251135774343,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20957347750663757,
      "num_tokens": 34359442.0,
      "step": 14990
    },
    {
      "entropy": 5.265705680847168,
      "epoch": 1.4404418828049952,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004798483324223533,
      "loss": 5.076,
      "mean_token_accuracy": 0.20944440215826035,
      "num_tokens": 34371129.0,
      "step": 14995
    },
    {
      "entropy": 5.251073408126831,
      "epoch": 1.440922190201729,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004798341487309932,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21382581442594528,
      "num_tokens": 34381565.0,
      "step": 15000
    },
    {
      "epoch": 1.440922190201729,
      "eval_entropy": 5.133266845137709,
      "eval_loss": 5.163461685180664,
      "eval_mean_token_accuracy": 0.21123866147944284,
      "eval_num_tokens": 34381565.0,
      "eval_runtime": 26.7947,
      "eval_samples_per_second": 1224.681,
      "eval_steps_per_second": 153.09,
      "step": 15000
    },
    {
      "entropy": 5.305964469909668,
      "epoch": 1.441402497598463,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047981996028399233,
      "loss": 5.0093,
      "mean_token_accuracy": 0.2117380127310753,
      "num_tokens": 34392145.0,
      "step": 15005
    },
    {
      "entropy": 5.2658977031707765,
      "epoch": 1.4418828049951968,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004798057670816802,
      "loss": 5.164,
      "mean_token_accuracy": 0.20149679034948348,
      "num_tokens": 34403193.0,
      "step": 15010
    },
    {
      "entropy": 5.211975526809693,
      "epoch": 1.4423631123919307,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004797915691243863,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20298822820186616,
      "num_tokens": 34414289.0,
      "step": 15015
    },
    {
      "entropy": 5.326760625839233,
      "epoch": 1.4428434197886648,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004797773664124403,
      "loss": 5.1525,
      "mean_token_accuracy": 0.2018410697579384,
      "num_tokens": 34424439.0,
      "step": 15020
    },
    {
      "entropy": 5.264997339248657,
      "epoch": 1.4433237271853987,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047976315894617195,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20948225110769272,
      "num_tokens": 34437299.0,
      "step": 15025
    },
    {
      "entropy": 5.232417917251587,
      "epoch": 1.4438040345821326,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000479748946725911,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22084558010101318,
      "num_tokens": 34448397.0,
      "step": 15030
    },
    {
      "entropy": 5.225221157073975,
      "epoch": 1.4442843419788665,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004797347297519875,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20993768870830537,
      "num_tokens": 34459389.0,
      "step": 15035
    },
    {
      "entropy": 5.318525695800782,
      "epoch": 1.4447646493756003,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047972050802473154,
      "loss": 5.1029,
      "mean_token_accuracy": 0.2109626367688179,
      "num_tokens": 34470610.0,
      "step": 15040
    },
    {
      "entropy": 5.215408706665039,
      "epoch": 1.4452449567723342,
      "grad_norm": 1.5,
      "learning_rate": 0.0004797062815444733,
      "loss": 5.0286,
      "mean_token_accuracy": 0.20810011476278306,
      "num_tokens": 34482100.0,
      "step": 15045
    },
    {
      "entropy": 5.217554759979248,
      "epoch": 1.4457252641690683,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000479692050311543,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20242715179920195,
      "num_tokens": 34493086.0,
      "step": 15050
    },
    {
      "entropy": 5.238348293304443,
      "epoch": 1.4462055715658022,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004796778143262711,
      "loss": 4.9872,
      "mean_token_accuracy": 0.21600108295679094,
      "num_tokens": 34505030.0,
      "step": 15055
    },
    {
      "entropy": 5.313660669326782,
      "epoch": 1.446685878962536,
      "grad_norm": 1.5,
      "learning_rate": 0.0004796635735889882,
      "loss": 5.1749,
      "mean_token_accuracy": 0.19981304407119752,
      "num_tokens": 34516790.0,
      "step": 15060
    },
    {
      "entropy": 5.43970890045166,
      "epoch": 1.44716618635927,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047964932810002476,
      "loss": 5.1725,
      "mean_token_accuracy": 0.1968842104077339,
      "num_tokens": 34530333.0,
      "step": 15065
    },
    {
      "entropy": 5.284123325347901,
      "epoch": 1.4476464937560038,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004796350778597117,
      "loss": 5.0665,
      "mean_token_accuracy": 0.21023496240377426,
      "num_tokens": 34540945.0,
      "step": 15070
    },
    {
      "entropy": 5.2773909091949465,
      "epoch": 1.4481268011527377,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004796208228683796,
      "loss": 5.0547,
      "mean_token_accuracy": 0.20392337590456008,
      "num_tokens": 34551449.0,
      "step": 15075
    },
    {
      "entropy": 5.281215381622315,
      "epoch": 1.4486071085494716,
      "grad_norm": 2.09375,
      "learning_rate": 0.00047960656312635977,
      "loss": 5.1059,
      "mean_token_accuracy": 0.1994602620601654,
      "num_tokens": 34562841.0,
      "step": 15080
    },
    {
      "entropy": 5.243448734283447,
      "epoch": 1.4490874159462055,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004795922986339831,
      "loss": 5.0696,
      "mean_token_accuracy": 0.20891574621200562,
      "num_tokens": 34574408.0,
      "step": 15085
    },
    {
      "entropy": 5.3178098678588865,
      "epoch": 1.4495677233429394,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047957802939158057,
      "loss": 5.1262,
      "mean_token_accuracy": 0.20618992745876313,
      "num_tokens": 34586243.0,
      "step": 15090
    },
    {
      "entropy": 5.319449186325073,
      "epoch": 1.4500480307396733,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004795637553994838,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20813077688217163,
      "num_tokens": 34596631.0,
      "step": 15095
    },
    {
      "entropy": 5.225552940368653,
      "epoch": 1.4505283381364074,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047954947665802404,
      "loss": 5.0402,
      "mean_token_accuracy": 0.21303620785474778,
      "num_tokens": 34607952.0,
      "step": 15100
    },
    {
      "entropy": 5.213094091415405,
      "epoch": 1.4510086455331412,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004795351931675329,
      "loss": 4.981,
      "mean_token_accuracy": 0.21471683979034423,
      "num_tokens": 34618355.0,
      "step": 15105
    },
    {
      "entropy": 5.24654712677002,
      "epoch": 1.4514889529298751,
      "grad_norm": 2.796875,
      "learning_rate": 0.0004795209049283419,
      "loss": 5.0676,
      "mean_token_accuracy": 0.2041841059923172,
      "num_tokens": 34629402.0,
      "step": 15110
    },
    {
      "entropy": 5.386718082427978,
      "epoch": 1.451969260326609,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004795066119407827,
      "loss": 5.1946,
      "mean_token_accuracy": 0.2045331761240959,
      "num_tokens": 34642523.0,
      "step": 15115
    },
    {
      "entropy": 5.212902164459228,
      "epoch": 1.4524495677233429,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004794923142051873,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21584298014640807,
      "num_tokens": 34653714.0,
      "step": 15120
    },
    {
      "entropy": 5.182364654541016,
      "epoch": 1.452929875120077,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047947801172188755,
      "loss": 5.036,
      "mean_token_accuracy": 0.2145914539694786,
      "num_tokens": 34664899.0,
      "step": 15125
    },
    {
      "entropy": 5.332994508743286,
      "epoch": 1.4534101825168109,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004794637044912155,
      "loss": 5.0754,
      "mean_token_accuracy": 0.2075731113553047,
      "num_tokens": 34674888.0,
      "step": 15130
    },
    {
      "entropy": 5.209937715530396,
      "epoch": 1.4538904899135447,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004794493925135034,
      "loss": 4.9905,
      "mean_token_accuracy": 0.2192991480231285,
      "num_tokens": 34686446.0,
      "step": 15135
    },
    {
      "entropy": 5.28668475151062,
      "epoch": 1.4543707973102786,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047943507578908357,
      "loss": 5.1737,
      "mean_token_accuracy": 0.20910208225250243,
      "num_tokens": 34697592.0,
      "step": 15140
    },
    {
      "entropy": 5.289400386810303,
      "epoch": 1.4548511047070125,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004794207543182883,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20559440404176713,
      "num_tokens": 34709305.0,
      "step": 15145
    },
    {
      "entropy": 5.3380097389221195,
      "epoch": 1.4553314121037464,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047940642810145005,
      "loss": 5.1334,
      "mean_token_accuracy": 0.20686376243829727,
      "num_tokens": 34721606.0,
      "step": 15150
    },
    {
      "entropy": 5.272457504272461,
      "epoch": 1.4558117195004803,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047939209713890156,
      "loss": 5.0309,
      "mean_token_accuracy": 0.2123032346367836,
      "num_tokens": 34732125.0,
      "step": 15155
    },
    {
      "entropy": 5.2216168403625485,
      "epoch": 1.4562920268972142,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047937776143097547,
      "loss": 5.0543,
      "mean_token_accuracy": 0.2082948014140129,
      "num_tokens": 34743317.0,
      "step": 15160
    },
    {
      "entropy": 5.252063512802124,
      "epoch": 1.456772334293948,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004793634209780047,
      "loss": 5.0502,
      "mean_token_accuracy": 0.20208578407764435,
      "num_tokens": 34754546.0,
      "step": 15165
    },
    {
      "entropy": 5.268165588378906,
      "epoch": 1.457252641690682,
      "grad_norm": 1.375,
      "learning_rate": 0.0004793490757803221,
      "loss": 4.9745,
      "mean_token_accuracy": 0.21133567541837692,
      "num_tokens": 34766027.0,
      "step": 15170
    },
    {
      "entropy": 5.266101264953614,
      "epoch": 1.457732949087416,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047933472583826063,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20642741322517394,
      "num_tokens": 34777750.0,
      "step": 15175
    },
    {
      "entropy": 5.297080039978027,
      "epoch": 1.45821325648415,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004793203711521537,
      "loss": 5.1403,
      "mean_token_accuracy": 0.2117946356534958,
      "num_tokens": 34788924.0,
      "step": 15180
    },
    {
      "entropy": 5.303713798522949,
      "epoch": 1.4586935638808838,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047930601172233446,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21331677138805388,
      "num_tokens": 34800153.0,
      "step": 15185
    },
    {
      "entropy": 5.21817979812622,
      "epoch": 1.4591738712776177,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047929164754913624,
      "loss": 5.0755,
      "mean_token_accuracy": 0.20696344673633577,
      "num_tokens": 34811597.0,
      "step": 15190
    },
    {
      "entropy": 5.2976597309112545,
      "epoch": 1.4596541786743515,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004792772786328926,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2146271750330925,
      "num_tokens": 34822661.0,
      "step": 15195
    },
    {
      "entropy": 5.368168544769287,
      "epoch": 1.4601344860710854,
      "grad_norm": 1.375,
      "learning_rate": 0.00047926290497393714,
      "loss": 5.1503,
      "mean_token_accuracy": 0.20504291653633117,
      "num_tokens": 34833554.0,
      "step": 15200
    },
    {
      "entropy": 5.29137921333313,
      "epoch": 1.4606147934678195,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004792485265726036,
      "loss": 5.1182,
      "mean_token_accuracy": 0.21235002279281617,
      "num_tokens": 34846060.0,
      "step": 15205
    },
    {
      "entropy": 5.259025621414184,
      "epoch": 1.4610951008645534,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004792341434292257,
      "loss": 5.0791,
      "mean_token_accuracy": 0.2074443742632866,
      "num_tokens": 34857324.0,
      "step": 15210
    },
    {
      "entropy": 5.265723371505738,
      "epoch": 1.4615754082612873,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004792197555441374,
      "loss": 5.0234,
      "mean_token_accuracy": 0.2169741615653038,
      "num_tokens": 34869317.0,
      "step": 15215
    },
    {
      "entropy": 5.273576354980468,
      "epoch": 1.4620557156580212,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004792053629176729,
      "loss": 5.0356,
      "mean_token_accuracy": 0.21141736805438996,
      "num_tokens": 34880961.0,
      "step": 15220
    },
    {
      "entropy": 5.246563386917114,
      "epoch": 1.462536023054755,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004791909655501662,
      "loss": 5.048,
      "mean_token_accuracy": 0.2085846558213234,
      "num_tokens": 34893518.0,
      "step": 15225
    },
    {
      "entropy": 5.290282106399536,
      "epoch": 1.463016330451489,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004791765634419516,
      "loss": 5.1404,
      "mean_token_accuracy": 0.20116308629512786,
      "num_tokens": 34906201.0,
      "step": 15230
    },
    {
      "entropy": 5.253640222549438,
      "epoch": 1.4634966378482228,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047916215659336343,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2156451940536499,
      "num_tokens": 34916841.0,
      "step": 15235
    },
    {
      "entropy": 5.301081132888794,
      "epoch": 1.4639769452449567,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004791477450047363,
      "loss": 5.1299,
      "mean_token_accuracy": 0.20015170723199843,
      "num_tokens": 34928081.0,
      "step": 15240
    },
    {
      "entropy": 5.301649570465088,
      "epoch": 1.4644572526416906,
      "grad_norm": 1.71875,
      "learning_rate": 0.00047913332867640464,
      "loss": 5.0563,
      "mean_token_accuracy": 0.21239689588546753,
      "num_tokens": 34939072.0,
      "step": 15245
    },
    {
      "entropy": 5.279038524627685,
      "epoch": 1.4649375600384245,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004791189076087033,
      "loss": 5.0667,
      "mean_token_accuracy": 0.2010880798101425,
      "num_tokens": 34950898.0,
      "step": 15250
    },
    {
      "entropy": 5.3510161399841305,
      "epoch": 1.4654178674351586,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047910448180196703,
      "loss": 5.1218,
      "mean_token_accuracy": 0.20341092348098755,
      "num_tokens": 34961560.0,
      "step": 15255
    },
    {
      "entropy": 5.277447891235352,
      "epoch": 1.4658981748318924,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004790900512565307,
      "loss": 5.0944,
      "mean_token_accuracy": 0.21668130159378052,
      "num_tokens": 34973604.0,
      "step": 15260
    },
    {
      "entropy": 5.264111709594727,
      "epoch": 1.4663784822286263,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004790756159727294,
      "loss": 5.1268,
      "mean_token_accuracy": 0.20534325689077376,
      "num_tokens": 34985369.0,
      "step": 15265
    },
    {
      "entropy": 5.428138685226441,
      "epoch": 1.4668587896253602,
      "grad_norm": 1.75,
      "learning_rate": 0.00047906117595089835,
      "loss": 5.162,
      "mean_token_accuracy": 0.20002368241548538,
      "num_tokens": 34996746.0,
      "step": 15270
    },
    {
      "entropy": 5.297950649261475,
      "epoch": 1.467339097022094,
      "grad_norm": 1.375,
      "learning_rate": 0.0004790467311913727,
      "loss": 5.0902,
      "mean_token_accuracy": 0.21288189589977263,
      "num_tokens": 35008032.0,
      "step": 15275
    },
    {
      "entropy": 5.264452171325684,
      "epoch": 1.4678194044188282,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004790322816944879,
      "loss": 5.14,
      "mean_token_accuracy": 0.20516570657491684,
      "num_tokens": 35019423.0,
      "step": 15280
    },
    {
      "entropy": 5.212979125976562,
      "epoch": 1.468299711815562,
      "grad_norm": 1.875,
      "learning_rate": 0.0004790178274605793,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21871455311775206,
      "num_tokens": 35029985.0,
      "step": 15285
    },
    {
      "entropy": 5.301221752166748,
      "epoch": 1.468780019212296,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047900336848998254,
      "loss": 5.0823,
      "mean_token_accuracy": 0.21213140934705735,
      "num_tokens": 35040580.0,
      "step": 15290
    },
    {
      "entropy": 5.328329038619995,
      "epoch": 1.4692603266090298,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004789889047830334,
      "loss": 5.1254,
      "mean_token_accuracy": 0.20806695073843,
      "num_tokens": 35051414.0,
      "step": 15295
    },
    {
      "entropy": 5.2469557762146,
      "epoch": 1.4697406340057637,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047897443634006766,
      "loss": 5.0538,
      "mean_token_accuracy": 0.20995538234710692,
      "num_tokens": 35062483.0,
      "step": 15300
    },
    {
      "entropy": 5.286959314346314,
      "epoch": 1.4702209414024976,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004789599631614211,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21461502760648726,
      "num_tokens": 35073536.0,
      "step": 15305
    },
    {
      "entropy": 5.361008644104004,
      "epoch": 1.4707012487992315,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004789454852474298,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20300833880901337,
      "num_tokens": 35085636.0,
      "step": 15310
    },
    {
      "entropy": 5.265331411361695,
      "epoch": 1.4711815561959654,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004789310025984299,
      "loss": 5.0261,
      "mean_token_accuracy": 0.20781455636024476,
      "num_tokens": 35096660.0,
      "step": 15315
    },
    {
      "entropy": 5.2564716815948485,
      "epoch": 1.4716618635926992,
      "grad_norm": 1.765625,
      "learning_rate": 0.00047891651521475776,
      "loss": 5.1574,
      "mean_token_accuracy": 0.20880174040794372,
      "num_tokens": 35109657.0,
      "step": 15320
    },
    {
      "entropy": 5.285657548904419,
      "epoch": 1.4721421709894331,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047890202309674963,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20403075814247132,
      "num_tokens": 35120399.0,
      "step": 15325
    },
    {
      "entropy": 5.229372262954712,
      "epoch": 1.4726224783861672,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047888752624474195,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21126300990581512,
      "num_tokens": 35130105.0,
      "step": 15330
    },
    {
      "entropy": 5.180285120010376,
      "epoch": 1.473102785782901,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004788730246590714,
      "loss": 5.0143,
      "mean_token_accuracy": 0.21084060668945312,
      "num_tokens": 35142547.0,
      "step": 15335
    },
    {
      "entropy": 5.228218030929566,
      "epoch": 1.473583093179635,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047885851834007456,
      "loss": 4.9535,
      "mean_token_accuracy": 0.2190911203622818,
      "num_tokens": 35154603.0,
      "step": 15340
    },
    {
      "entropy": 5.231724548339844,
      "epoch": 1.4740634005763689,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047884400728808824,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20942963510751725,
      "num_tokens": 35165727.0,
      "step": 15345
    },
    {
      "entropy": 5.230791282653809,
      "epoch": 1.4745437079731027,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004788294915034494,
      "loss": 5.0854,
      "mean_token_accuracy": 0.20756847113370896,
      "num_tokens": 35177642.0,
      "step": 15350
    },
    {
      "entropy": 5.239924430847168,
      "epoch": 1.4750240153698366,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000478814970986495,
      "loss": 5.0651,
      "mean_token_accuracy": 0.215902940928936,
      "num_tokens": 35189377.0,
      "step": 15355
    },
    {
      "entropy": 5.279846048355102,
      "epoch": 1.4755043227665707,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047880044573756213,
      "loss": 5.1178,
      "mean_token_accuracy": 0.2091296285390854,
      "num_tokens": 35201244.0,
      "step": 15360
    },
    {
      "entropy": 5.318231058120728,
      "epoch": 1.4759846301633046,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047878591575698816,
      "loss": 5.0806,
      "mean_token_accuracy": 0.21272775530815125,
      "num_tokens": 35212491.0,
      "step": 15365
    },
    {
      "entropy": 5.219100284576416,
      "epoch": 1.4764649375600385,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004787713810451103,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21304252743721008,
      "num_tokens": 35225552.0,
      "step": 15370
    },
    {
      "entropy": 5.237793064117431,
      "epoch": 1.4769452449567724,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047875684160226606,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21291471719741822,
      "num_tokens": 35237961.0,
      "step": 15375
    },
    {
      "entropy": 5.214386749267578,
      "epoch": 1.4774255523535063,
      "grad_norm": 1.375,
      "learning_rate": 0.000478742297428793,
      "loss": 5.0873,
      "mean_token_accuracy": 0.20448803454637526,
      "num_tokens": 35249502.0,
      "step": 15380
    },
    {
      "entropy": 5.322081899642944,
      "epoch": 1.4779058597502401,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047872774852502877,
      "loss": 5.1412,
      "mean_token_accuracy": 0.20390703082084655,
      "num_tokens": 35262978.0,
      "step": 15385
    },
    {
      "entropy": 5.300397205352783,
      "epoch": 1.478386167146974,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004787131948913112,
      "loss": 5.1611,
      "mean_token_accuracy": 0.2024422511458397,
      "num_tokens": 35273289.0,
      "step": 15390
    },
    {
      "entropy": 5.296235036849976,
      "epoch": 1.478866474543708,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047869863652797806,
      "loss": 5.0756,
      "mean_token_accuracy": 0.2075693428516388,
      "num_tokens": 35283827.0,
      "step": 15395
    },
    {
      "entropy": 5.320119428634643,
      "epoch": 1.4793467819404418,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004786840734353675,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20777103453874587,
      "num_tokens": 35293264.0,
      "step": 15400
    },
    {
      "entropy": 5.235420227050781,
      "epoch": 1.4798270893371757,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047866950561381756,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21415072530508042,
      "num_tokens": 35304380.0,
      "step": 15405
    },
    {
      "entropy": 5.257650518417359,
      "epoch": 1.4803073967339098,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004786549330636665,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21200386434793472,
      "num_tokens": 35316805.0,
      "step": 15410
    },
    {
      "entropy": 5.253449153900147,
      "epoch": 1.4807877041306436,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047864035578525256,
      "loss": 5.0568,
      "mean_token_accuracy": 0.2062395393848419,
      "num_tokens": 35328645.0,
      "step": 15415
    },
    {
      "entropy": 5.269427728652954,
      "epoch": 1.4812680115273775,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004786257737789143,
      "loss": 5.0526,
      "mean_token_accuracy": 0.20431149303913115,
      "num_tokens": 35339983.0,
      "step": 15420
    },
    {
      "entropy": 5.349887561798096,
      "epoch": 1.4817483189241114,
      "grad_norm": 1.375,
      "learning_rate": 0.0004786111870449902,
      "loss": 5.1712,
      "mean_token_accuracy": 0.19953811317682266,
      "num_tokens": 35350720.0,
      "step": 15425
    },
    {
      "entropy": 5.341303443908691,
      "epoch": 1.4822286263208453,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047859659558381894,
      "loss": 5.1057,
      "mean_token_accuracy": 0.20577046871185303,
      "num_tokens": 35361435.0,
      "step": 15430
    },
    {
      "entropy": 5.259214162826538,
      "epoch": 1.4827089337175792,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047858199939573935,
      "loss": 5.1709,
      "mean_token_accuracy": 0.20134855806827545,
      "num_tokens": 35372829.0,
      "step": 15435
    },
    {
      "entropy": 5.302060174942016,
      "epoch": 1.4831892411143133,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047856739848109014,
      "loss": 5.0627,
      "mean_token_accuracy": 0.2088722586631775,
      "num_tokens": 35383270.0,
      "step": 15440
    },
    {
      "entropy": 5.404217195510864,
      "epoch": 1.4836695485110472,
      "grad_norm": 2.0,
      "learning_rate": 0.00047855279284021046,
      "loss": 5.1603,
      "mean_token_accuracy": 0.19722400307655336,
      "num_tokens": 35396308.0,
      "step": 15445
    },
    {
      "entropy": 5.256095743179321,
      "epoch": 1.484149855907781,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047853818247343933,
      "loss": 5.0819,
      "mean_token_accuracy": 0.20537677854299546,
      "num_tokens": 35408944.0,
      "step": 15450
    },
    {
      "entropy": 5.302842712402343,
      "epoch": 1.484630163304515,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00047852356738111606,
      "loss": 5.1385,
      "mean_token_accuracy": 0.19910149425268173,
      "num_tokens": 35422227.0,
      "step": 15455
    },
    {
      "entropy": 5.3131180763244625,
      "epoch": 1.4851104707012488,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004785089475635799,
      "loss": 5.1191,
      "mean_token_accuracy": 0.20303317457437514,
      "num_tokens": 35434110.0,
      "step": 15460
    },
    {
      "entropy": 5.309653091430664,
      "epoch": 1.4855907780979827,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047849432302117024,
      "loss": 5.0758,
      "mean_token_accuracy": 0.2114386036992073,
      "num_tokens": 35445194.0,
      "step": 15465
    },
    {
      "entropy": 5.327378606796264,
      "epoch": 1.4860710854947166,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047847969375422656,
      "loss": 5.0797,
      "mean_token_accuracy": 0.20937245488166809,
      "num_tokens": 35455953.0,
      "step": 15470
    },
    {
      "entropy": 5.234394311904907,
      "epoch": 1.4865513928914504,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004784650597630887,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21788979768753053,
      "num_tokens": 35466932.0,
      "step": 15475
    },
    {
      "entropy": 5.172629976272583,
      "epoch": 1.4870317002881843,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047845042104809635,
      "loss": 5.0431,
      "mean_token_accuracy": 0.20887638628482819,
      "num_tokens": 35477755.0,
      "step": 15480
    },
    {
      "entropy": 5.241114091873169,
      "epoch": 1.4875120076849184,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004784357776095892,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2165546402335167,
      "num_tokens": 35489913.0,
      "step": 15485
    },
    {
      "entropy": 5.307248115539551,
      "epoch": 1.4879923150816523,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004784211294479075,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20971139073371886,
      "num_tokens": 35500314.0,
      "step": 15490
    },
    {
      "entropy": 5.379426622390747,
      "epoch": 1.4884726224783862,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004784064765633912,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19546196460723878,
      "num_tokens": 35514717.0,
      "step": 15495
    },
    {
      "entropy": 5.256443691253662,
      "epoch": 1.48895292987512,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047839181895638057,
      "loss": 5.0606,
      "mean_token_accuracy": 0.20679570138454437,
      "num_tokens": 35526122.0,
      "step": 15500
    },
    {
      "entropy": 5.26741738319397,
      "epoch": 1.489433237271854,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047837715662721575,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20276448130607605,
      "num_tokens": 35538270.0,
      "step": 15505
    },
    {
      "entropy": 5.379626131057739,
      "epoch": 1.4899135446685878,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004783624895762372,
      "loss": 5.166,
      "mean_token_accuracy": 0.20006187856197358,
      "num_tokens": 35549598.0,
      "step": 15510
    },
    {
      "entropy": 5.2867998600006105,
      "epoch": 1.490393852065322,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047834781780378563,
      "loss": 5.0654,
      "mean_token_accuracy": 0.20930274277925492,
      "num_tokens": 35559623.0,
      "step": 15515
    },
    {
      "entropy": 5.216621780395508,
      "epoch": 1.4908741594620558,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004783331413102015,
      "loss": 5.0583,
      "mean_token_accuracy": 0.20890338271856307,
      "num_tokens": 35571578.0,
      "step": 15520
    },
    {
      "entropy": 5.3204789638519285,
      "epoch": 1.4913544668587897,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047831846009582557,
      "loss": 5.1406,
      "mean_token_accuracy": 0.20226700007915496,
      "num_tokens": 35583570.0,
      "step": 15525
    },
    {
      "entropy": 5.342668199539185,
      "epoch": 1.4918347742555236,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004783037741609988,
      "loss": 5.1061,
      "mean_token_accuracy": 0.20622288733720778,
      "num_tokens": 35595360.0,
      "step": 15530
    },
    {
      "entropy": 5.286706161499024,
      "epoch": 1.4923150816522575,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004782890835060621,
      "loss": 5.065,
      "mean_token_accuracy": 0.20960413962602614,
      "num_tokens": 35607326.0,
      "step": 15535
    },
    {
      "entropy": 5.339612913131714,
      "epoch": 1.4927953890489913,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004782743881313564,
      "loss": 5.0989,
      "mean_token_accuracy": 0.20399579107761384,
      "num_tokens": 35618504.0,
      "step": 15540
    },
    {
      "entropy": 5.283051824569702,
      "epoch": 1.4932756964457252,
      "grad_norm": 2.21875,
      "learning_rate": 0.00047825968803722315,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21066266894340516,
      "num_tokens": 35631111.0,
      "step": 15545
    },
    {
      "entropy": 5.291447925567627,
      "epoch": 1.493756003842459,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004782449832240035,
      "loss": 5.0637,
      "mean_token_accuracy": 0.21451565027236938,
      "num_tokens": 35643434.0,
      "step": 15550
    },
    {
      "entropy": 5.339574813842773,
      "epoch": 1.494236311239193,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004782302736920387,
      "loss": 5.1316,
      "mean_token_accuracy": 0.2066397711634636,
      "num_tokens": 35654531.0,
      "step": 15555
    },
    {
      "entropy": 5.290758752822876,
      "epoch": 1.4947166186359269,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004782155594416705,
      "loss": 5.0561,
      "mean_token_accuracy": 0.21272497326135636,
      "num_tokens": 35665488.0,
      "step": 15560
    },
    {
      "entropy": 5.311057758331299,
      "epoch": 1.495196926032661,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047820084047324045,
      "loss": 5.1335,
      "mean_token_accuracy": 0.20088960230350494,
      "num_tokens": 35677110.0,
      "step": 15565
    },
    {
      "entropy": 5.33135576248169,
      "epoch": 1.4956772334293948,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047818611678709027,
      "loss": 5.1463,
      "mean_token_accuracy": 0.20527877807617187,
      "num_tokens": 35688694.0,
      "step": 15570
    },
    {
      "entropy": 5.253985261917114,
      "epoch": 1.4961575408261287,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004781713883835618,
      "loss": 5.0471,
      "mean_token_accuracy": 0.21116443127393722,
      "num_tokens": 35699497.0,
      "step": 15575
    },
    {
      "entropy": 5.329415607452392,
      "epoch": 1.4966378482228626,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047815665526299695,
      "loss": 5.1148,
      "mean_token_accuracy": 0.21109289228916167,
      "num_tokens": 35711034.0,
      "step": 15580
    },
    {
      "entropy": 5.268395280838012,
      "epoch": 1.4971181556195965,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004781419174257378,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21558750420808792,
      "num_tokens": 35721804.0,
      "step": 15585
    },
    {
      "entropy": 5.2015190601348875,
      "epoch": 1.4975984630163304,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004781271748721266,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21348223388195037,
      "num_tokens": 35732897.0,
      "step": 15590
    },
    {
      "entropy": 5.12036361694336,
      "epoch": 1.4980787704130645,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004781124276025055,
      "loss": 4.9133,
      "mean_token_accuracy": 0.2235010579228401,
      "num_tokens": 35743792.0,
      "step": 15595
    },
    {
      "entropy": 5.311236619949341,
      "epoch": 1.4985590778097984,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000478097675617217,
      "loss": 5.0781,
      "mean_token_accuracy": 0.20348804593086242,
      "num_tokens": 35754659.0,
      "step": 15600
    },
    {
      "entropy": 5.3396155834198,
      "epoch": 1.4990393852065322,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047808291891660357,
      "loss": 5.0725,
      "mean_token_accuracy": 0.21164245009422303,
      "num_tokens": 35766269.0,
      "step": 15605
    },
    {
      "entropy": 5.294134092330933,
      "epoch": 1.4995196926032661,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047806815750100774,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20552843809127808,
      "num_tokens": 35777634.0,
      "step": 15610
    },
    {
      "entropy": 5.257100009918213,
      "epoch": 1.5,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004780533913707723,
      "loss": 5.1116,
      "mean_token_accuracy": 0.2068232223391533,
      "num_tokens": 35788689.0,
      "step": 15615
    },
    {
      "entropy": 5.240042877197266,
      "epoch": 1.5004803073967339,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047803862052624006,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2107036903500557,
      "num_tokens": 35799365.0,
      "step": 15620
    },
    {
      "entropy": 5.251093864440918,
      "epoch": 1.5009606147934678,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047802384496775397,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2142290636897087,
      "num_tokens": 35810682.0,
      "step": 15625
    },
    {
      "entropy": 5.279915189743042,
      "epoch": 1.5014409221902016,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004780090646956571,
      "loss": 5.1032,
      "mean_token_accuracy": 0.20221735686063766,
      "num_tokens": 35822193.0,
      "step": 15630
    },
    {
      "entropy": 5.238804769515991,
      "epoch": 1.5019212295869355,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047799427971029245,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20516337752342223,
      "num_tokens": 35835486.0,
      "step": 15635
    },
    {
      "entropy": 5.301007413864136,
      "epoch": 1.5024015369836694,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004779794900120034,
      "loss": 5.0668,
      "mean_token_accuracy": 0.20400108844041825,
      "num_tokens": 35846245.0,
      "step": 15640
    },
    {
      "entropy": 5.308107185363769,
      "epoch": 1.5028818443804035,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004779646956011334,
      "loss": 5.1995,
      "mean_token_accuracy": 0.20816765427589418,
      "num_tokens": 35857914.0,
      "step": 15645
    },
    {
      "entropy": 5.34468150138855,
      "epoch": 1.5033621517771374,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047794989647802574,
      "loss": 5.114,
      "mean_token_accuracy": 0.20237296372652053,
      "num_tokens": 35870235.0,
      "step": 15650
    },
    {
      "entropy": 5.297115421295166,
      "epoch": 1.5038424591738713,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047793509264302424,
      "loss": 5.1154,
      "mean_token_accuracy": 0.20965593457221984,
      "num_tokens": 35882154.0,
      "step": 15655
    },
    {
      "entropy": 5.285312223434448,
      "epoch": 1.5043227665706052,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047792028409647237,
      "loss": 5.0558,
      "mean_token_accuracy": 0.21763764321804047,
      "num_tokens": 35894192.0,
      "step": 15660
    },
    {
      "entropy": 5.291246938705444,
      "epoch": 1.5048030739673393,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047790547083871414,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20638402551412582,
      "num_tokens": 35906017.0,
      "step": 15665
    },
    {
      "entropy": 5.370081996917724,
      "epoch": 1.5052833813640731,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047789065287009335,
      "loss": 5.1382,
      "mean_token_accuracy": 0.2023767948150635,
      "num_tokens": 35917671.0,
      "step": 15670
    },
    {
      "entropy": 5.218689823150635,
      "epoch": 1.505763688760807,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004778758301909542,
      "loss": 4.989,
      "mean_token_accuracy": 0.21625009775161744,
      "num_tokens": 35928421.0,
      "step": 15675
    },
    {
      "entropy": 5.302108383178711,
      "epoch": 1.506243996157541,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004778610028016405,
      "loss": 5.1576,
      "mean_token_accuracy": 0.19806920289993285,
      "num_tokens": 35939689.0,
      "step": 15680
    },
    {
      "entropy": 5.319546222686768,
      "epoch": 1.5067243035542748,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004778461707024967,
      "loss": 5.1669,
      "mean_token_accuracy": 0.20428922027349472,
      "num_tokens": 35952342.0,
      "step": 15685
    },
    {
      "entropy": 5.432180309295655,
      "epoch": 1.5072046109510087,
      "grad_norm": 1.25,
      "learning_rate": 0.0004778313338938672,
      "loss": 5.2185,
      "mean_token_accuracy": 0.18969587981700897,
      "num_tokens": 35964726.0,
      "step": 15690
    },
    {
      "entropy": 5.314167118072509,
      "epoch": 1.5076849183477425,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047781649237609643,
      "loss": 5.0679,
      "mean_token_accuracy": 0.2140672653913498,
      "num_tokens": 35976932.0,
      "step": 15695
    },
    {
      "entropy": 5.177508735656739,
      "epoch": 1.5081652257444764,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004778016461495289,
      "loss": 4.9877,
      "mean_token_accuracy": 0.21566703617572786,
      "num_tokens": 35988371.0,
      "step": 15700
    },
    {
      "entropy": 5.1939784526824955,
      "epoch": 1.5086455331412103,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004777867952145094,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21708894222974778,
      "num_tokens": 36001310.0,
      "step": 15705
    },
    {
      "entropy": 5.380842065811157,
      "epoch": 1.5091258405379442,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004777719395713826,
      "loss": 5.1793,
      "mean_token_accuracy": 0.2018696367740631,
      "num_tokens": 36011975.0,
      "step": 15710
    },
    {
      "entropy": 5.314497995376587,
      "epoch": 1.509606147934678,
      "grad_norm": 2.296875,
      "learning_rate": 0.00047775707922049354,
      "loss": 5.1289,
      "mean_token_accuracy": 0.2025901824235916,
      "num_tokens": 36024987.0,
      "step": 15715
    },
    {
      "entropy": 5.2834936618804935,
      "epoch": 1.510086455331412,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004777422141621871,
      "loss": 5.0594,
      "mean_token_accuracy": 0.21496993452310562,
      "num_tokens": 36036003.0,
      "step": 15720
    },
    {
      "entropy": 5.321617317199707,
      "epoch": 1.510566762728146,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004777273443968085,
      "loss": 5.1046,
      "mean_token_accuracy": 0.2051626592874527,
      "num_tokens": 36047948.0,
      "step": 15725
    },
    {
      "entropy": 5.259036493301392,
      "epoch": 1.51104707012488,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004777124699247029,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21570937037467958,
      "num_tokens": 36059197.0,
      "step": 15730
    },
    {
      "entropy": 5.356362247467041,
      "epoch": 1.5115273775216138,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004776975907462157,
      "loss": 5.1536,
      "mean_token_accuracy": 0.2007529005408287,
      "num_tokens": 36072167.0,
      "step": 15735
    },
    {
      "entropy": 5.428820323944092,
      "epoch": 1.5120076849183477,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004776827068616924,
      "loss": 5.2142,
      "mean_token_accuracy": 0.20043275207281114,
      "num_tokens": 36083805.0,
      "step": 15740
    },
    {
      "entropy": 5.261936664581299,
      "epoch": 1.5124879923150818,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004776678182714785,
      "loss": 5.0334,
      "mean_token_accuracy": 0.2192924901843071,
      "num_tokens": 36095450.0,
      "step": 15745
    },
    {
      "entropy": 5.2254188537597654,
      "epoch": 1.5129682997118157,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047765292497591955,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21666243225336074,
      "num_tokens": 36106319.0,
      "step": 15750
    },
    {
      "entropy": 5.213980007171631,
      "epoch": 1.5134486071085496,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047763802697536146,
      "loss": 4.8768,
      "mean_token_accuracy": 0.2263483852148056,
      "num_tokens": 36117015.0,
      "step": 15755
    },
    {
      "entropy": 5.2086950778961185,
      "epoch": 1.5139289145052834,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047762312427015015,
      "loss": 5.06,
      "mean_token_accuracy": 0.2043185532093048,
      "num_tokens": 36127308.0,
      "step": 15760
    },
    {
      "entropy": 5.209178447723389,
      "epoch": 1.5144092219020173,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00047760821686063153,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2166517823934555,
      "num_tokens": 36139811.0,
      "step": 15765
    },
    {
      "entropy": 5.309093379974366,
      "epoch": 1.5148895292987512,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047759330474715173,
      "loss": 5.114,
      "mean_token_accuracy": 0.20372600555419923,
      "num_tokens": 36151330.0,
      "step": 15770
    },
    {
      "entropy": 5.272545194625854,
      "epoch": 1.515369836695485,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047757838793005704,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20881228595972062,
      "num_tokens": 36162436.0,
      "step": 15775
    },
    {
      "entropy": 5.384903907775879,
      "epoch": 1.515850144092219,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047756346640969366,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19001675993204117,
      "num_tokens": 36175642.0,
      "step": 15780
    },
    {
      "entropy": 5.273928260803222,
      "epoch": 1.5163304514889528,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047754854018640803,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21238639652729036,
      "num_tokens": 36186925.0,
      "step": 15785
    },
    {
      "entropy": 5.234462308883667,
      "epoch": 1.5168107588856867,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047753360926054684,
      "loss": 5.1206,
      "mean_token_accuracy": 0.20451428294181823,
      "num_tokens": 36199280.0,
      "step": 15790
    },
    {
      "entropy": 5.25186710357666,
      "epoch": 1.5172910662824206,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047751867363245653,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21442606896162034,
      "num_tokens": 36210276.0,
      "step": 15795
    },
    {
      "entropy": 5.308847427368164,
      "epoch": 1.5177713736791547,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004775037333024841,
      "loss": 5.0939,
      "mean_token_accuracy": 0.21022081077098848,
      "num_tokens": 36221211.0,
      "step": 15800
    },
    {
      "entropy": 5.184071493148804,
      "epoch": 1.5182516810758886,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004774887882709762,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21422246247529983,
      "num_tokens": 36233465.0,
      "step": 15805
    },
    {
      "entropy": 5.244770431518555,
      "epoch": 1.5187319884726225,
      "grad_norm": 1.125,
      "learning_rate": 0.00047747383853827995,
      "loss": 5.0085,
      "mean_token_accuracy": 0.21385688185691834,
      "num_tokens": 36245594.0,
      "step": 15810
    },
    {
      "entropy": 5.290256834030151,
      "epoch": 1.5192122958693564,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004774588841047424,
      "loss": 5.1658,
      "mean_token_accuracy": 0.20378238260746,
      "num_tokens": 36257776.0,
      "step": 15815
    },
    {
      "entropy": 5.228780794143677,
      "epoch": 1.5196926032660905,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004774439249707108,
      "loss": 5.0301,
      "mean_token_accuracy": 0.20076089203357697,
      "num_tokens": 36269056.0,
      "step": 15820
    },
    {
      "entropy": 5.312287473678589,
      "epoch": 1.5201729106628243,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004774289611365323,
      "loss": 5.0286,
      "mean_token_accuracy": 0.21069958060979843,
      "num_tokens": 36280700.0,
      "step": 15825
    },
    {
      "entropy": 5.290642881393433,
      "epoch": 1.5206532180595582,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047741399260255434,
      "loss": 5.1105,
      "mean_token_accuracy": 0.2072647675871849,
      "num_tokens": 36291708.0,
      "step": 15830
    },
    {
      "entropy": 5.25113525390625,
      "epoch": 1.521133525456292,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047739901936912467,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2113155946135521,
      "num_tokens": 36303830.0,
      "step": 15835
    },
    {
      "entropy": 5.239939832687378,
      "epoch": 1.521613832853026,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004773840414365907,
      "loss": 5.054,
      "mean_token_accuracy": 0.20686713457107545,
      "num_tokens": 36316514.0,
      "step": 15840
    },
    {
      "entropy": 5.359360122680664,
      "epoch": 1.5220941402497599,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047736905880530026,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20628927201032637,
      "num_tokens": 36328234.0,
      "step": 15845
    },
    {
      "entropy": 5.224095439910888,
      "epoch": 1.5225744476464937,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004773540714756012,
      "loss": 5.0476,
      "mean_token_accuracy": 0.2123289853334427,
      "num_tokens": 36340383.0,
      "step": 15850
    },
    {
      "entropy": 5.235283374786377,
      "epoch": 1.5230547550432276,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047733907944784144,
      "loss": 5.0387,
      "mean_token_accuracy": 0.2122533932328224,
      "num_tokens": 36351970.0,
      "step": 15855
    },
    {
      "entropy": 5.2437056541442875,
      "epoch": 1.5235350624399615,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004773240827223691,
      "loss": 5.0417,
      "mean_token_accuracy": 0.21088377386331558,
      "num_tokens": 36363163.0,
      "step": 15860
    },
    {
      "entropy": 5.288681173324585,
      "epoch": 1.5240153698366954,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004773090812995323,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20718586444854736,
      "num_tokens": 36375079.0,
      "step": 15865
    },
    {
      "entropy": 5.348295879364014,
      "epoch": 1.5244956772334293,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047729407517967945,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2144293427467346,
      "num_tokens": 36386402.0,
      "step": 15870
    },
    {
      "entropy": 5.189296436309815,
      "epoch": 1.5249759846301632,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047727906436315884,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2155287355184555,
      "num_tokens": 36397973.0,
      "step": 15875
    },
    {
      "entropy": 5.2684698581695555,
      "epoch": 1.5254562920268973,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047726404885031895,
      "loss": 5.1513,
      "mean_token_accuracy": 0.2061857908964157,
      "num_tokens": 36409941.0,
      "step": 15880
    },
    {
      "entropy": 5.366504812240601,
      "epoch": 1.5259365994236311,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047724902864150845,
      "loss": 5.1267,
      "mean_token_accuracy": 0.20563569217920302,
      "num_tokens": 36420609.0,
      "step": 15885
    },
    {
      "entropy": 5.274050235748291,
      "epoch": 1.526416906820365,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047723400373707607,
      "loss": 5.0355,
      "mean_token_accuracy": 0.21531638950109483,
      "num_tokens": 36430854.0,
      "step": 15890
    },
    {
      "entropy": 5.174607944488526,
      "epoch": 1.526897214217099,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004772189741373707,
      "loss": 4.9675,
      "mean_token_accuracy": 0.2161065921187401,
      "num_tokens": 36442225.0,
      "step": 15895
    },
    {
      "entropy": 5.425528860092163,
      "epoch": 1.527377521613833,
      "grad_norm": 1.25,
      "learning_rate": 0.00047720393984274117,
      "loss": 5.1917,
      "mean_token_accuracy": 0.20283327996730804,
      "num_tokens": 36454488.0,
      "step": 15900
    },
    {
      "entropy": 5.307177448272705,
      "epoch": 1.5278578290105669,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047718890085353654,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21385109275579453,
      "num_tokens": 36466197.0,
      "step": 15905
    },
    {
      "entropy": 5.264429569244385,
      "epoch": 1.5283381364073008,
      "grad_norm": 1.375,
      "learning_rate": 0.000477173857170106,
      "loss": 5.1278,
      "mean_token_accuracy": 0.2023659110069275,
      "num_tokens": 36478159.0,
      "step": 15910
    },
    {
      "entropy": 5.284144496917724,
      "epoch": 1.5288184438040346,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047715880879279894,
      "loss": 5.1859,
      "mean_token_accuracy": 0.20323507338762284,
      "num_tokens": 36489621.0,
      "step": 15915
    },
    {
      "entropy": 5.326041889190674,
      "epoch": 1.5292987512007685,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004771437557219646,
      "loss": 5.1407,
      "mean_token_accuracy": 0.20648288428783418,
      "num_tokens": 36500629.0,
      "step": 15920
    },
    {
      "entropy": 5.341318130493164,
      "epoch": 1.5297790585975024,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004771286979579524,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20869407504796983,
      "num_tokens": 36512110.0,
      "step": 15925
    },
    {
      "entropy": 5.194867134094238,
      "epoch": 1.5302593659942363,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004771136355011121,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21974752396345137,
      "num_tokens": 36522499.0,
      "step": 15930
    },
    {
      "entropy": 5.258851051330566,
      "epoch": 1.5307396733909702,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047709856835179333,
      "loss": 5.1065,
      "mean_token_accuracy": 0.2018541917204857,
      "num_tokens": 36533465.0,
      "step": 15935
    },
    {
      "entropy": 5.257210063934326,
      "epoch": 1.531219980787704,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047708349651034586,
      "loss": 5.0497,
      "mean_token_accuracy": 0.2098389357328415,
      "num_tokens": 36545213.0,
      "step": 15940
    },
    {
      "entropy": 5.140851926803589,
      "epoch": 1.531700288184438,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047706841997711974,
      "loss": 4.9181,
      "mean_token_accuracy": 0.2193902850151062,
      "num_tokens": 36555656.0,
      "step": 15945
    },
    {
      "entropy": 5.246776342391968,
      "epoch": 1.5321805955811718,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047705333875246495,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20277945548295975,
      "num_tokens": 36568027.0,
      "step": 15950
    },
    {
      "entropy": 5.254421949386597,
      "epoch": 1.532660902977906,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047703825283673153,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21252552270889283,
      "num_tokens": 36579434.0,
      "step": 15955
    },
    {
      "entropy": 5.285301399230957,
      "epoch": 1.5331412103746398,
      "grad_norm": 1.375,
      "learning_rate": 0.0004770231622302699,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20386855751276017,
      "num_tokens": 36591068.0,
      "step": 15960
    },
    {
      "entropy": 5.2406610488891605,
      "epoch": 1.5336215177713737,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047700806693343016,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2093963220715523,
      "num_tokens": 36603013.0,
      "step": 15965
    },
    {
      "entropy": 5.3147179126739506,
      "epoch": 1.5341018251681076,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047699296694656316,
      "loss": 5.1481,
      "mean_token_accuracy": 0.1987731099128723,
      "num_tokens": 36614565.0,
      "step": 15970
    },
    {
      "entropy": 5.266924571990967,
      "epoch": 1.5345821325648417,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004769778622700192,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2138664171099663,
      "num_tokens": 36625685.0,
      "step": 15975
    },
    {
      "entropy": 5.118635845184326,
      "epoch": 1.5350624399615755,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047696275290414885,
      "loss": 4.9711,
      "mean_token_accuracy": 0.21178223490715026,
      "num_tokens": 36637278.0,
      "step": 15980
    },
    {
      "entropy": 5.228858089447021,
      "epoch": 1.5355427473583094,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047694763884930324,
      "loss": 5.054,
      "mean_token_accuracy": 0.21276589930057527,
      "num_tokens": 36649527.0,
      "step": 15985
    },
    {
      "entropy": 5.299609565734864,
      "epoch": 1.5360230547550433,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047693252010583314,
      "loss": 5.0175,
      "mean_token_accuracy": 0.2079237163066864,
      "num_tokens": 36660653.0,
      "step": 15990
    },
    {
      "entropy": 5.255507516860962,
      "epoch": 1.5365033621517772,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004769173966740895,
      "loss": 5.036,
      "mean_token_accuracy": 0.21136794239282608,
      "num_tokens": 36672340.0,
      "step": 15995
    },
    {
      "entropy": 5.339828586578369,
      "epoch": 1.536983669548511,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047690226855442346,
      "loss": 5.1572,
      "mean_token_accuracy": 0.19855871498584748,
      "num_tokens": 36684468.0,
      "step": 16000
    },
    {
      "entropy": 5.280662965774536,
      "epoch": 1.537463976945245,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004768871357471863,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20143088400363923,
      "num_tokens": 36695921.0,
      "step": 16005
    },
    {
      "entropy": 5.188514041900635,
      "epoch": 1.5379442843419788,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047687199825272936,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2142717495560646,
      "num_tokens": 36707029.0,
      "step": 16010
    },
    {
      "entropy": 5.168916320800781,
      "epoch": 1.5384245917387127,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047685685607140403,
      "loss": 4.9427,
      "mean_token_accuracy": 0.21739741116762162,
      "num_tokens": 36717899.0,
      "step": 16015
    },
    {
      "entropy": 5.235824537277222,
      "epoch": 1.5389048991354466,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047684170920356185,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20655650347471238,
      "num_tokens": 36728691.0,
      "step": 16020
    },
    {
      "entropy": 5.338007831573487,
      "epoch": 1.5393852065321805,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004768265576495546,
      "loss": 5.1606,
      "mean_token_accuracy": 0.20265884101390838,
      "num_tokens": 36741122.0,
      "step": 16025
    },
    {
      "entropy": 5.32447624206543,
      "epoch": 1.5398655139289144,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047681140140973396,
      "loss": 5.1761,
      "mean_token_accuracy": 0.20372225046157838,
      "num_tokens": 36752841.0,
      "step": 16030
    },
    {
      "entropy": 5.309655427932739,
      "epoch": 1.5403458213256485,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004767962404844517,
      "loss": 5.07,
      "mean_token_accuracy": 0.2140301287174225,
      "num_tokens": 36763030.0,
      "step": 16035
    },
    {
      "entropy": 5.336140155792236,
      "epoch": 1.5408261287223823,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047678107487406015,
      "loss": 5.1773,
      "mean_token_accuracy": 0.2017665073275566,
      "num_tokens": 36774442.0,
      "step": 16040
    },
    {
      "entropy": 5.290905284881592,
      "epoch": 1.5413064361191162,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047676590457891116,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20406427532434462,
      "num_tokens": 36785650.0,
      "step": 16045
    },
    {
      "entropy": 5.233311128616333,
      "epoch": 1.54178674351585,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004767507295993569,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20976938009262086,
      "num_tokens": 36797021.0,
      "step": 16050
    },
    {
      "entropy": 5.24737401008606,
      "epoch": 1.5422670509125842,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004767355499357498,
      "loss": 5.0366,
      "mean_token_accuracy": 0.21222082525491714,
      "num_tokens": 36807593.0,
      "step": 16055
    },
    {
      "entropy": 5.213295841217041,
      "epoch": 1.542747358309318,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004767203655884423,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21032444834709169,
      "num_tokens": 36820185.0,
      "step": 16060
    },
    {
      "entropy": 5.201232671737671,
      "epoch": 1.543227665706052,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004767051765577869,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2195556178689003,
      "num_tokens": 36831350.0,
      "step": 16065
    },
    {
      "entropy": 5.160927248001099,
      "epoch": 1.5437079731027858,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047668998284413624,
      "loss": 4.9337,
      "mean_token_accuracy": 0.21568286567926406,
      "num_tokens": 36842475.0,
      "step": 16070
    },
    {
      "entropy": 5.229477119445801,
      "epoch": 1.5441882804995197,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047667478444784306,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2159205123782158,
      "num_tokens": 36853385.0,
      "step": 16075
    },
    {
      "entropy": 5.261432743072509,
      "epoch": 1.5446685878962536,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004766595813692602,
      "loss": 5.0516,
      "mean_token_accuracy": 0.2112713173031807,
      "num_tokens": 36866339.0,
      "step": 16080
    },
    {
      "entropy": 5.308079862594605,
      "epoch": 1.5451488952929875,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047664437360874076,
      "loss": 5.027,
      "mean_token_accuracy": 0.21027812957763672,
      "num_tokens": 36877943.0,
      "step": 16085
    },
    {
      "entropy": 5.313610219955445,
      "epoch": 1.5456292026897214,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047662916116663766,
      "loss": 5.0776,
      "mean_token_accuracy": 0.208051660656929,
      "num_tokens": 36889343.0,
      "step": 16090
    },
    {
      "entropy": 5.296591854095459,
      "epoch": 1.5461095100864553,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047661394404330417,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20525844395160675,
      "num_tokens": 36900727.0,
      "step": 16095
    },
    {
      "entropy": 5.194662761688233,
      "epoch": 1.5465898174831891,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047659872223909357,
      "loss": 5.0083,
      "mean_token_accuracy": 0.21567150056362153,
      "num_tokens": 36911959.0,
      "step": 16100
    },
    {
      "entropy": 5.222887802124023,
      "epoch": 1.547070124879923,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004765834957543592,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21292728632688523,
      "num_tokens": 36924964.0,
      "step": 16105
    },
    {
      "entropy": 5.29455246925354,
      "epoch": 1.547550432276657,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047656826458945475,
      "loss": 5.1379,
      "mean_token_accuracy": 0.19735962599515916,
      "num_tokens": 36937033.0,
      "step": 16110
    },
    {
      "entropy": 5.24070782661438,
      "epoch": 1.548030739673391,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047655302874473365,
      "loss": 5.0067,
      "mean_token_accuracy": 0.21257947981357575,
      "num_tokens": 36948745.0,
      "step": 16115
    },
    {
      "entropy": 5.271581840515137,
      "epoch": 1.5485110470701249,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004765377882205498,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21620627343654633,
      "num_tokens": 36961210.0,
      "step": 16120
    },
    {
      "entropy": 5.301712226867676,
      "epoch": 1.5489913544668588,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004765225430172568,
      "loss": 5.0994,
      "mean_token_accuracy": 0.2087859705090523,
      "num_tokens": 36973720.0,
      "step": 16125
    },
    {
      "entropy": 5.313554525375366,
      "epoch": 1.5494716618635929,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004765072931352089,
      "loss": 5.142,
      "mean_token_accuracy": 0.20792601704597474,
      "num_tokens": 36986899.0,
      "step": 16130
    },
    {
      "entropy": 5.233713960647583,
      "epoch": 1.5499519692603267,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047649203857476,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20893001109361647,
      "num_tokens": 36998324.0,
      "step": 16135
    },
    {
      "entropy": 5.324620294570923,
      "epoch": 1.5504322766570606,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047647677933626423,
      "loss": 5.0977,
      "mean_token_accuracy": 0.20946022868156433,
      "num_tokens": 37009779.0,
      "step": 16140
    },
    {
      "entropy": 5.3257222175598145,
      "epoch": 1.5509125840537945,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047646151542007583,
      "loss": 5.1256,
      "mean_token_accuracy": 0.20851856619119644,
      "num_tokens": 37022473.0,
      "step": 16145
    },
    {
      "entropy": 5.3737406730651855,
      "epoch": 1.5513928914505284,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004764462468265494,
      "loss": 5.1733,
      "mean_token_accuracy": 0.20351529866456985,
      "num_tokens": 37033456.0,
      "step": 16150
    },
    {
      "entropy": 5.307319021224975,
      "epoch": 1.5518731988472623,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047643097355603913,
      "loss": 5.0911,
      "mean_token_accuracy": 0.2019636735320091,
      "num_tokens": 37044983.0,
      "step": 16155
    },
    {
      "entropy": 5.2649799346923825,
      "epoch": 1.5523535062439962,
      "grad_norm": 1.25,
      "learning_rate": 0.0004764156956088999,
      "loss": 5.0103,
      "mean_token_accuracy": 0.2107056200504303,
      "num_tokens": 37055992.0,
      "step": 16160
    },
    {
      "entropy": 5.28945484161377,
      "epoch": 1.55283381364073,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004764004129854863,
      "loss": 5.0612,
      "mean_token_accuracy": 0.21935467422008514,
      "num_tokens": 37068227.0,
      "step": 16165
    },
    {
      "entropy": 5.24505615234375,
      "epoch": 1.553314121037464,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047638512568615307,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20980026423931122,
      "num_tokens": 37080221.0,
      "step": 16170
    },
    {
      "entropy": 5.275307083129883,
      "epoch": 1.5537944284341978,
      "grad_norm": 1.25,
      "learning_rate": 0.0004763698337112553,
      "loss": 5.065,
      "mean_token_accuracy": 0.20427938997745515,
      "num_tokens": 37092223.0,
      "step": 16175
    },
    {
      "entropy": 5.313152074813843,
      "epoch": 1.5542747358309317,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004763545370611479,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20058134198188782,
      "num_tokens": 37104169.0,
      "step": 16180
    },
    {
      "entropy": 5.3100502490997314,
      "epoch": 1.5547550432276656,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047633923573618605,
      "loss": 5.1343,
      "mean_token_accuracy": 0.21133484840393066,
      "num_tokens": 37115784.0,
      "step": 16185
    },
    {
      "entropy": 5.213553047180175,
      "epoch": 1.5552353506243997,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000476323929736725,
      "loss": 4.9743,
      "mean_token_accuracy": 0.21303585469722747,
      "num_tokens": 37128018.0,
      "step": 16190
    },
    {
      "entropy": 5.269845485687256,
      "epoch": 1.5557156580211335,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047630861906312004,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20416450053453444,
      "num_tokens": 37139330.0,
      "step": 16195
    },
    {
      "entropy": 5.2091107845306395,
      "epoch": 1.5561959654178674,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004762933037157268,
      "loss": 4.9173,
      "mean_token_accuracy": 0.22389539182186127,
      "num_tokens": 37151531.0,
      "step": 16200
    },
    {
      "entropy": 5.324626302719116,
      "epoch": 1.5566762728146013,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047627798369490076,
      "loss": 5.0656,
      "mean_token_accuracy": 0.21038767993450164,
      "num_tokens": 37162442.0,
      "step": 16205
    },
    {
      "entropy": 5.265551280975342,
      "epoch": 1.5571565802113354,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047626265900099757,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21128504127264022,
      "num_tokens": 37173737.0,
      "step": 16210
    },
    {
      "entropy": 5.166654253005982,
      "epoch": 1.5576368876080693,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047624732963437314,
      "loss": 5.0187,
      "mean_token_accuracy": 0.20972042232751847,
      "num_tokens": 37185983.0,
      "step": 16215
    },
    {
      "entropy": 5.355982494354248,
      "epoch": 1.5581171950048032,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047623199559538324,
      "loss": 5.1252,
      "mean_token_accuracy": 0.19982872605323793,
      "num_tokens": 37197119.0,
      "step": 16220
    },
    {
      "entropy": 5.41767578125,
      "epoch": 1.558597502401537,
      "grad_norm": 1.421875,
      "learning_rate": 0.000476216656884384,
      "loss": 5.1997,
      "mean_token_accuracy": 0.20376020073890685,
      "num_tokens": 37209516.0,
      "step": 16225
    },
    {
      "entropy": 5.147062158584594,
      "epoch": 1.559077809798271,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047620131350173135,
      "loss": 4.998,
      "mean_token_accuracy": 0.21488342583179473,
      "num_tokens": 37220559.0,
      "step": 16230
    },
    {
      "entropy": 5.298512125015259,
      "epoch": 1.5595581171950048,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004761859654477817,
      "loss": 5.0877,
      "mean_token_accuracy": 0.20814523994922637,
      "num_tokens": 37232232.0,
      "step": 16235
    },
    {
      "entropy": 5.21407380104065,
      "epoch": 1.5600384245917387,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004761706127228914,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2175306797027588,
      "num_tokens": 37242819.0,
      "step": 16240
    },
    {
      "entropy": 5.289794158935547,
      "epoch": 1.5605187319884726,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004761552553274168,
      "loss": 5.0877,
      "mean_token_accuracy": 0.20818599164485932,
      "num_tokens": 37254940.0,
      "step": 16245
    },
    {
      "entropy": 5.293817329406738,
      "epoch": 1.5609990393852065,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004761398932617144,
      "loss": 5.1003,
      "mean_token_accuracy": 0.20271167606115342,
      "num_tokens": 37266220.0,
      "step": 16250
    },
    {
      "entropy": 5.356972503662109,
      "epoch": 1.5614793467819403,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000476124526526141,
      "loss": 5.1119,
      "mean_token_accuracy": 0.20252732634544374,
      "num_tokens": 37278136.0,
      "step": 16255
    },
    {
      "entropy": 5.327190256118774,
      "epoch": 1.5619596541786742,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047610915512105327,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20780642479658126,
      "num_tokens": 37290358.0,
      "step": 16260
    },
    {
      "entropy": 5.323932123184204,
      "epoch": 1.562439961575408,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004760937790468082,
      "loss": 5.1327,
      "mean_token_accuracy": 0.20940637588500977,
      "num_tokens": 37300719.0,
      "step": 16265
    },
    {
      "entropy": 5.237097978591919,
      "epoch": 1.5629202689721422,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004760783983037627,
      "loss": 5.0859,
      "mean_token_accuracy": 0.2086906537413597,
      "num_tokens": 37311917.0,
      "step": 16270
    },
    {
      "entropy": 5.3001546382904055,
      "epoch": 1.563400576368876,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004760630128922738,
      "loss": 5.049,
      "mean_token_accuracy": 0.21187455356121063,
      "num_tokens": 37325076.0,
      "step": 16275
    },
    {
      "entropy": 5.280317068099976,
      "epoch": 1.56388088376561,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004760476228126989,
      "loss": 5.0696,
      "mean_token_accuracy": 0.20440367013216018,
      "num_tokens": 37336064.0,
      "step": 16280
    },
    {
      "entropy": 5.310790300369263,
      "epoch": 1.5643611911623438,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004760322280653951,
      "loss": 5.0574,
      "mean_token_accuracy": 0.2066957786679268,
      "num_tokens": 37347343.0,
      "step": 16285
    },
    {
      "entropy": 5.2565226554870605,
      "epoch": 1.564841498559078,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004760168286507199,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21340460777282716,
      "num_tokens": 37359067.0,
      "step": 16290
    },
    {
      "entropy": 5.234368753433228,
      "epoch": 1.5653218059558118,
      "grad_norm": 1.375,
      "learning_rate": 0.00047600142456903085,
      "loss": 5.0371,
      "mean_token_accuracy": 0.20297729969024658,
      "num_tokens": 37369811.0,
      "step": 16295
    },
    {
      "entropy": 5.186442136764526,
      "epoch": 1.5658021133525457,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047598601582068555,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22256670892238617,
      "num_tokens": 37380944.0,
      "step": 16300
    },
    {
      "entropy": 5.303704071044922,
      "epoch": 1.5662824207492796,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004759706024060418,
      "loss": 5.0852,
      "mean_token_accuracy": 0.19702230989933014,
      "num_tokens": 37392099.0,
      "step": 16305
    },
    {
      "entropy": 5.237500286102295,
      "epoch": 1.5667627281460135,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004759551843254575,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2110443815588951,
      "num_tokens": 37404254.0,
      "step": 16310
    },
    {
      "entropy": 5.248646450042725,
      "epoch": 1.5672430355427474,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047593976157929034,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2169898122549057,
      "num_tokens": 37416112.0,
      "step": 16315
    },
    {
      "entropy": 5.21478304862976,
      "epoch": 1.5677233429394812,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004759243341678987,
      "loss": 5.0053,
      "mean_token_accuracy": 0.21101684719324112,
      "num_tokens": 37428350.0,
      "step": 16320
    },
    {
      "entropy": 5.273245477676392,
      "epoch": 1.5682036503362151,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004759089020916407,
      "loss": 5.0826,
      "mean_token_accuracy": 0.21234282851219177,
      "num_tokens": 37440059.0,
      "step": 16325
    },
    {
      "entropy": 5.333327579498291,
      "epoch": 1.568683957732949,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047589346535087444,
      "loss": 5.1015,
      "mean_token_accuracy": 0.20307673066854476,
      "num_tokens": 37450880.0,
      "step": 16330
    },
    {
      "entropy": 5.229888391494751,
      "epoch": 1.5691642651296829,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004758780239459586,
      "loss": 5.0177,
      "mean_token_accuracy": 0.2121018201112747,
      "num_tokens": 37463484.0,
      "step": 16335
    },
    {
      "entropy": 5.300388431549072,
      "epoch": 1.5696445725264168,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004758625778772514,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21022880226373672,
      "num_tokens": 37474952.0,
      "step": 16340
    },
    {
      "entropy": 5.354442882537842,
      "epoch": 1.5701248799231509,
      "grad_norm": 1.625,
      "learning_rate": 0.00047584712714511166,
      "loss": 5.1433,
      "mean_token_accuracy": 0.20635926723480225,
      "num_tokens": 37486858.0,
      "step": 16345
    },
    {
      "entropy": 5.262277698516845,
      "epoch": 1.5706051873198847,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047583167174989797,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21443158239126206,
      "num_tokens": 37497852.0,
      "step": 16350
    },
    {
      "entropy": 5.231578636169433,
      "epoch": 1.5710854947166186,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004758162116919692,
      "loss": 5.081,
      "mean_token_accuracy": 0.2050024688243866,
      "num_tokens": 37509051.0,
      "step": 16355
    },
    {
      "entropy": 5.346564817428589,
      "epoch": 1.5715658021133525,
      "grad_norm": 1.765625,
      "learning_rate": 0.00047580074697168434,
      "loss": 5.1949,
      "mean_token_accuracy": 0.20822606682777406,
      "num_tokens": 37520415.0,
      "step": 16360
    },
    {
      "entropy": 5.324740791320801,
      "epoch": 1.5720461095100866,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047578527758940236,
      "loss": 5.0879,
      "mean_token_accuracy": 0.20940931141376495,
      "num_tokens": 37532087.0,
      "step": 16365
    },
    {
      "entropy": 5.314472961425781,
      "epoch": 1.5725264169068205,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004757698035454825,
      "loss": 5.1177,
      "mean_token_accuracy": 0.2047325611114502,
      "num_tokens": 37545219.0,
      "step": 16370
    },
    {
      "entropy": 5.296878147125244,
      "epoch": 1.5730067243035544,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004757543248402839,
      "loss": 5.0641,
      "mean_token_accuracy": 0.2068307191133499,
      "num_tokens": 37558138.0,
      "step": 16375
    },
    {
      "entropy": 5.250341272354126,
      "epoch": 1.5734870317002883,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047573884147416597,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21454728692770003,
      "num_tokens": 37568975.0,
      "step": 16380
    },
    {
      "entropy": 5.217583799362183,
      "epoch": 1.5739673390970221,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004757233534474883,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21710899472236633,
      "num_tokens": 37580774.0,
      "step": 16385
    },
    {
      "entropy": 5.176962518692017,
      "epoch": 1.574447646493756,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004757078607606103,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21621515303850175,
      "num_tokens": 37592858.0,
      "step": 16390
    },
    {
      "entropy": 5.281411504745483,
      "epoch": 1.57492795389049,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004756923634138918,
      "loss": 5.0725,
      "mean_token_accuracy": 0.20920645743608474,
      "num_tokens": 37604005.0,
      "step": 16395
    },
    {
      "entropy": 5.26087589263916,
      "epoch": 1.5754082612872238,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047567686140769264,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21374571323394775,
      "num_tokens": 37614629.0,
      "step": 16400
    },
    {
      "entropy": 5.253297805786133,
      "epoch": 1.5758885686839577,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047566135474237247,
      "loss": 5.012,
      "mean_token_accuracy": 0.21769997030496596,
      "num_tokens": 37626118.0,
      "step": 16405
    },
    {
      "entropy": 5.368995094299317,
      "epoch": 1.5763688760806915,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047564584341829166,
      "loss": 5.1228,
      "mean_token_accuracy": 0.1988277867436409,
      "num_tokens": 37638514.0,
      "step": 16410
    },
    {
      "entropy": 5.355324220657349,
      "epoch": 1.5768491834774254,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047563032743581,
      "loss": 5.1268,
      "mean_token_accuracy": 0.20276721566915512,
      "num_tokens": 37649602.0,
      "step": 16415
    },
    {
      "entropy": 5.3163281917572025,
      "epoch": 1.5773294908741593,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047561480679528804,
      "loss": 5.0791,
      "mean_token_accuracy": 0.2077769860625267,
      "num_tokens": 37661428.0,
      "step": 16420
    },
    {
      "entropy": 5.3220458984375,
      "epoch": 1.5778097982708934,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004755992814970859,
      "loss": 5.0784,
      "mean_token_accuracy": 0.20953404456377028,
      "num_tokens": 37673608.0,
      "step": 16425
    },
    {
      "entropy": 5.3145702362060545,
      "epoch": 1.5782901056676273,
      "grad_norm": 1.21875,
      "learning_rate": 0.000475583751541564,
      "loss": 5.2129,
      "mean_token_accuracy": 0.20594584494829177,
      "num_tokens": 37684957.0,
      "step": 16430
    },
    {
      "entropy": 5.256752014160156,
      "epoch": 1.5787704130643612,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047556821692908315,
      "loss": 5.1042,
      "mean_token_accuracy": 0.20484959185123444,
      "num_tokens": 37697773.0,
      "step": 16435
    },
    {
      "entropy": 5.332219743728638,
      "epoch": 1.579250720461095,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004755526776600038,
      "loss": 5.0927,
      "mean_token_accuracy": 0.21434657126665116,
      "num_tokens": 37708315.0,
      "step": 16440
    },
    {
      "entropy": 5.259537553787231,
      "epoch": 1.5797310278578292,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047553713373468684,
      "loss": 5.0599,
      "mean_token_accuracy": 0.21168979555368422,
      "num_tokens": 37720440.0,
      "step": 16445
    },
    {
      "entropy": 5.152255868911743,
      "epoch": 1.580211335254563,
      "grad_norm": 1.375,
      "learning_rate": 0.00047552158515349306,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21748272478580474,
      "num_tokens": 37732332.0,
      "step": 16450
    },
    {
      "entropy": 5.258172607421875,
      "epoch": 1.580691642651297,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047550603191678356,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20722013413906099,
      "num_tokens": 37744770.0,
      "step": 16455
    },
    {
      "entropy": 5.220672225952148,
      "epoch": 1.5811719500480308,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004754904740249194,
      "loss": 4.9602,
      "mean_token_accuracy": 0.2158677890896797,
      "num_tokens": 37755958.0,
      "step": 16460
    },
    {
      "entropy": 5.2421502590179445,
      "epoch": 1.5816522574447647,
      "grad_norm": 1.25,
      "learning_rate": 0.00047547491147826156,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20977760702371598,
      "num_tokens": 37767182.0,
      "step": 16465
    },
    {
      "entropy": 5.298257970809937,
      "epoch": 1.5821325648414986,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004754593442771718,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20752999782562256,
      "num_tokens": 37778299.0,
      "step": 16470
    },
    {
      "entropy": 5.3512509822845455,
      "epoch": 1.5826128722382324,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047544377242201115,
      "loss": 5.1202,
      "mean_token_accuracy": 0.209932903945446,
      "num_tokens": 37789879.0,
      "step": 16475
    },
    {
      "entropy": 5.277742910385132,
      "epoch": 1.5830931796349663,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047542819591314136,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20411130785942078,
      "num_tokens": 37800726.0,
      "step": 16480
    },
    {
      "entropy": 5.245403289794922,
      "epoch": 1.5835734870317002,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004754126147509241,
      "loss": 5.0559,
      "mean_token_accuracy": 0.21008965671062468,
      "num_tokens": 37812298.0,
      "step": 16485
    },
    {
      "entropy": 5.207891798019409,
      "epoch": 1.584053794428434,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047539702893572086,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2234435871243477,
      "num_tokens": 37824433.0,
      "step": 16490
    },
    {
      "entropy": 5.285567092895508,
      "epoch": 1.584534101825168,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047538143846789376,
      "loss": 5.0593,
      "mean_token_accuracy": 0.21146337389945985,
      "num_tokens": 37835647.0,
      "step": 16495
    },
    {
      "entropy": 5.297049045562744,
      "epoch": 1.585014409221902,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004753658433478047,
      "loss": 5.1325,
      "mean_token_accuracy": 0.20734177231788636,
      "num_tokens": 37847721.0,
      "step": 16500
    },
    {
      "entropy": 5.261937236785888,
      "epoch": 1.585494716618636,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047535024357581564,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21704123616218568,
      "num_tokens": 37860111.0,
      "step": 16505
    },
    {
      "entropy": 5.264764404296875,
      "epoch": 1.5859750240153698,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004753346391522889,
      "loss": 5.1111,
      "mean_token_accuracy": 0.2045636460185051,
      "num_tokens": 37872321.0,
      "step": 16510
    },
    {
      "entropy": 5.3495707511901855,
      "epoch": 1.5864553314121037,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047531903007758667,
      "loss": 5.1816,
      "mean_token_accuracy": 0.20022435188293458,
      "num_tokens": 37885195.0,
      "step": 16515
    },
    {
      "entropy": 5.283568954467773,
      "epoch": 1.5869356388088378,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004753034163520714,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21376079469919204,
      "num_tokens": 37895487.0,
      "step": 16520
    },
    {
      "entropy": 5.148827981948853,
      "epoch": 1.5874159462055717,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047528779797610557,
      "loss": 5.0943,
      "mean_token_accuracy": 0.21099332571029664,
      "num_tokens": 37908344.0,
      "step": 16525
    },
    {
      "entropy": 5.270332765579224,
      "epoch": 1.5878962536023056,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047527217495005184,
      "loss": 5.0429,
      "mean_token_accuracy": 0.20828526616096496,
      "num_tokens": 37920220.0,
      "step": 16530
    },
    {
      "entropy": 5.245607471466064,
      "epoch": 1.5883765609990395,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047525654727427285,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22014959752559662,
      "num_tokens": 37930874.0,
      "step": 16535
    },
    {
      "entropy": 5.193498611450195,
      "epoch": 1.5888568683957733,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004752409149491315,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21008171141147614,
      "num_tokens": 37942501.0,
      "step": 16540
    },
    {
      "entropy": 5.232227420806884,
      "epoch": 1.5893371757925072,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047522527797499075,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20897990465164185,
      "num_tokens": 37954408.0,
      "step": 16545
    },
    {
      "entropy": 5.279903078079224,
      "epoch": 1.589817483189241,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004752096363522135,
      "loss": 5.099,
      "mean_token_accuracy": 0.20462866872549057,
      "num_tokens": 37966088.0,
      "step": 16550
    },
    {
      "entropy": 5.294344902038574,
      "epoch": 1.590297790585975,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047519399008116305,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20048427432775498,
      "num_tokens": 37977726.0,
      "step": 16555
    },
    {
      "entropy": 5.285877466201782,
      "epoch": 1.5907780979827089,
      "grad_norm": 1.25,
      "learning_rate": 0.0004751783391622026,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20376301407814026,
      "num_tokens": 37989024.0,
      "step": 16560
    },
    {
      "entropy": 5.3090849876403805,
      "epoch": 1.5912584053794427,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004751626835956955,
      "loss": 5.1019,
      "mean_token_accuracy": 0.20732163786888122,
      "num_tokens": 37999870.0,
      "step": 16565
    },
    {
      "entropy": 5.309950542449951,
      "epoch": 1.5917387127761766,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004751470233820053,
      "loss": 5.1545,
      "mean_token_accuracy": 0.20655997842550278,
      "num_tokens": 38010945.0,
      "step": 16570
    },
    {
      "entropy": 5.246731901168824,
      "epoch": 1.5922190201729105,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004751313585214955,
      "loss": 5.0181,
      "mean_token_accuracy": 0.21244566142559052,
      "num_tokens": 38021584.0,
      "step": 16575
    },
    {
      "entropy": 5.278038787841797,
      "epoch": 1.5926993275696446,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004751156890145298,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20392366647720336,
      "num_tokens": 38034407.0,
      "step": 16580
    },
    {
      "entropy": 5.381368589401245,
      "epoch": 1.5931796349663785,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000475100014861472,
      "loss": 5.1242,
      "mean_token_accuracy": 0.19443738758563994,
      "num_tokens": 38047061.0,
      "step": 16585
    },
    {
      "entropy": 5.389515972137451,
      "epoch": 1.5936599423631124,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004750843360626861,
      "loss": 5.1569,
      "mean_token_accuracy": 0.20738481730222702,
      "num_tokens": 38059168.0,
      "step": 16590
    },
    {
      "entropy": 5.2732805728912355,
      "epoch": 1.5941402497598463,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004750686526185359,
      "loss": 5.0287,
      "mean_token_accuracy": 0.21070666909217833,
      "num_tokens": 38070514.0,
      "step": 16595
    },
    {
      "entropy": 5.386032056808472,
      "epoch": 1.5946205571565804,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047505296452938584,
      "loss": 5.1479,
      "mean_token_accuracy": 0.2041761189699173,
      "num_tokens": 38081995.0,
      "step": 16600
    },
    {
      "entropy": 5.201404476165772,
      "epoch": 1.5951008645533142,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047503727179559995,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2142452597618103,
      "num_tokens": 38094371.0,
      "step": 16605
    },
    {
      "entropy": 5.1635422706604,
      "epoch": 1.5955811719500481,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047502157441754256,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2145054429769516,
      "num_tokens": 38106340.0,
      "step": 16610
    },
    {
      "entropy": 5.283494377136231,
      "epoch": 1.596061479346782,
      "grad_norm": 1.375,
      "learning_rate": 0.0004750058723955781,
      "loss": 5.0255,
      "mean_token_accuracy": 0.20727859139442445,
      "num_tokens": 38117641.0,
      "step": 16615
    },
    {
      "entropy": 5.303549814224243,
      "epoch": 1.5965417867435159,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004749901657300713,
      "loss": 5.0238,
      "mean_token_accuracy": 0.21383918225765228,
      "num_tokens": 38129628.0,
      "step": 16620
    },
    {
      "entropy": 5.194593620300293,
      "epoch": 1.5970220941402498,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047497445442138667,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21010721921920777,
      "num_tokens": 38140843.0,
      "step": 16625
    },
    {
      "entropy": 5.2041994571685795,
      "epoch": 1.5975024015369836,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047495873846988896,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21936007887125014,
      "num_tokens": 38152666.0,
      "step": 16630
    },
    {
      "entropy": 5.19046835899353,
      "epoch": 1.5979827089337175,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004749430178759431,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21735272258520127,
      "num_tokens": 38164102.0,
      "step": 16635
    },
    {
      "entropy": 5.186071395874023,
      "epoch": 1.5984630163304514,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047492729263991413,
      "loss": 5.0112,
      "mean_token_accuracy": 0.21523601263761521,
      "num_tokens": 38175211.0,
      "step": 16640
    },
    {
      "entropy": 5.236240243911743,
      "epoch": 1.5989433237271853,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047491156276216695,
      "loss": 5.0119,
      "mean_token_accuracy": 0.21814688593149184,
      "num_tokens": 38186968.0,
      "step": 16645
    },
    {
      "entropy": 5.305980205535889,
      "epoch": 1.5994236311239192,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047489582824306704,
      "loss": 5.043,
      "mean_token_accuracy": 0.20733065009117127,
      "num_tokens": 38198740.0,
      "step": 16650
    },
    {
      "entropy": 5.222610902786255,
      "epoch": 1.5999039385206533,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047488008908297955,
      "loss": 4.9647,
      "mean_token_accuracy": 0.21612946987152098,
      "num_tokens": 38209110.0,
      "step": 16655
    },
    {
      "entropy": 5.176356554031372,
      "epoch": 1.6003842459173871,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004748643452822699,
      "loss": 4.9521,
      "mean_token_accuracy": 0.21685307919979097,
      "num_tokens": 38221840.0,
      "step": 16660
    },
    {
      "entropy": 5.307543325424194,
      "epoch": 1.600864553314121,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004748485968413036,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20611382126808167,
      "num_tokens": 38232752.0,
      "step": 16665
    },
    {
      "entropy": 5.237250947952271,
      "epoch": 1.601344860710855,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047483284376044634,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21094163358211518,
      "num_tokens": 38243394.0,
      "step": 16670
    },
    {
      "entropy": 5.259588050842285,
      "epoch": 1.601825168107589,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004748170860400638,
      "loss": 5.1263,
      "mean_token_accuracy": 0.21209454983472825,
      "num_tokens": 38254317.0,
      "step": 16675
    },
    {
      "entropy": 5.273589563369751,
      "epoch": 1.602305475504323,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047480132368052185,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2174384355545044,
      "num_tokens": 38266886.0,
      "step": 16680
    },
    {
      "entropy": 5.351137399673462,
      "epoch": 1.6027857829010568,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047478555668218643,
      "loss": 5.1799,
      "mean_token_accuracy": 0.20333765000104903,
      "num_tokens": 38279975.0,
      "step": 16685
    },
    {
      "entropy": 5.314376354217529,
      "epoch": 1.6032660902977907,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004747697850454237,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20322198569774627,
      "num_tokens": 38290237.0,
      "step": 16690
    },
    {
      "entropy": 5.28905029296875,
      "epoch": 1.6037463976945245,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004747540087705997,
      "loss": 4.9729,
      "mean_token_accuracy": 0.2110549822449684,
      "num_tokens": 38300833.0,
      "step": 16695
    },
    {
      "entropy": 5.205938386917114,
      "epoch": 1.6042267050912584,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004747382278580808,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21970429867506028,
      "num_tokens": 38311926.0,
      "step": 16700
    },
    {
      "entropy": 5.327702045440674,
      "epoch": 1.6047070124879923,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004747224423082333,
      "loss": 5.1622,
      "mean_token_accuracy": 0.20488552898168563,
      "num_tokens": 38323053.0,
      "step": 16705
    },
    {
      "entropy": 5.297062397003174,
      "epoch": 1.6051873198847262,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047470665212142384,
      "loss": 5.1019,
      "mean_token_accuracy": 0.20423240065574647,
      "num_tokens": 38334959.0,
      "step": 16710
    },
    {
      "entropy": 5.330727243423462,
      "epoch": 1.60566762728146,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047469085729801887,
      "loss": 5.1153,
      "mean_token_accuracy": 0.20353061854839324,
      "num_tokens": 38346402.0,
      "step": 16715
    },
    {
      "entropy": 5.324340438842773,
      "epoch": 1.606147934678194,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047467505783838515,
      "loss": 5.059,
      "mean_token_accuracy": 0.20843278616666794,
      "num_tokens": 38358549.0,
      "step": 16720
    },
    {
      "entropy": 5.320571851730347,
      "epoch": 1.6066282420749278,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004746592537428895,
      "loss": 5.0654,
      "mean_token_accuracy": 0.20796492993831633,
      "num_tokens": 38370840.0,
      "step": 16725
    },
    {
      "entropy": 5.275411653518677,
      "epoch": 1.6071085494716617,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047464344501189877,
      "loss": 5.0532,
      "mean_token_accuracy": 0.20882368981838226,
      "num_tokens": 38382980.0,
      "step": 16730
    },
    {
      "entropy": 5.310444355010986,
      "epoch": 1.6075888568683958,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047462763164578015,
      "loss": 4.975,
      "mean_token_accuracy": 0.2197318822145462,
      "num_tokens": 38393836.0,
      "step": 16735
    },
    {
      "entropy": 5.173428392410278,
      "epoch": 1.6080691642651297,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004746118136449007,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22010448724031448,
      "num_tokens": 38405770.0,
      "step": 16740
    },
    {
      "entropy": 5.276047277450561,
      "epoch": 1.6085494716618636,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004745959910096276,
      "loss": 5.155,
      "mean_token_accuracy": 0.19893281459808348,
      "num_tokens": 38418847.0,
      "step": 16745
    },
    {
      "entropy": 5.227228879928589,
      "epoch": 1.6090297790585975,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047458016374032837,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21467060297727586,
      "num_tokens": 38431532.0,
      "step": 16750
    },
    {
      "entropy": 5.35255012512207,
      "epoch": 1.6095100864553316,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004745643318373703,
      "loss": 5.2052,
      "mean_token_accuracy": 0.2069998174905777,
      "num_tokens": 38442071.0,
      "step": 16755
    },
    {
      "entropy": 5.291540718078613,
      "epoch": 1.6099903938520654,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047454849530112106,
      "loss": 5.1153,
      "mean_token_accuracy": 0.20244176536798478,
      "num_tokens": 38454979.0,
      "step": 16760
    },
    {
      "entropy": 5.2175980567932125,
      "epoch": 1.6104707012487993,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047453265413194826,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2209549739956856,
      "num_tokens": 38465579.0,
      "step": 16765
    },
    {
      "entropy": 5.207346630096436,
      "epoch": 1.6109510086455332,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047451680833021973,
      "loss": 5.0139,
      "mean_token_accuracy": 0.21299715340137482,
      "num_tokens": 38477139.0,
      "step": 16770
    },
    {
      "entropy": 5.226201009750366,
      "epoch": 1.611431316042267,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004745009578963034,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2222629204392433,
      "num_tokens": 38487638.0,
      "step": 16775
    },
    {
      "entropy": 5.155400705337525,
      "epoch": 1.611911623439001,
      "grad_norm": 1.25,
      "learning_rate": 0.00047448510283056716,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2282456323504448,
      "num_tokens": 38498395.0,
      "step": 16780
    },
    {
      "entropy": 5.240451812744141,
      "epoch": 1.6123919308357348,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047446924313337925,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21857232898473739,
      "num_tokens": 38509372.0,
      "step": 16785
    },
    {
      "entropy": 5.207878160476684,
      "epoch": 1.6128722382324687,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047445337880510773,
      "loss": 4.9958,
      "mean_token_accuracy": 0.20910652875900268,
      "num_tokens": 38520651.0,
      "step": 16790
    },
    {
      "entropy": 5.276827383041382,
      "epoch": 1.6133525456292026,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004744375098461211,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20441794246435166,
      "num_tokens": 38531461.0,
      "step": 16795
    },
    {
      "entropy": 5.230743646621704,
      "epoch": 1.6138328530259365,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004744216362567876,
      "loss": 4.9437,
      "mean_token_accuracy": 0.22445514053106308,
      "num_tokens": 38542707.0,
      "step": 16800
    },
    {
      "entropy": 5.220602083206177,
      "epoch": 1.6143131604226704,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047440575803747595,
      "loss": 4.942,
      "mean_token_accuracy": 0.22282014936208724,
      "num_tokens": 38553664.0,
      "step": 16805
    },
    {
      "entropy": 5.2639378070831295,
      "epoch": 1.6147934678194045,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047438987518855463,
      "loss": 5.0345,
      "mean_token_accuracy": 0.2077275410294533,
      "num_tokens": 38564891.0,
      "step": 16810
    },
    {
      "entropy": 5.260013437271118,
      "epoch": 1.6152737752161384,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004743739877103926,
      "loss": 5.0344,
      "mean_token_accuracy": 0.2090534508228302,
      "num_tokens": 38576535.0,
      "step": 16815
    },
    {
      "entropy": 5.278713274002075,
      "epoch": 1.6157540826128722,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004743580956033585,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20421395003795623,
      "num_tokens": 38588160.0,
      "step": 16820
    },
    {
      "entropy": 5.233013725280761,
      "epoch": 1.6162343900096061,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047434219886782135,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2204900488257408,
      "num_tokens": 38599281.0,
      "step": 16825
    },
    {
      "entropy": 5.251700162887573,
      "epoch": 1.6167146974063402,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004743262975041504,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21127479076385497,
      "num_tokens": 38611823.0,
      "step": 16830
    },
    {
      "entropy": 5.2695873260498045,
      "epoch": 1.617195004803074,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004743103915127146,
      "loss": 5.0369,
      "mean_token_accuracy": 0.21127189844846725,
      "num_tokens": 38623181.0,
      "step": 16835
    },
    {
      "entropy": 5.293290281295777,
      "epoch": 1.617675312199808,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047429448089388336,
      "loss": 5.0531,
      "mean_token_accuracy": 0.21046160459518432,
      "num_tokens": 38633149.0,
      "step": 16840
    },
    {
      "entropy": 5.279803371429443,
      "epoch": 1.6181556195965419,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047427856564802605,
      "loss": 5.0359,
      "mean_token_accuracy": 0.20733281821012498,
      "num_tokens": 38644204.0,
      "step": 16845
    },
    {
      "entropy": 5.295021915435791,
      "epoch": 1.6186359269932757,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004742626457755122,
      "loss": 5.0199,
      "mean_token_accuracy": 0.21473308950662612,
      "num_tokens": 38655287.0,
      "step": 16850
    },
    {
      "entropy": 5.164002370834351,
      "epoch": 1.6191162343900096,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004742467212767114,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21801784932613372,
      "num_tokens": 38665732.0,
      "step": 16855
    },
    {
      "entropy": 5.27867488861084,
      "epoch": 1.6195965417867435,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004742307921519933,
      "loss": 5.1495,
      "mean_token_accuracy": 0.19636489152908326,
      "num_tokens": 38677289.0,
      "step": 16860
    },
    {
      "entropy": 5.255645895004273,
      "epoch": 1.6200768491834774,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047421485840172794,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2140633836388588,
      "num_tokens": 38689581.0,
      "step": 16865
    },
    {
      "entropy": 5.340537595748901,
      "epoch": 1.6205571565802113,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004741989200262851,
      "loss": 5.0869,
      "mean_token_accuracy": 0.2052430748939514,
      "num_tokens": 38701327.0,
      "step": 16870
    },
    {
      "entropy": 5.261393976211548,
      "epoch": 1.6210374639769451,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004741829770260347,
      "loss": 5.0701,
      "mean_token_accuracy": 0.21008166521787644,
      "num_tokens": 38713431.0,
      "step": 16875
    },
    {
      "entropy": 5.297986125946045,
      "epoch": 1.621517771373679,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047416702940134714,
      "loss": 5.0886,
      "mean_token_accuracy": 0.208334718644619,
      "num_tokens": 38724091.0,
      "step": 16880
    },
    {
      "entropy": 5.2804368019104,
      "epoch": 1.621998078770413,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047415107715259255,
      "loss": 5.1338,
      "mean_token_accuracy": 0.21043217927217484,
      "num_tokens": 38736219.0,
      "step": 16885
    },
    {
      "entropy": 5.270649051666259,
      "epoch": 1.622478386167147,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047413512028014125,
      "loss": 4.9962,
      "mean_token_accuracy": 0.208334381878376,
      "num_tokens": 38747512.0,
      "step": 16890
    },
    {
      "entropy": 5.30160231590271,
      "epoch": 1.622958693563881,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004741191587843638,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20620564818382264,
      "num_tokens": 38758777.0,
      "step": 16895
    },
    {
      "entropy": 5.326295518875122,
      "epoch": 1.6234390009606148,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004741031926656308,
      "loss": 5.1041,
      "mean_token_accuracy": 0.2127739205956459,
      "num_tokens": 38770094.0,
      "step": 16900
    },
    {
      "entropy": 5.30284366607666,
      "epoch": 1.6239193083573487,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004740872219243128,
      "loss": 5.0223,
      "mean_token_accuracy": 0.2147320419549942,
      "num_tokens": 38781419.0,
      "step": 16905
    },
    {
      "entropy": 5.242863655090332,
      "epoch": 1.6243996157540828,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004740712465607807,
      "loss": 5.0236,
      "mean_token_accuracy": 0.2077527403831482,
      "num_tokens": 38794303.0,
      "step": 16910
    },
    {
      "entropy": 5.264979887008667,
      "epoch": 1.6248799231508166,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004740552665754054,
      "loss": 5.0708,
      "mean_token_accuracy": 0.21243511140346527,
      "num_tokens": 38805387.0,
      "step": 16915
    },
    {
      "entropy": 5.273353528976441,
      "epoch": 1.6253602305475505,
      "grad_norm": 1.375,
      "learning_rate": 0.00047403928196855776,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21344126909971237,
      "num_tokens": 38815307.0,
      "step": 16920
    },
    {
      "entropy": 5.26421160697937,
      "epoch": 1.6258405379442844,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047402329274060916,
      "loss": 5.0409,
      "mean_token_accuracy": 0.21460300385951997,
      "num_tokens": 38826585.0,
      "step": 16925
    },
    {
      "entropy": 5.218717956542969,
      "epoch": 1.6263208453410183,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004740072988919306,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2221046581864357,
      "num_tokens": 38838406.0,
      "step": 16930
    },
    {
      "entropy": 5.251041555404663,
      "epoch": 1.6268011527377522,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004739913004228936,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21273664087057115,
      "num_tokens": 38849952.0,
      "step": 16935
    },
    {
      "entropy": 5.286770915985107,
      "epoch": 1.627281460134486,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004739752973338694,
      "loss": 5.1065,
      "mean_token_accuracy": 0.2061857670545578,
      "num_tokens": 38862761.0,
      "step": 16940
    },
    {
      "entropy": 5.251566219329834,
      "epoch": 1.62776176753122,
      "grad_norm": 1.9375,
      "learning_rate": 0.00047395928962522965,
      "loss": 5.0163,
      "mean_token_accuracy": 0.2063119575381279,
      "num_tokens": 38874422.0,
      "step": 16945
    },
    {
      "entropy": 5.2898224830627445,
      "epoch": 1.6282420749279538,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047394327729734595,
      "loss": 5.062,
      "mean_token_accuracy": 0.2077808663249016,
      "num_tokens": 38885751.0,
      "step": 16950
    },
    {
      "entropy": 5.310858154296875,
      "epoch": 1.6287223823246877,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004739272603505901,
      "loss": 5.0429,
      "mean_token_accuracy": 0.21619021743535996,
      "num_tokens": 38897539.0,
      "step": 16955
    },
    {
      "entropy": 5.310702753067017,
      "epoch": 1.6292026897214216,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000473911238785334,
      "loss": 5.032,
      "mean_token_accuracy": 0.2101557433605194,
      "num_tokens": 38908439.0,
      "step": 16960
    },
    {
      "entropy": 5.228362560272217,
      "epoch": 1.6296829971181557,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004738952126019496,
      "loss": 5.0171,
      "mean_token_accuracy": 0.2079322963953018,
      "num_tokens": 38920881.0,
      "step": 16965
    },
    {
      "entropy": 5.240253162384033,
      "epoch": 1.6301633045148896,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004738791818008089,
      "loss": 5.0027,
      "mean_token_accuracy": 0.22017588317394257,
      "num_tokens": 38932097.0,
      "step": 16970
    },
    {
      "entropy": 5.347792482376098,
      "epoch": 1.6306436119116234,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004738631463822841,
      "loss": 5.2127,
      "mean_token_accuracy": 0.20434884130954742,
      "num_tokens": 38945122.0,
      "step": 16975
    },
    {
      "entropy": 5.211736726760864,
      "epoch": 1.6311239193083573,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047384710634674766,
      "loss": 5.0527,
      "mean_token_accuracy": 0.2168850928544998,
      "num_tokens": 38956136.0,
      "step": 16980
    },
    {
      "entropy": 5.308329629898071,
      "epoch": 1.6316042267050914,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047383106169457184,
      "loss": 5.0444,
      "mean_token_accuracy": 0.2087881252169609,
      "num_tokens": 38968650.0,
      "step": 16985
    },
    {
      "entropy": 5.258285427093506,
      "epoch": 1.6320845341018253,
      "grad_norm": 1.25,
      "learning_rate": 0.0004738150124261292,
      "loss": 4.9647,
      "mean_token_accuracy": 0.21761055141687394,
      "num_tokens": 38979834.0,
      "step": 16990
    },
    {
      "entropy": 5.190436744689942,
      "epoch": 1.6325648414985592,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047379895854179226,
      "loss": 4.9621,
      "mean_token_accuracy": 0.21543731540441513,
      "num_tokens": 38991548.0,
      "step": 16995
    },
    {
      "entropy": 5.256105613708496,
      "epoch": 1.633045148895293,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004737829000419338,
      "loss": 5.0299,
      "mean_token_accuracy": 0.2132558599114418,
      "num_tokens": 39003344.0,
      "step": 17000
    },
    {
      "entropy": 5.262235307693482,
      "epoch": 1.633525456292027,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047376683692692666,
      "loss": 5.0475,
      "mean_token_accuracy": 0.21691738367080687,
      "num_tokens": 39015278.0,
      "step": 17005
    },
    {
      "entropy": 5.275490999221802,
      "epoch": 1.6340057636887608,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004737507691971439,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2093028724193573,
      "num_tokens": 39027634.0,
      "step": 17010
    },
    {
      "entropy": 5.337136745452881,
      "epoch": 1.6344860710854947,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047373469685295833,
      "loss": 5.1073,
      "mean_token_accuracy": 0.20952678620815277,
      "num_tokens": 39038226.0,
      "step": 17015
    },
    {
      "entropy": 5.281513595581055,
      "epoch": 1.6349663784822286,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047371861989474326,
      "loss": 5.1001,
      "mean_token_accuracy": 0.20848229825496672,
      "num_tokens": 39050627.0,
      "step": 17020
    },
    {
      "entropy": 5.2767401218414305,
      "epoch": 1.6354466858789625,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004737025383228719,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20673838108778,
      "num_tokens": 39061734.0,
      "step": 17025
    },
    {
      "entropy": 5.243673515319824,
      "epoch": 1.6359269932756964,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047368645213771764,
      "loss": 5.0843,
      "mean_token_accuracy": 0.20264798998832703,
      "num_tokens": 39074288.0,
      "step": 17030
    },
    {
      "entropy": 5.24128212928772,
      "epoch": 1.6364073006724302,
      "grad_norm": 1.125,
      "learning_rate": 0.000473670361339654,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20323335826396943,
      "num_tokens": 39084967.0,
      "step": 17035
    },
    {
      "entropy": 5.264110040664673,
      "epoch": 1.6368876080691641,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004736542659290544,
      "loss": 5.0022,
      "mean_token_accuracy": 0.21481208205223085,
      "num_tokens": 39095093.0,
      "step": 17040
    },
    {
      "entropy": 5.303515958786011,
      "epoch": 1.6373679154658982,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004736381659062927,
      "loss": 5.062,
      "mean_token_accuracy": 0.21389075964689255,
      "num_tokens": 39106952.0,
      "step": 17045
    },
    {
      "entropy": 5.216152048110962,
      "epoch": 1.637848222862632,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047362206127174255,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21349318474531173,
      "num_tokens": 39117682.0,
      "step": 17050
    },
    {
      "entropy": 5.281610774993896,
      "epoch": 1.638328530259366,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047360595202577786,
      "loss": 5.1413,
      "mean_token_accuracy": 0.20124550312757492,
      "num_tokens": 39129767.0,
      "step": 17055
    },
    {
      "entropy": 5.235630655288697,
      "epoch": 1.6388088376560999,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047358983816877284,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2131196990609169,
      "num_tokens": 39140906.0,
      "step": 17060
    },
    {
      "entropy": 5.316436147689819,
      "epoch": 1.639289145052834,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004735737197011015,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2078072652220726,
      "num_tokens": 39151884.0,
      "step": 17065
    },
    {
      "entropy": 5.130867910385132,
      "epoch": 1.6397694524495678,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047355759662313793,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21289096176624298,
      "num_tokens": 39162986.0,
      "step": 17070
    },
    {
      "entropy": 5.274908876419067,
      "epoch": 1.6402497598463017,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004735414689352566,
      "loss": 5.0618,
      "mean_token_accuracy": 0.21092267632484435,
      "num_tokens": 39174280.0,
      "step": 17075
    },
    {
      "entropy": 5.322296667098999,
      "epoch": 1.6407300672430356,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004735253366378318,
      "loss": 5.0382,
      "mean_token_accuracy": 0.2190268874168396,
      "num_tokens": 39185499.0,
      "step": 17080
    },
    {
      "entropy": 5.172384357452392,
      "epoch": 1.6412103746397695,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004735091997312383,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21579830795526506,
      "num_tokens": 39196322.0,
      "step": 17085
    },
    {
      "entropy": 5.206001138687133,
      "epoch": 1.6416906820365034,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047349305821585067,
      "loss": 5.04,
      "mean_token_accuracy": 0.2090488463640213,
      "num_tokens": 39207332.0,
      "step": 17090
    },
    {
      "entropy": 5.306821012496949,
      "epoch": 1.6421709894332372,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004734769120920435,
      "loss": 5.0984,
      "mean_token_accuracy": 0.20602259635925294,
      "num_tokens": 39217305.0,
      "step": 17095
    },
    {
      "entropy": 5.310189342498779,
      "epoch": 1.6426512968299711,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004734607613601919,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21020877212285996,
      "num_tokens": 39228895.0,
      "step": 17100
    },
    {
      "entropy": 5.241233682632446,
      "epoch": 1.643131604226705,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047344460602067077,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21040278673171997,
      "num_tokens": 39240859.0,
      "step": 17105
    },
    {
      "entropy": 5.233461141586304,
      "epoch": 1.643611911623439,
      "grad_norm": 1.328125,
      "learning_rate": 0.000473428446073855,
      "loss": 4.934,
      "mean_token_accuracy": 0.2252657189965248,
      "num_tokens": 39251556.0,
      "step": 17110
    },
    {
      "entropy": 5.218749475479126,
      "epoch": 1.6440922190201728,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047341228152012003,
      "loss": 4.9081,
      "mean_token_accuracy": 0.212572817504406,
      "num_tokens": 39263919.0,
      "step": 17115
    },
    {
      "entropy": 5.313381814956665,
      "epoch": 1.6445725264169067,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000473396112359841,
      "loss": 5.1968,
      "mean_token_accuracy": 0.20562588274478913,
      "num_tokens": 39276873.0,
      "step": 17120
    },
    {
      "entropy": 5.268904972076416,
      "epoch": 1.6450528338136408,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047337993859339334,
      "loss": 4.9691,
      "mean_token_accuracy": 0.22207941859960556,
      "num_tokens": 39288912.0,
      "step": 17125
    },
    {
      "entropy": 5.196393251419067,
      "epoch": 1.6455331412103746,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047336376022115255,
      "loss": 5.0112,
      "mean_token_accuracy": 0.20876137763261796,
      "num_tokens": 39301115.0,
      "step": 17130
    },
    {
      "entropy": 5.259602451324463,
      "epoch": 1.6460134486071085,
      "grad_norm": 1.90625,
      "learning_rate": 0.00047334757724349437,
      "loss": 5.101,
      "mean_token_accuracy": 0.20254710018634797,
      "num_tokens": 39313535.0,
      "step": 17135
    },
    {
      "entropy": 5.304538297653198,
      "epoch": 1.6464937560038426,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004733313896607943,
      "loss": 5.069,
      "mean_token_accuracy": 0.20653516054153442,
      "num_tokens": 39325289.0,
      "step": 17140
    },
    {
      "entropy": 5.264994049072266,
      "epoch": 1.6469740634005765,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004733151974734284,
      "loss": 4.9906,
      "mean_token_accuracy": 0.2167246550321579,
      "num_tokens": 39337382.0,
      "step": 17145
    },
    {
      "entropy": 5.253612899780274,
      "epoch": 1.6474543707973104,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047329900068177245,
      "loss": 5.001,
      "mean_token_accuracy": 0.21116073280572892,
      "num_tokens": 39348548.0,
      "step": 17150
    },
    {
      "entropy": 5.2151390552520756,
      "epoch": 1.6479346781940443,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047328279928620244,
      "loss": 5.021,
      "mean_token_accuracy": 0.20906397700309753,
      "num_tokens": 39359788.0,
      "step": 17155
    },
    {
      "entropy": 5.245170211791992,
      "epoch": 1.6484149855907781,
      "grad_norm": 1.25,
      "learning_rate": 0.0004732665932870947,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21631735414266587,
      "num_tokens": 39370533.0,
      "step": 17160
    },
    {
      "entropy": 5.260925245285034,
      "epoch": 1.648895292987512,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047325038268482544,
      "loss": 5.0598,
      "mean_token_accuracy": 0.2126273214817047,
      "num_tokens": 39381157.0,
      "step": 17165
    },
    {
      "entropy": 5.2953489303588865,
      "epoch": 1.649375600384246,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004732341674797709,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21181332617998122,
      "num_tokens": 39391280.0,
      "step": 17170
    },
    {
      "entropy": 5.163144063949585,
      "epoch": 1.6498559077809798,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047321794767230766,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21328102499246598,
      "num_tokens": 39402097.0,
      "step": 17175
    },
    {
      "entropy": 5.287734651565552,
      "epoch": 1.6503362151777137,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047320172326281224,
      "loss": 5.127,
      "mean_token_accuracy": 0.21006289273500442,
      "num_tokens": 39414590.0,
      "step": 17180
    },
    {
      "entropy": 5.265344190597534,
      "epoch": 1.6508165225744476,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047318549425166134,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21206119507551194,
      "num_tokens": 39425933.0,
      "step": 17185
    },
    {
      "entropy": 5.240490818023682,
      "epoch": 1.6512968299711814,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004731692606392318,
      "loss": 5.0234,
      "mean_token_accuracy": 0.21674046218395232,
      "num_tokens": 39437059.0,
      "step": 17190
    },
    {
      "entropy": 5.227220487594605,
      "epoch": 1.6517771373679153,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004731530224259004,
      "loss": 5.0902,
      "mean_token_accuracy": 0.21029784977436067,
      "num_tokens": 39449582.0,
      "step": 17195
    },
    {
      "entropy": 5.387343311309815,
      "epoch": 1.6522574447646494,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004731367796120442,
      "loss": 5.0971,
      "mean_token_accuracy": 0.2061827763915062,
      "num_tokens": 39461147.0,
      "step": 17200
    },
    {
      "entropy": 5.3197108745574955,
      "epoch": 1.6527377521613833,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004731205321980404,
      "loss": 5.0703,
      "mean_token_accuracy": 0.20503018498420716,
      "num_tokens": 39471881.0,
      "step": 17205
    },
    {
      "entropy": 5.2026856422424315,
      "epoch": 1.6532180595581172,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047310428018426616,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20668946653604509,
      "num_tokens": 39483632.0,
      "step": 17210
    },
    {
      "entropy": 5.291299295425415,
      "epoch": 1.653698366954851,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004730880235710987,
      "loss": 5.0491,
      "mean_token_accuracy": 0.20785726606845856,
      "num_tokens": 39495587.0,
      "step": 17215
    },
    {
      "entropy": 5.261475944519043,
      "epoch": 1.6541786743515852,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004730717623589155,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20801530182361602,
      "num_tokens": 39505503.0,
      "step": 17220
    },
    {
      "entropy": 5.252955913543701,
      "epoch": 1.654658981748319,
      "grad_norm": 1.125,
      "learning_rate": 0.0004730554965480942,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21527422666549684,
      "num_tokens": 39516681.0,
      "step": 17225
    },
    {
      "entropy": 5.236886930465698,
      "epoch": 1.655139289145053,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004730392261390124,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21085644215345384,
      "num_tokens": 39528451.0,
      "step": 17230
    },
    {
      "entropy": 5.244349956512451,
      "epoch": 1.6556195965417868,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004730229511320478,
      "loss": 5.0697,
      "mean_token_accuracy": 0.20715661197900773,
      "num_tokens": 39540306.0,
      "step": 17235
    },
    {
      "entropy": 5.231801843643188,
      "epoch": 1.6560999039385207,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047300667152757827,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21241377294063568,
      "num_tokens": 39551260.0,
      "step": 17240
    },
    {
      "entropy": 5.302245378494263,
      "epoch": 1.6565802113352546,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047299038732598184,
      "loss": 5.1044,
      "mean_token_accuracy": 0.20715683698654175,
      "num_tokens": 39562991.0,
      "step": 17245
    },
    {
      "entropy": 5.224659252166748,
      "epoch": 1.6570605187319885,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047297409852763644,
      "loss": 4.9693,
      "mean_token_accuracy": 0.21658762395381928,
      "num_tokens": 39574629.0,
      "step": 17250
    },
    {
      "entropy": 5.217621088027954,
      "epoch": 1.6575408261287223,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004729578051329204,
      "loss": 5.0876,
      "mean_token_accuracy": 0.20619237571954727,
      "num_tokens": 39585599.0,
      "step": 17255
    },
    {
      "entropy": 5.282077312469482,
      "epoch": 1.6580211335254562,
      "grad_norm": 1.703125,
      "learning_rate": 0.00047294150714221185,
      "loss": 5.032,
      "mean_token_accuracy": 0.21100164353847503,
      "num_tokens": 39597225.0,
      "step": 17260
    },
    {
      "entropy": 5.317398548126221,
      "epoch": 1.65850144092219,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004729252045558894,
      "loss": 5.072,
      "mean_token_accuracy": 0.21067775338888167,
      "num_tokens": 39609296.0,
      "step": 17265
    },
    {
      "entropy": 5.185437250137329,
      "epoch": 1.658981748318924,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047290889737433133,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22053769528865813,
      "num_tokens": 39619369.0,
      "step": 17270
    },
    {
      "entropy": 5.197672367095947,
      "epoch": 1.6594620557156579,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047289258559791633,
      "loss": 4.971,
      "mean_token_accuracy": 0.21856908351182938,
      "num_tokens": 39630438.0,
      "step": 17275
    },
    {
      "entropy": 5.25117073059082,
      "epoch": 1.659942363112392,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047287626922702317,
      "loss": 5.0497,
      "mean_token_accuracy": 0.20963386446237564,
      "num_tokens": 39643339.0,
      "step": 17280
    },
    {
      "entropy": 5.265747737884522,
      "epoch": 1.6604226705091258,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047285994826203054,
      "loss": 5.023,
      "mean_token_accuracy": 0.21861969530582429,
      "num_tokens": 39654495.0,
      "step": 17285
    },
    {
      "entropy": 5.214551115036011,
      "epoch": 1.6609029779058597,
      "grad_norm": 1.5,
      "learning_rate": 0.0004728436227033175,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21828881949186324,
      "num_tokens": 39665870.0,
      "step": 17290
    },
    {
      "entropy": 5.215697717666626,
      "epoch": 1.6613832853025938,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047282729255126294,
      "loss": 4.952,
      "mean_token_accuracy": 0.22071336805820466,
      "num_tokens": 39676761.0,
      "step": 17295
    },
    {
      "entropy": 5.267432975769043,
      "epoch": 1.6618635926993277,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004728109578062461,
      "loss": 5.0139,
      "mean_token_accuracy": 0.213969849050045,
      "num_tokens": 39687515.0,
      "step": 17300
    },
    {
      "entropy": 5.183841037750244,
      "epoch": 1.6623439000960616,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047279461846864626,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2195118173956871,
      "num_tokens": 39699599.0,
      "step": 17305
    },
    {
      "entropy": 5.22152419090271,
      "epoch": 1.6628242074927955,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047277827453884265,
      "loss": 4.9665,
      "mean_token_accuracy": 0.22828542590141296,
      "num_tokens": 39710122.0,
      "step": 17310
    },
    {
      "entropy": 5.310120964050293,
      "epoch": 1.6633045148895294,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047276192601721477,
      "loss": 5.1114,
      "mean_token_accuracy": 0.20437227934598923,
      "num_tokens": 39722884.0,
      "step": 17315
    },
    {
      "entropy": 5.2060455799102785,
      "epoch": 1.6637848222862632,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004727455729041422,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21905168145895004,
      "num_tokens": 39734464.0,
      "step": 17320
    },
    {
      "entropy": 5.250668239593506,
      "epoch": 1.6642651296829971,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047272921520000465,
      "loss": 5.0398,
      "mean_token_accuracy": 0.2118943989276886,
      "num_tokens": 39745948.0,
      "step": 17325
    },
    {
      "entropy": 5.232309103012085,
      "epoch": 1.664745437079731,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004727128529051819,
      "loss": 5.009,
      "mean_token_accuracy": 0.21044907867908477,
      "num_tokens": 39757024.0,
      "step": 17330
    },
    {
      "entropy": 5.170520305633545,
      "epoch": 1.6652257444764649,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004726964860200537,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2167271926999092,
      "num_tokens": 39767205.0,
      "step": 17335
    },
    {
      "entropy": 5.283964967727661,
      "epoch": 1.6657060518731988,
      "grad_norm": 1.375,
      "learning_rate": 0.0004726801145450002,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21094982773065568,
      "num_tokens": 39779543.0,
      "step": 17340
    },
    {
      "entropy": 5.22049732208252,
      "epoch": 1.6661863592699326,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004726637384804014,
      "loss": 4.9541,
      "mean_token_accuracy": 0.21696168631315232,
      "num_tokens": 39790827.0,
      "step": 17345
    },
    {
      "entropy": 5.155150604248047,
      "epoch": 1.6666666666666665,
      "grad_norm": 1.25,
      "learning_rate": 0.0004726473578266375,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22023144215345383,
      "num_tokens": 39801995.0,
      "step": 17350
    },
    {
      "entropy": 5.242689180374145,
      "epoch": 1.6671469740634006,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047263097258408893,
      "loss": 5.0201,
      "mean_token_accuracy": 0.21151920258998871,
      "num_tokens": 39813319.0,
      "step": 17355
    },
    {
      "entropy": 5.290485525131226,
      "epoch": 1.6676272814601345,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004726145827531359,
      "loss": 4.9692,
      "mean_token_accuracy": 0.224749793112278,
      "num_tokens": 39824178.0,
      "step": 17360
    },
    {
      "entropy": 5.172863531112671,
      "epoch": 1.6681075888568684,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047259818833415916,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2105391427874565,
      "num_tokens": 39835452.0,
      "step": 17365
    },
    {
      "entropy": 5.180167579650879,
      "epoch": 1.6685878962536023,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047258178932753917,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21109419018030168,
      "num_tokens": 39848083.0,
      "step": 17370
    },
    {
      "entropy": 5.320585584640503,
      "epoch": 1.6690682036503364,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047256538573365675,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20762102603912352,
      "num_tokens": 39859132.0,
      "step": 17375
    },
    {
      "entropy": 5.217416048049927,
      "epoch": 1.6695485110470702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004725489775528928,
      "loss": 4.9952,
      "mean_token_accuracy": 0.21453192532062532,
      "num_tokens": 39870237.0,
      "step": 17380
    },
    {
      "entropy": 5.175440740585327,
      "epoch": 1.6700288184438041,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047253256478562805,
      "loss": 5.0043,
      "mean_token_accuracy": 0.21115336269140245,
      "num_tokens": 39883009.0,
      "step": 17385
    },
    {
      "entropy": 5.252021646499633,
      "epoch": 1.670509125840538,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047251614743224374,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21227517873048782,
      "num_tokens": 39894444.0,
      "step": 17390
    },
    {
      "entropy": 5.320438432693481,
      "epoch": 1.670989433237272,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047249972549312107,
      "loss": 5.0684,
      "mean_token_accuracy": 0.20599901527166367,
      "num_tokens": 39905312.0,
      "step": 17395
    },
    {
      "entropy": 5.265308427810669,
      "epoch": 1.6714697406340058,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004724832989686411,
      "loss": 5.0742,
      "mean_token_accuracy": 0.2087402045726776,
      "num_tokens": 39916724.0,
      "step": 17400
    },
    {
      "entropy": 5.257494497299194,
      "epoch": 1.6719500480307397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047246686785918545,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2121524140238762,
      "num_tokens": 39928919.0,
      "step": 17405
    },
    {
      "entropy": 5.300972175598145,
      "epoch": 1.6724303554274735,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047245043216513546,
      "loss": 5.1434,
      "mean_token_accuracy": 0.20387590378522874,
      "num_tokens": 39940301.0,
      "step": 17410
    },
    {
      "entropy": 5.228685522079468,
      "epoch": 1.6729106628242074,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004724339918868727,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21534408777952194,
      "num_tokens": 39951269.0,
      "step": 17415
    },
    {
      "entropy": 5.299399375915527,
      "epoch": 1.6733909702209413,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000472417547024779,
      "loss": 5.0772,
      "mean_token_accuracy": 0.21561664640903472,
      "num_tokens": 39963057.0,
      "step": 17420
    },
    {
      "entropy": 5.369837093353271,
      "epoch": 1.6738712776176752,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047240109757923593,
      "loss": 5.1016,
      "mean_token_accuracy": 0.2089596852660179,
      "num_tokens": 39973876.0,
      "step": 17425
    },
    {
      "entropy": 5.293364000320435,
      "epoch": 1.674351585014409,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004723846435506256,
      "loss": 5.0574,
      "mean_token_accuracy": 0.2080368861556053,
      "num_tokens": 39984477.0,
      "step": 17430
    },
    {
      "entropy": 5.222610092163086,
      "epoch": 1.6748318924111432,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047236818493932994,
      "loss": 5.023,
      "mean_token_accuracy": 0.21111055314540864,
      "num_tokens": 39996266.0,
      "step": 17435
    },
    {
      "entropy": 5.270339918136597,
      "epoch": 1.675312199807877,
      "grad_norm": 1.25,
      "learning_rate": 0.0004723517217457311,
      "loss": 5.0846,
      "mean_token_accuracy": 0.20994069725275039,
      "num_tokens": 40008618.0,
      "step": 17440
    },
    {
      "entropy": 5.266600751876831,
      "epoch": 1.675792507204611,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004723352539702113,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21478998512029648,
      "num_tokens": 40020290.0,
      "step": 17445
    },
    {
      "entropy": 5.22064790725708,
      "epoch": 1.6762728146013448,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004723187816131529,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2196679472923279,
      "num_tokens": 40031167.0,
      "step": 17450
    },
    {
      "entropy": 5.187341260910034,
      "epoch": 1.676753121998079,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004723023046749383,
      "loss": 5.0127,
      "mean_token_accuracy": 0.21875148564577102,
      "num_tokens": 40043278.0,
      "step": 17455
    },
    {
      "entropy": 5.260714101791382,
      "epoch": 1.6772334293948128,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047228582315595,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21428411453962326,
      "num_tokens": 40055022.0,
      "step": 17460
    },
    {
      "entropy": 5.271437931060791,
      "epoch": 1.6777137367915467,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004722693370565708,
      "loss": 5.0721,
      "mean_token_accuracy": 0.2104582294821739,
      "num_tokens": 40066556.0,
      "step": 17465
    },
    {
      "entropy": 5.36696310043335,
      "epoch": 1.6781940441882806,
      "grad_norm": 1.125,
      "learning_rate": 0.00047225284637718323,
      "loss": 5.1234,
      "mean_token_accuracy": 0.20200545489788055,
      "num_tokens": 40078015.0,
      "step": 17470
    },
    {
      "entropy": 5.225201940536499,
      "epoch": 1.6786743515850144,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004722363511181703,
      "loss": 5.0073,
      "mean_token_accuracy": 0.20932909101247787,
      "num_tokens": 40090061.0,
      "step": 17475
    },
    {
      "entropy": 5.306127595901489,
      "epoch": 1.6791546589817483,
      "grad_norm": 1.15625,
      "learning_rate": 0.000472219851279915,
      "loss": 5.0834,
      "mean_token_accuracy": 0.20633242428302764,
      "num_tokens": 40102569.0,
      "step": 17480
    },
    {
      "entropy": 5.268912601470947,
      "epoch": 1.6796349663784822,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004722033468628004,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21026518642902375,
      "num_tokens": 40114883.0,
      "step": 17485
    },
    {
      "entropy": 5.1828179359436035,
      "epoch": 1.680115273775216,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004721868378672098,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22258042991161348,
      "num_tokens": 40126337.0,
      "step": 17490
    },
    {
      "entropy": 5.189944791793823,
      "epoch": 1.68059558117195,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004721703242935261,
      "loss": 4.9893,
      "mean_token_accuracy": 0.21965228617191315,
      "num_tokens": 40137438.0,
      "step": 17495
    },
    {
      "entropy": 5.236634016036987,
      "epoch": 1.6810758885686838,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004721538061421331,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20900608599185944,
      "num_tokens": 40148061.0,
      "step": 17500
    },
    {
      "entropy": 5.271089839935303,
      "epoch": 1.6815561959654177,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047213728341341407,
      "loss": 5.0414,
      "mean_token_accuracy": 0.2093170017004013,
      "num_tokens": 40159118.0,
      "step": 17505
    },
    {
      "entropy": 5.201540899276734,
      "epoch": 1.6820365033621518,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004721207561077527,
      "loss": 4.9753,
      "mean_token_accuracy": 0.21675143241882325,
      "num_tokens": 40170027.0,
      "step": 17510
    },
    {
      "entropy": 5.1682991027832035,
      "epoch": 1.6825168107588857,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004721042242255327,
      "loss": 5.003,
      "mean_token_accuracy": 0.21124687492847444,
      "num_tokens": 40181819.0,
      "step": 17515
    },
    {
      "entropy": 5.2825909614562985,
      "epoch": 1.6829971181556196,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047208768776713805,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2125624895095825,
      "num_tokens": 40191870.0,
      "step": 17520
    },
    {
      "entropy": 5.240515184402466,
      "epoch": 1.6834774255523535,
      "grad_norm": 1.25,
      "learning_rate": 0.0004720711467329523,
      "loss": 5.0091,
      "mean_token_accuracy": 0.21227465122938155,
      "num_tokens": 40203191.0,
      "step": 17525
    },
    {
      "entropy": 5.325783109664917,
      "epoch": 1.6839577329490876,
      "grad_norm": 1.5,
      "learning_rate": 0.0004720546011233599,
      "loss": 5.1616,
      "mean_token_accuracy": 0.202537639439106,
      "num_tokens": 40214217.0,
      "step": 17530
    },
    {
      "entropy": 5.221359300613403,
      "epoch": 1.6844380403458215,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004720380509387446,
      "loss": 5.0378,
      "mean_token_accuracy": 0.2149007111787796,
      "num_tokens": 40225711.0,
      "step": 17535
    },
    {
      "entropy": 5.341517686843872,
      "epoch": 1.6849183477425553,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000472021496179491,
      "loss": 5.1473,
      "mean_token_accuracy": 0.20318531692028047,
      "num_tokens": 40235748.0,
      "step": 17540
    },
    {
      "entropy": 5.31803822517395,
      "epoch": 1.6853986551392892,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047200493684598316,
      "loss": 5.0912,
      "mean_token_accuracy": 0.20382725894451142,
      "num_tokens": 40246090.0,
      "step": 17545
    },
    {
      "entropy": 5.274109649658203,
      "epoch": 1.685878962536023,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047198837293860573,
      "loss": 5.0826,
      "mean_token_accuracy": 0.21297947317361832,
      "num_tokens": 40257709.0,
      "step": 17550
    },
    {
      "entropy": 5.277770376205444,
      "epoch": 1.686359269932757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004719718044577432,
      "loss": 5.0209,
      "mean_token_accuracy": 0.21575209349393845,
      "num_tokens": 40268670.0,
      "step": 17555
    },
    {
      "entropy": 5.245500278472901,
      "epoch": 1.6868395773294909,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047195523140378034,
      "loss": 4.9401,
      "mean_token_accuracy": 0.219527992606163,
      "num_tokens": 40279289.0,
      "step": 17560
    },
    {
      "entropy": 5.276642560958862,
      "epoch": 1.6873198847262247,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047193865377710177,
      "loss": 5.1554,
      "mean_token_accuracy": 0.1988054856657982,
      "num_tokens": 40289953.0,
      "step": 17565
    },
    {
      "entropy": 5.332764911651611,
      "epoch": 1.6878001921229586,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047192207157809246,
      "loss": 5.0775,
      "mean_token_accuracy": 0.20764612406492233,
      "num_tokens": 40302180.0,
      "step": 17570
    },
    {
      "entropy": 5.20483808517456,
      "epoch": 1.6882804995196925,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047190548480713736,
      "loss": 4.9308,
      "mean_token_accuracy": 0.21693507581949234,
      "num_tokens": 40312556.0,
      "step": 17575
    },
    {
      "entropy": 5.223858451843261,
      "epoch": 1.6887608069164264,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047188889346462163,
      "loss": 5.0425,
      "mean_token_accuracy": 0.21080678850412368,
      "num_tokens": 40323213.0,
      "step": 17580
    },
    {
      "entropy": 5.23965015411377,
      "epoch": 1.6892411143131603,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047187229755093037,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20284025222063065,
      "num_tokens": 40334567.0,
      "step": 17585
    },
    {
      "entropy": 5.3243879795074465,
      "epoch": 1.6897214217098944,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000471855697066449,
      "loss": 5.0715,
      "mean_token_accuracy": 0.2066202938556671,
      "num_tokens": 40345606.0,
      "step": 17590
    },
    {
      "entropy": 5.272208976745605,
      "epoch": 1.6902017291066282,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047183909201156297,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21176956295967103,
      "num_tokens": 40356366.0,
      "step": 17595
    },
    {
      "entropy": 5.157226800918579,
      "epoch": 1.6906820365033621,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004718224823866576,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21795445084571838,
      "num_tokens": 40367451.0,
      "step": 17600
    },
    {
      "entropy": 5.149046325683594,
      "epoch": 1.691162343900096,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004718058681921186,
      "loss": 4.9076,
      "mean_token_accuracy": 0.21818415224552154,
      "num_tokens": 40379377.0,
      "step": 17605
    },
    {
      "entropy": 5.26991753578186,
      "epoch": 1.6916426512968301,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047178924942833185,
      "loss": 5.0503,
      "mean_token_accuracy": 0.2108687251806259,
      "num_tokens": 40390615.0,
      "step": 17610
    },
    {
      "entropy": 5.273028469085693,
      "epoch": 1.692122958693564,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004717726260956831,
      "loss": 5.0624,
      "mean_token_accuracy": 0.20707659423351288,
      "num_tokens": 40402475.0,
      "step": 17615
    },
    {
      "entropy": 5.283695554733276,
      "epoch": 1.6926032660902979,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004717559981945581,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2055098906159401,
      "num_tokens": 40414311.0,
      "step": 17620
    },
    {
      "entropy": 5.205201292037964,
      "epoch": 1.6930835734870318,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004717393657253432,
      "loss": 4.9417,
      "mean_token_accuracy": 0.21965151876211167,
      "num_tokens": 40425859.0,
      "step": 17625
    },
    {
      "entropy": 5.309840250015259,
      "epoch": 1.6935638808837656,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004717227286884243,
      "loss": 5.0699,
      "mean_token_accuracy": 0.2065918892621994,
      "num_tokens": 40436679.0,
      "step": 17630
    },
    {
      "entropy": 5.32344651222229,
      "epoch": 1.6940441882804995,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004717060870841879,
      "loss": 5.1046,
      "mean_token_accuracy": 0.20305901765823364,
      "num_tokens": 40446985.0,
      "step": 17635
    },
    {
      "entropy": 5.216515922546387,
      "epoch": 1.6945244956772334,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004716894409130202,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21672162264585496,
      "num_tokens": 40458777.0,
      "step": 17640
    },
    {
      "entropy": 5.2992652416229244,
      "epoch": 1.6950048030739673,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004716727901753078,
      "loss": 5.0896,
      "mean_token_accuracy": 0.201206973195076,
      "num_tokens": 40469493.0,
      "step": 17645
    },
    {
      "entropy": 5.2455024242401125,
      "epoch": 1.6954851104707012,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004716561348714371,
      "loss": 5.1035,
      "mean_token_accuracy": 0.21021593511104583,
      "num_tokens": 40480889.0,
      "step": 17650
    },
    {
      "entropy": 5.223057603836059,
      "epoch": 1.695965417867435,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047163947500179494,
      "loss": 4.9224,
      "mean_token_accuracy": 0.21474961191415787,
      "num_tokens": 40491630.0,
      "step": 17655
    },
    {
      "entropy": 5.406704521179199,
      "epoch": 1.696445725264169,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004716228105667681,
      "loss": 5.1808,
      "mean_token_accuracy": 0.19069262892007827,
      "num_tokens": 40503204.0,
      "step": 17660
    },
    {
      "entropy": 5.192758703231812,
      "epoch": 1.696926032660903,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004716061415667435,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21487909853458403,
      "num_tokens": 40513917.0,
      "step": 17665
    },
    {
      "entropy": 5.292790699005127,
      "epoch": 1.697406340057637,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000471589468002108,
      "loss": 5.0954,
      "mean_token_accuracy": 0.20588869899511336,
      "num_tokens": 40526501.0,
      "step": 17670
    },
    {
      "entropy": 5.234438419342041,
      "epoch": 1.6978866474543708,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004715727898732488,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21391747444868087,
      "num_tokens": 40536641.0,
      "step": 17675
    },
    {
      "entropy": 5.264958715438842,
      "epoch": 1.6983669548511047,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047155610718055315,
      "loss": 5.021,
      "mean_token_accuracy": 0.2182985797524452,
      "num_tokens": 40547497.0,
      "step": 17680
    },
    {
      "entropy": 5.216670370101928,
      "epoch": 1.6988472622478388,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047153941992440833,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21291295886039735,
      "num_tokens": 40559212.0,
      "step": 17685
    },
    {
      "entropy": 5.197024393081665,
      "epoch": 1.6993275696445727,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004715227281052018,
      "loss": 4.9481,
      "mean_token_accuracy": 0.21620489060878753,
      "num_tokens": 40571692.0,
      "step": 17690
    },
    {
      "entropy": 5.250081729888916,
      "epoch": 1.6998078770413065,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004715060317233211,
      "loss": 5.0272,
      "mean_token_accuracy": 0.21214864552021026,
      "num_tokens": 40583751.0,
      "step": 17695
    },
    {
      "entropy": 5.3393310546875,
      "epoch": 1.7002881844380404,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004714893307791538,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21281693428754805,
      "num_tokens": 40594782.0,
      "step": 17700
    },
    {
      "entropy": 5.241069793701172,
      "epoch": 1.7007684918347743,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047147262527308766,
      "loss": 4.9968,
      "mean_token_accuracy": 0.21480080038309096,
      "num_tokens": 40604583.0,
      "step": 17705
    },
    {
      "entropy": 5.19742112159729,
      "epoch": 1.7012487992315082,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004714559152055106,
      "loss": 5.053,
      "mean_token_accuracy": 0.20500172376632692,
      "num_tokens": 40617957.0,
      "step": 17710
    },
    {
      "entropy": 5.266496276855468,
      "epoch": 1.701729106628242,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004714392005768106,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2177741691470146,
      "num_tokens": 40628016.0,
      "step": 17715
    },
    {
      "entropy": 5.358376169204712,
      "epoch": 1.702209414024976,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004714224813873756,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20627815425395965,
      "num_tokens": 40639339.0,
      "step": 17720
    },
    {
      "entropy": 5.320933866500854,
      "epoch": 1.7026897214217098,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047140575763759393,
      "loss": 5.115,
      "mean_token_accuracy": 0.2098211646080017,
      "num_tokens": 40651634.0,
      "step": 17725
    },
    {
      "entropy": 5.2674033641815186,
      "epoch": 1.7031700288184437,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047138902932785363,
      "loss": 5.0811,
      "mean_token_accuracy": 0.204517363011837,
      "num_tokens": 40662742.0,
      "step": 17730
    },
    {
      "entropy": 5.277242374420166,
      "epoch": 1.7036503362151776,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047137229645854333,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20914596915245057,
      "num_tokens": 40673586.0,
      "step": 17735
    },
    {
      "entropy": 5.225255346298217,
      "epoch": 1.7041306436119115,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004713555590300513,
      "loss": 4.911,
      "mean_token_accuracy": 0.2164280667901039,
      "num_tokens": 40684238.0,
      "step": 17740
    },
    {
      "entropy": 5.258769989013672,
      "epoch": 1.7046109510086456,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004713388170427664,
      "loss": 5.1527,
      "mean_token_accuracy": 0.20387216806411743,
      "num_tokens": 40696154.0,
      "step": 17745
    },
    {
      "entropy": 5.314235210418701,
      "epoch": 1.7050912584053795,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004713220704970771,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21217306852340698,
      "num_tokens": 40708328.0,
      "step": 17750
    },
    {
      "entropy": 5.331616830825806,
      "epoch": 1.7055715658021133,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047130531939337236,
      "loss": 5.1571,
      "mean_token_accuracy": 0.2081201121211052,
      "num_tokens": 40719067.0,
      "step": 17755
    },
    {
      "entropy": 5.119943857192993,
      "epoch": 1.7060518731988472,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047128856373204086,
      "loss": 4.8382,
      "mean_token_accuracy": 0.22838073670864106,
      "num_tokens": 40730638.0,
      "step": 17760
    },
    {
      "entropy": 5.263867521286011,
      "epoch": 1.7065321805955813,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047127180351347184,
      "loss": 5.0209,
      "mean_token_accuracy": 0.21050103157758712,
      "num_tokens": 40743245.0,
      "step": 17765
    },
    {
      "entropy": 5.285775947570801,
      "epoch": 1.7070124879923152,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004712550387380544,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2135525807738304,
      "num_tokens": 40753769.0,
      "step": 17770
    },
    {
      "entropy": 5.186306810379028,
      "epoch": 1.707492795389049,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004712382694061776,
      "loss": 5.0484,
      "mean_token_accuracy": 0.21187084764242173,
      "num_tokens": 40765470.0,
      "step": 17775
    },
    {
      "entropy": 5.244655656814575,
      "epoch": 1.707973102785783,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047122149551823096,
      "loss": 5.0612,
      "mean_token_accuracy": 0.2029837876558304,
      "num_tokens": 40776427.0,
      "step": 17780
    },
    {
      "entropy": 5.285497570037842,
      "epoch": 1.7084534101825168,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004712047170746039,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20866076946258544,
      "num_tokens": 40788707.0,
      "step": 17785
    },
    {
      "entropy": 5.357574796676635,
      "epoch": 1.7089337175792507,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047118793407568586,
      "loss": 5.1458,
      "mean_token_accuracy": 0.20369352102279664,
      "num_tokens": 40800092.0,
      "step": 17790
    },
    {
      "entropy": 5.1695140361785885,
      "epoch": 1.7094140249759846,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047117114652186657,
      "loss": 4.9552,
      "mean_token_accuracy": 0.2189222738146782,
      "num_tokens": 40811389.0,
      "step": 17795
    },
    {
      "entropy": 5.218140363693237,
      "epoch": 1.7098943323727185,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047115435441353573,
      "loss": 5.004,
      "mean_token_accuracy": 0.21120177507400512,
      "num_tokens": 40821438.0,
      "step": 17800
    },
    {
      "entropy": 5.214805936813354,
      "epoch": 1.7103746397694524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047113755775108333,
      "loss": 4.9885,
      "mean_token_accuracy": 0.21069856733083725,
      "num_tokens": 40833335.0,
      "step": 17805
    },
    {
      "entropy": 5.358754396438599,
      "epoch": 1.7108549471661862,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047112075653489913,
      "loss": 5.1276,
      "mean_token_accuracy": 0.21213698983192444,
      "num_tokens": 40844090.0,
      "step": 17810
    },
    {
      "entropy": 5.272863388061523,
      "epoch": 1.7113352545629201,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004711039507653734,
      "loss": 5.0198,
      "mean_token_accuracy": 0.20617565363645554,
      "num_tokens": 40855176.0,
      "step": 17815
    },
    {
      "entropy": 5.268869543075562,
      "epoch": 1.7118155619596542,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004710871404428961,
      "loss": 5.1348,
      "mean_token_accuracy": 0.21078982949256897,
      "num_tokens": 40866190.0,
      "step": 17820
    },
    {
      "entropy": 5.24475212097168,
      "epoch": 1.7122958693563881,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047107032556785786,
      "loss": 4.9848,
      "mean_token_accuracy": 0.20978621393442154,
      "num_tokens": 40876273.0,
      "step": 17825
    },
    {
      "entropy": 5.2347798347473145,
      "epoch": 1.712776176753122,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047105350614064874,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22042171210050582,
      "num_tokens": 40887843.0,
      "step": 17830
    },
    {
      "entropy": 5.252465105056762,
      "epoch": 1.7132564841498559,
      "grad_norm": 1.375,
      "learning_rate": 0.00047103668216165944,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20834531038999557,
      "num_tokens": 40898648.0,
      "step": 17835
    },
    {
      "entropy": 5.188456249237061,
      "epoch": 1.71373679154659,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047101985363128045,
      "loss": 4.9952,
      "mean_token_accuracy": 0.21525215953588486,
      "num_tokens": 40910817.0,
      "step": 17840
    },
    {
      "entropy": 5.291696405410766,
      "epoch": 1.7142170989433239,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047100302054990255,
      "loss": 5.0719,
      "mean_token_accuracy": 0.21371753960847856,
      "num_tokens": 40922520.0,
      "step": 17845
    },
    {
      "entropy": 5.286196374893189,
      "epoch": 1.7146974063400577,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004709861829179165,
      "loss": 5.0549,
      "mean_token_accuracy": 0.21199713200330733,
      "num_tokens": 40934188.0,
      "step": 17850
    },
    {
      "entropy": 5.269313764572144,
      "epoch": 1.7151777137367916,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047096934073571325,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21290006190538407,
      "num_tokens": 40944520.0,
      "step": 17855
    },
    {
      "entropy": 5.3243907451629635,
      "epoch": 1.7156580211335255,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047095249400368384,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1974412590265274,
      "num_tokens": 40955874.0,
      "step": 17860
    },
    {
      "entropy": 5.269866323471069,
      "epoch": 1.7161383285302594,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047093564272221927,
      "loss": 5.0324,
      "mean_token_accuracy": 0.21494160294532777,
      "num_tokens": 40966697.0,
      "step": 17865
    },
    {
      "entropy": 5.192908763885498,
      "epoch": 1.7166186359269933,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047091878689171105,
      "loss": 5.0157,
      "mean_token_accuracy": 0.213586288690567,
      "num_tokens": 40978107.0,
      "step": 17870
    },
    {
      "entropy": 5.325576591491699,
      "epoch": 1.7170989433237271,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004709019265125502,
      "loss": 5.0314,
      "mean_token_accuracy": 0.20897511690855025,
      "num_tokens": 40991535.0,
      "step": 17875
    },
    {
      "entropy": 5.331446743011474,
      "epoch": 1.717579250720461,
      "grad_norm": 1.25,
      "learning_rate": 0.00047088506158512837,
      "loss": 5.0464,
      "mean_token_accuracy": 0.2134338527917862,
      "num_tokens": 41001333.0,
      "step": 17880
    },
    {
      "entropy": 5.152120304107666,
      "epoch": 1.718059558117195,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047086819210983714,
      "loss": 4.9158,
      "mean_token_accuracy": 0.2208703339099884,
      "num_tokens": 41012854.0,
      "step": 17885
    },
    {
      "entropy": 5.165986490249634,
      "epoch": 1.7185398655139288,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047085131808706813,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21885813176631927,
      "num_tokens": 41024041.0,
      "step": 17890
    },
    {
      "entropy": 5.292533206939697,
      "epoch": 1.7190201729106627,
      "grad_norm": 1.375,
      "learning_rate": 0.0004708344395172129,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21375389844179155,
      "num_tokens": 41034788.0,
      "step": 17895
    },
    {
      "entropy": 5.306055164337158,
      "epoch": 1.7195004803073968,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004708175564006636,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21829918324947356,
      "num_tokens": 41046513.0,
      "step": 17900
    },
    {
      "entropy": 5.348610877990723,
      "epoch": 1.7199807877041307,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004708006687378121,
      "loss": 5.1466,
      "mean_token_accuracy": 0.19843295067548752,
      "num_tokens": 41059054.0,
      "step": 17905
    },
    {
      "entropy": 5.291535520553589,
      "epoch": 1.7204610951008645,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004707837765290505,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21096409112215042,
      "num_tokens": 41070171.0,
      "step": 17910
    },
    {
      "entropy": 5.204817867279052,
      "epoch": 1.7209414024975984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004707668797747709,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21286309212446214,
      "num_tokens": 41080963.0,
      "step": 17915
    },
    {
      "entropy": 5.261023092269897,
      "epoch": 1.7214217098943325,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004707499784753657,
      "loss": 5.062,
      "mean_token_accuracy": 0.21039356887340546,
      "num_tokens": 41092658.0,
      "step": 17920
    },
    {
      "entropy": 5.292727279663086,
      "epoch": 1.7219020172910664,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004707330726312273,
      "loss": 5.0828,
      "mean_token_accuracy": 0.20666903555393218,
      "num_tokens": 41104521.0,
      "step": 17925
    },
    {
      "entropy": 5.2403936862945555,
      "epoch": 1.7223823246878003,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047071616224274803,
      "loss": 5.029,
      "mean_token_accuracy": 0.21640972346067427,
      "num_tokens": 41116420.0,
      "step": 17930
    },
    {
      "entropy": 5.225432586669922,
      "epoch": 1.7228626320845342,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004706992473103207,
      "loss": 5.0082,
      "mean_token_accuracy": 0.2125321075320244,
      "num_tokens": 41127747.0,
      "step": 17935
    },
    {
      "entropy": 5.247241687774658,
      "epoch": 1.723342939481268,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047068232783433806,
      "loss": 4.987,
      "mean_token_accuracy": 0.2121317133307457,
      "num_tokens": 41138498.0,
      "step": 17940
    },
    {
      "entropy": 5.290298414230347,
      "epoch": 1.723823246878002,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004706654038151927,
      "loss": 5.0887,
      "mean_token_accuracy": 0.20892859995365143,
      "num_tokens": 41149781.0,
      "step": 17945
    },
    {
      "entropy": 5.279216003417969,
      "epoch": 1.7243035542747358,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004706484752532777,
      "loss": 5.023,
      "mean_token_accuracy": 0.21500371694564818,
      "num_tokens": 41160421.0,
      "step": 17950
    },
    {
      "entropy": 5.2552684307098385,
      "epoch": 1.7247838616714697,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004706315421489861,
      "loss": 5.0118,
      "mean_token_accuracy": 0.21512119770050048,
      "num_tokens": 41171192.0,
      "step": 17955
    },
    {
      "entropy": 5.255807876586914,
      "epoch": 1.7252641690682036,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004706146045027109,
      "loss": 4.9958,
      "mean_token_accuracy": 0.22131290435791015,
      "num_tokens": 41182396.0,
      "step": 17960
    },
    {
      "entropy": 5.292353725433349,
      "epoch": 1.7257444764649374,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004705976623148455,
      "loss": 5.1396,
      "mean_token_accuracy": 0.1990826591849327,
      "num_tokens": 41194939.0,
      "step": 17965
    },
    {
      "entropy": 5.27990689277649,
      "epoch": 1.7262247838616713,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047058071558578324,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20766075998544692,
      "num_tokens": 41206670.0,
      "step": 17970
    },
    {
      "entropy": 5.344282770156861,
      "epoch": 1.7267050912584054,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004705637643159175,
      "loss": 5.2273,
      "mean_token_accuracy": 0.2059547170996666,
      "num_tokens": 41220200.0,
      "step": 17975
    },
    {
      "entropy": 5.278282165527344,
      "epoch": 1.7271853986551393,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047054680850564185,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2120529979467392,
      "num_tokens": 41231697.0,
      "step": 17980
    },
    {
      "entropy": 5.234078073501587,
      "epoch": 1.7276657060518732,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004705298481553499,
      "loss": 5.026,
      "mean_token_accuracy": 0.20734040886163713,
      "num_tokens": 41243835.0,
      "step": 17985
    },
    {
      "entropy": 5.291547107696533,
      "epoch": 1.728146013448607,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047051288326543553,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21486423164606094,
      "num_tokens": 41255147.0,
      "step": 17990
    },
    {
      "entropy": 5.417861127853394,
      "epoch": 1.7286263208453412,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047049591383629247,
      "loss": 5.1242,
      "mean_token_accuracy": 0.19683542996644973,
      "num_tokens": 41267863.0,
      "step": 17995
    },
    {
      "entropy": 5.305896663665772,
      "epoch": 1.729106628242075,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047047893986831493,
      "loss": 5.0426,
      "mean_token_accuracy": 0.2066734328866005,
      "num_tokens": 41278506.0,
      "step": 18000
    },
    {
      "epoch": 1.729106628242075,
      "eval_entropy": 5.044664518360043,
      "eval_loss": 5.099826812744141,
      "eval_mean_token_accuracy": 0.21678114187304767,
      "eval_num_tokens": 41278506.0,
      "eval_runtime": 26.494,
      "eval_samples_per_second": 1238.582,
      "eval_steps_per_second": 154.827,
      "step": 18000
    },
    {
      "entropy": 5.241024541854858,
      "epoch": 1.729586935638809,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047046196136189686,
      "loss": 5.0376,
      "mean_token_accuracy": 0.21125493049621583,
      "num_tokens": 41289825.0,
      "step": 18005
    },
    {
      "entropy": 5.237089443206787,
      "epoch": 1.7300672430355428,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004704449783174323,
      "loss": 4.9988,
      "mean_token_accuracy": 0.21374676674604415,
      "num_tokens": 41300451.0,
      "step": 18010
    },
    {
      "entropy": 5.282814168930054,
      "epoch": 1.7305475504322767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004704279907353158,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21976561099290848,
      "num_tokens": 41312762.0,
      "step": 18015
    },
    {
      "entropy": 5.21120810508728,
      "epoch": 1.7310278578290106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047041099861594167,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21858959794044494,
      "num_tokens": 41323464.0,
      "step": 18020
    },
    {
      "entropy": 5.244153165817261,
      "epoch": 1.7315081652257445,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004703940019597044,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21283236593008042,
      "num_tokens": 41335361.0,
      "step": 18025
    },
    {
      "entropy": 5.271466875076294,
      "epoch": 1.7319884726224783,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047037700076699857,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21386536210775375,
      "num_tokens": 41345353.0,
      "step": 18030
    },
    {
      "entropy": 5.256780052185059,
      "epoch": 1.7324687800192122,
      "grad_norm": 1.375,
      "learning_rate": 0.000470359995038219,
      "loss": 5.0901,
      "mean_token_accuracy": 0.21522050946950913,
      "num_tokens": 41356578.0,
      "step": 18035
    },
    {
      "entropy": 5.3121990203857425,
      "epoch": 1.732949087415946,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004703429847737604,
      "loss": 5.0503,
      "mean_token_accuracy": 0.21347695142030715,
      "num_tokens": 41367668.0,
      "step": 18040
    },
    {
      "entropy": 5.273126173019409,
      "epoch": 1.73342939481268,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004703259699740177,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21362841129302979,
      "num_tokens": 41379746.0,
      "step": 18045
    },
    {
      "entropy": 5.3077721118927,
      "epoch": 1.7339097022094139,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047030895063938607,
      "loss": 5.0979,
      "mean_token_accuracy": 0.20613100081682206,
      "num_tokens": 41390251.0,
      "step": 18050
    },
    {
      "entropy": 5.32034330368042,
      "epoch": 1.734390009606148,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047029192677026043,
      "loss": 5.0011,
      "mean_token_accuracy": 0.21293154060840608,
      "num_tokens": 41401689.0,
      "step": 18055
    },
    {
      "entropy": 5.332354164123535,
      "epoch": 1.7348703170028819,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004702748983670363,
      "loss": 5.0383,
      "mean_token_accuracy": 0.20695865601301194,
      "num_tokens": 41414001.0,
      "step": 18060
    },
    {
      "entropy": 5.3374395847320555,
      "epoch": 1.7353506243996157,
      "grad_norm": 1.375,
      "learning_rate": 0.0004702578654301088,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20650572776794435,
      "num_tokens": 41426012.0,
      "step": 18065
    },
    {
      "entropy": 5.32809739112854,
      "epoch": 1.7358309317963496,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004702408279598734,
      "loss": 5.1124,
      "mean_token_accuracy": 0.20945894420146943,
      "num_tokens": 41436799.0,
      "step": 18070
    },
    {
      "entropy": 5.194957971572876,
      "epoch": 1.7363112391930837,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004702237859567258,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2248876303434372,
      "num_tokens": 41448802.0,
      "step": 18075
    },
    {
      "entropy": 5.159519338607788,
      "epoch": 1.7367915465898176,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004702067394210616,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22305767834186555,
      "num_tokens": 41459576.0,
      "step": 18080
    },
    {
      "entropy": 5.317985725402832,
      "epoch": 1.7372718539865515,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047018968835327643,
      "loss": 5.0528,
      "mean_token_accuracy": 0.20701833814382553,
      "num_tokens": 41471105.0,
      "step": 18085
    },
    {
      "entropy": 5.3002800941467285,
      "epoch": 1.7377521613832854,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004701726327537664,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21483348906040192,
      "num_tokens": 41482747.0,
      "step": 18090
    },
    {
      "entropy": 5.265456676483154,
      "epoch": 1.7382324687800192,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004701555726229274,
      "loss": 5.0478,
      "mean_token_accuracy": 0.20403186678886415,
      "num_tokens": 41494268.0,
      "step": 18095
    },
    {
      "entropy": 5.224048185348511,
      "epoch": 1.7387127761767531,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004701385079611555,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21645985245704652,
      "num_tokens": 41506144.0,
      "step": 18100
    },
    {
      "entropy": 5.212440347671508,
      "epoch": 1.739193083573487,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047012143876884677,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22631471157073973,
      "num_tokens": 41518165.0,
      "step": 18105
    },
    {
      "entropy": 5.2689821243286135,
      "epoch": 1.739673390970221,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004701043650463977,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21341103613376616,
      "num_tokens": 41529644.0,
      "step": 18110
    },
    {
      "entropy": 5.197230815887451,
      "epoch": 1.7401536983669548,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004700872867942046,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21980302035808563,
      "num_tokens": 41540540.0,
      "step": 18115
    },
    {
      "entropy": 5.237172222137451,
      "epoch": 1.7406340057636887,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000470070204012664,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20738894492387772,
      "num_tokens": 41551937.0,
      "step": 18120
    },
    {
      "entropy": 5.245754861831665,
      "epoch": 1.7411143131604225,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047005311670217256,
      "loss": 4.9382,
      "mean_token_accuracy": 0.22406237870454787,
      "num_tokens": 41561630.0,
      "step": 18125
    },
    {
      "entropy": 5.199283027648926,
      "epoch": 1.7415946205571564,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047003602486312687,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22160024642944337,
      "num_tokens": 41573152.0,
      "step": 18130
    },
    {
      "entropy": 5.296490097045899,
      "epoch": 1.7420749279538905,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004700189284959238,
      "loss": 5.0624,
      "mean_token_accuracy": 0.20071677267551422,
      "num_tokens": 41585280.0,
      "step": 18135
    },
    {
      "entropy": 5.292803955078125,
      "epoch": 1.7425552353506244,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047000182760096037,
      "loss": 4.9783,
      "mean_token_accuracy": 0.21720946133136748,
      "num_tokens": 41596308.0,
      "step": 18140
    },
    {
      "entropy": 5.201381254196167,
      "epoch": 1.7430355427473583,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004699847221786335,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21344539821147918,
      "num_tokens": 41607479.0,
      "step": 18145
    },
    {
      "entropy": 5.169917058944702,
      "epoch": 1.7435158501440924,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004699676122293403,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21237257421016692,
      "num_tokens": 41619832.0,
      "step": 18150
    },
    {
      "entropy": 5.322282409667968,
      "epoch": 1.7439961575408263,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004699504977534782,
      "loss": 5.1378,
      "mean_token_accuracy": 0.2060973674058914,
      "num_tokens": 41631715.0,
      "step": 18155
    },
    {
      "entropy": 5.326171875,
      "epoch": 1.7444764649375601,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004699333787514444,
      "loss": 5.0761,
      "mean_token_accuracy": 0.2093367874622345,
      "num_tokens": 41643454.0,
      "step": 18160
    },
    {
      "entropy": 5.267477893829346,
      "epoch": 1.744956772334294,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004699162552236363,
      "loss": 5.0356,
      "mean_token_accuracy": 0.21422340720891953,
      "num_tokens": 41655319.0,
      "step": 18165
    },
    {
      "entropy": 5.278668975830078,
      "epoch": 1.745437079731028,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046989912717045165,
      "loss": 5.0355,
      "mean_token_accuracy": 0.21109535843133925,
      "num_tokens": 41667082.0,
      "step": 18170
    },
    {
      "entropy": 5.29349799156189,
      "epoch": 1.7459173871277618,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046988199459228793,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21627289205789565,
      "num_tokens": 41678281.0,
      "step": 18175
    },
    {
      "entropy": 5.219664812088013,
      "epoch": 1.7463976945244957,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004698648574895429,
      "loss": 4.9975,
      "mean_token_accuracy": 0.21584915816783906,
      "num_tokens": 41690596.0,
      "step": 18180
    },
    {
      "entropy": 5.2811610221862795,
      "epoch": 1.7468780019212296,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046984771586261465,
      "loss": 5.0373,
      "mean_token_accuracy": 0.2115631580352783,
      "num_tokens": 41702441.0,
      "step": 18185
    },
    {
      "entropy": 5.24390025138855,
      "epoch": 1.7473583093179634,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000469830569711901,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21421640515327453,
      "num_tokens": 41714235.0,
      "step": 18190
    },
    {
      "entropy": 5.250346612930298,
      "epoch": 1.7478386167146973,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004698134190377999,
      "loss": 5.0302,
      "mean_token_accuracy": 0.2200736716389656,
      "num_tokens": 41725870.0,
      "step": 18195
    },
    {
      "entropy": 5.257828950881958,
      "epoch": 1.7483189241114312,
      "grad_norm": 1.125,
      "learning_rate": 0.00046979626384070983,
      "loss": 4.9786,
      "mean_token_accuracy": 0.2204352781176567,
      "num_tokens": 41736500.0,
      "step": 18200
    },
    {
      "entropy": 5.206278944015503,
      "epoch": 1.748799231508165,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000469779104121029,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2161783680319786,
      "num_tokens": 41748466.0,
      "step": 18205
    },
    {
      "entropy": 5.299764156341553,
      "epoch": 1.7492795389048992,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046976193987915553,
      "loss": 5.118,
      "mean_token_accuracy": 0.21007431894540787,
      "num_tokens": 41759728.0,
      "step": 18210
    },
    {
      "entropy": 5.264929389953613,
      "epoch": 1.749759846301633,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004697447711154883,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21152468621730805,
      "num_tokens": 41771628.0,
      "step": 18215
    },
    {
      "entropy": 5.329400539398193,
      "epoch": 1.750240153698367,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046972759783042576,
      "loss": 5.0488,
      "mean_token_accuracy": 0.2120545193552971,
      "num_tokens": 41783224.0,
      "step": 18220
    },
    {
      "entropy": 5.17046890258789,
      "epoch": 1.7507204610951008,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004697104200243666,
      "loss": 4.9762,
      "mean_token_accuracy": 0.2210172802209854,
      "num_tokens": 41794899.0,
      "step": 18225
    },
    {
      "entropy": 5.229033613204956,
      "epoch": 1.751200768491835,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004696932376977096,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21705750674009322,
      "num_tokens": 41806336.0,
      "step": 18230
    },
    {
      "entropy": 5.302998828887939,
      "epoch": 1.7516810758885688,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004696760508508538,
      "loss": 5.044,
      "mean_token_accuracy": 0.21656302511692047,
      "num_tokens": 41817443.0,
      "step": 18235
    },
    {
      "entropy": 5.263115739822387,
      "epoch": 1.7521613832853027,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046965885948419814,
      "loss": 5.0602,
      "mean_token_accuracy": 0.21218584924936296,
      "num_tokens": 41829740.0,
      "step": 18240
    },
    {
      "entropy": 5.220942735671997,
      "epoch": 1.7526416906820366,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004696416635981418,
      "loss": 4.968,
      "mean_token_accuracy": 0.22179571986198426,
      "num_tokens": 41840519.0,
      "step": 18245
    },
    {
      "entropy": 5.1602113246917725,
      "epoch": 1.7531219980787704,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000469624463193084,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2226407930254936,
      "num_tokens": 41850975.0,
      "step": 18250
    },
    {
      "entropy": 5.209735584259033,
      "epoch": 1.7536023054755043,
      "grad_norm": 1.34375,
      "learning_rate": 0.000469607258269424,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21733648031949998,
      "num_tokens": 41861854.0,
      "step": 18255
    },
    {
      "entropy": 5.322196769714355,
      "epoch": 1.7540826128722382,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004695900488275614,
      "loss": 5.0959,
      "mean_token_accuracy": 0.20987047106027604,
      "num_tokens": 41873279.0,
      "step": 18260
    },
    {
      "entropy": 5.3081639289855955,
      "epoch": 1.754562920268972,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004695728348678957,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2102254554629326,
      "num_tokens": 41885545.0,
      "step": 18265
    },
    {
      "entropy": 5.208811044692993,
      "epoch": 1.755043227665706,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004695556163908265,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21886218339204788,
      "num_tokens": 41896168.0,
      "step": 18270
    },
    {
      "entropy": 5.197315120697022,
      "epoch": 1.7555235350624399,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004695383933967536,
      "loss": 4.9677,
      "mean_token_accuracy": 0.22212957590818405,
      "num_tokens": 41907423.0,
      "step": 18275
    },
    {
      "entropy": 5.215720129013062,
      "epoch": 1.7560038424591737,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046952116588607694,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22171466648578644,
      "num_tokens": 41918617.0,
      "step": 18280
    },
    {
      "entropy": 5.225232648849487,
      "epoch": 1.7564841498559076,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004695039338591963,
      "loss": 4.9505,
      "mean_token_accuracy": 0.22022587358951567,
      "num_tokens": 41930235.0,
      "step": 18285
    },
    {
      "entropy": 5.227853727340698,
      "epoch": 1.7569644572526417,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000469486697316512,
      "loss": 5.0327,
      "mean_token_accuracy": 0.21550887674093247,
      "num_tokens": 41941308.0,
      "step": 18290
    },
    {
      "entropy": 5.352170991897583,
      "epoch": 1.7574447646493756,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000469469456258424,
      "loss": 5.1348,
      "mean_token_accuracy": 0.2080013632774353,
      "num_tokens": 41952406.0,
      "step": 18295
    },
    {
      "entropy": 5.321382236480713,
      "epoch": 1.7579250720461095,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004694522106853327,
      "loss": 5.0408,
      "mean_token_accuracy": 0.2021900087594986,
      "num_tokens": 41964642.0,
      "step": 18300
    },
    {
      "entropy": 5.233061075210571,
      "epoch": 1.7584053794428436,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046943496059763845,
      "loss": 5.0279,
      "mean_token_accuracy": 0.210664738714695,
      "num_tokens": 41975761.0,
      "step": 18305
    },
    {
      "entropy": 5.202634382247925,
      "epoch": 1.7588856868395775,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046941770599574176,
      "loss": 4.992,
      "mean_token_accuracy": 0.22007163166999816,
      "num_tokens": 41987596.0,
      "step": 18310
    },
    {
      "entropy": 5.213694095611572,
      "epoch": 1.7593659942363113,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004694004468800433,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2098413571715355,
      "num_tokens": 41999148.0,
      "step": 18315
    },
    {
      "entropy": 5.226025676727295,
      "epoch": 1.7598463016330452,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004693831832509437,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22208794057369233,
      "num_tokens": 42010030.0,
      "step": 18320
    },
    {
      "entropy": 5.266612815856933,
      "epoch": 1.760326609029779,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046936591510884375,
      "loss": 5.0901,
      "mean_token_accuracy": 0.2090602546930313,
      "num_tokens": 42022009.0,
      "step": 18325
    },
    {
      "entropy": 5.30504469871521,
      "epoch": 1.760806916426513,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046934864245414443,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2147595390677452,
      "num_tokens": 42033654.0,
      "step": 18330
    },
    {
      "entropy": 5.223488712310791,
      "epoch": 1.7612872238232469,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046933136528724676,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2140914022922516,
      "num_tokens": 42045521.0,
      "step": 18335
    },
    {
      "entropy": 5.166170501708985,
      "epoch": 1.7617675312199808,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004693140836085518,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21318133771419526,
      "num_tokens": 42057751.0,
      "step": 18340
    },
    {
      "entropy": 5.256892824172974,
      "epoch": 1.7622478386167146,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046929679741846076,
      "loss": 5.0673,
      "mean_token_accuracy": 0.21467916220426558,
      "num_tokens": 42068823.0,
      "step": 18345
    },
    {
      "entropy": 5.252368021011352,
      "epoch": 1.7627281460134485,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046927950671737505,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21481483727693557,
      "num_tokens": 42079856.0,
      "step": 18350
    },
    {
      "entropy": 5.2982253074646,
      "epoch": 1.7632084534101824,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046926221150569617,
      "loss": 5.0068,
      "mean_token_accuracy": 0.20960359126329423,
      "num_tokens": 42090560.0,
      "step": 18355
    },
    {
      "entropy": 5.264878463745117,
      "epoch": 1.7636887608069163,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004692449117838255,
      "loss": 5.0766,
      "mean_token_accuracy": 0.20549671947956086,
      "num_tokens": 42102252.0,
      "step": 18360
    },
    {
      "entropy": 5.305474662780762,
      "epoch": 1.7641690682036504,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004692276075521648,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20453374683856965,
      "num_tokens": 42113679.0,
      "step": 18365
    },
    {
      "entropy": 5.290595054626465,
      "epoch": 1.7646493756003843,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004692102988111158,
      "loss": 5.061,
      "mean_token_accuracy": 0.20987682938575744,
      "num_tokens": 42125919.0,
      "step": 18370
    },
    {
      "entropy": 5.389014863967896,
      "epoch": 1.7651296829971181,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046919298556108023,
      "loss": 5.1834,
      "mean_token_accuracy": 0.20681896656751633,
      "num_tokens": 42138189.0,
      "step": 18375
    },
    {
      "entropy": 5.314851236343384,
      "epoch": 1.765609990393852,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046917566780246036,
      "loss": 5.0852,
      "mean_token_accuracy": 0.21313934475183488,
      "num_tokens": 42150232.0,
      "step": 18380
    },
    {
      "entropy": 5.2676682472229,
      "epoch": 1.7660902977905861,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046915834553565793,
      "loss": 5.0093,
      "mean_token_accuracy": 0.20818854719400406,
      "num_tokens": 42160740.0,
      "step": 18385
    },
    {
      "entropy": 5.244115114212036,
      "epoch": 1.76657060518732,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004691410187610753,
      "loss": 5.0169,
      "mean_token_accuracy": 0.22447435259819032,
      "num_tokens": 42171761.0,
      "step": 18390
    },
    {
      "entropy": 5.210508251190186,
      "epoch": 1.767050912584054,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046912368747911465,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21784312427043914,
      "num_tokens": 42183362.0,
      "step": 18395
    },
    {
      "entropy": 5.332230186462402,
      "epoch": 1.7675312199807878,
      "grad_norm": 1.25,
      "learning_rate": 0.00046910635169017845,
      "loss": 5.1207,
      "mean_token_accuracy": 0.2064368724822998,
      "num_tokens": 42195610.0,
      "step": 18400
    },
    {
      "entropy": 5.267078876495361,
      "epoch": 1.7680115273775217,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004690890113946691,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2114761456847191,
      "num_tokens": 42207489.0,
      "step": 18405
    },
    {
      "entropy": 5.2627387046813965,
      "epoch": 1.7684918347742555,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004690716665929893,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20956328511238098,
      "num_tokens": 42217300.0,
      "step": 18410
    },
    {
      "entropy": 5.297606134414673,
      "epoch": 1.7689721421709894,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046905431728554164,
      "loss": 5.0468,
      "mean_token_accuracy": 0.2151705801486969,
      "num_tokens": 42229031.0,
      "step": 18415
    },
    {
      "entropy": 5.065530586242676,
      "epoch": 1.7694524495677233,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046903696347272894,
      "loss": 4.7799,
      "mean_token_accuracy": 0.23349689096212387,
      "num_tokens": 42239102.0,
      "step": 18420
    },
    {
      "entropy": 5.237277603149414,
      "epoch": 1.7699327569644572,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046901960515495413,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21556922495365144,
      "num_tokens": 42250436.0,
      "step": 18425
    },
    {
      "entropy": 5.2082127094268795,
      "epoch": 1.770413064361191,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004690022423326202,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2214997425675392,
      "num_tokens": 42260544.0,
      "step": 18430
    },
    {
      "entropy": 5.197844409942627,
      "epoch": 1.770893371757925,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004689848750061303,
      "loss": 4.98,
      "mean_token_accuracy": 0.21648634672164918,
      "num_tokens": 42272519.0,
      "step": 18435
    },
    {
      "entropy": 5.247015476226807,
      "epoch": 1.7713736791546588,
      "grad_norm": 1.25,
      "learning_rate": 0.0004689675031758876,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22213560789823533,
      "num_tokens": 42281722.0,
      "step": 18440
    },
    {
      "entropy": 5.241945123672485,
      "epoch": 1.771853986551393,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004689501268422954,
      "loss": 5.0812,
      "mean_token_accuracy": 0.2179307833313942,
      "num_tokens": 42294067.0,
      "step": 18445
    },
    {
      "entropy": 5.2518639087677,
      "epoch": 1.7723342939481268,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046893274600575725,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21819799691438674,
      "num_tokens": 42305547.0,
      "step": 18450
    },
    {
      "entropy": 5.336394596099853,
      "epoch": 1.7728146013448607,
      "grad_norm": 1.25,
      "learning_rate": 0.0004689153606666765,
      "loss": 5.0336,
      "mean_token_accuracy": 0.2030814155936241,
      "num_tokens": 42315557.0,
      "step": 18455
    },
    {
      "entropy": 5.291206932067871,
      "epoch": 1.7732949087415946,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000468897970825457,
      "loss": 5.1022,
      "mean_token_accuracy": 0.2059236541390419,
      "num_tokens": 42327374.0,
      "step": 18460
    },
    {
      "entropy": 5.2196080684661865,
      "epoch": 1.7737752161383287,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046888057648250233,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2209856852889061,
      "num_tokens": 42338889.0,
      "step": 18465
    },
    {
      "entropy": 5.25655689239502,
      "epoch": 1.7742555235350626,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004688631776382164,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21609259247779847,
      "num_tokens": 42348678.0,
      "step": 18470
    },
    {
      "entropy": 5.20499620437622,
      "epoch": 1.7747358309317964,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046884577429300305,
      "loss": 5.0549,
      "mean_token_accuracy": 0.21145761609077454,
      "num_tokens": 42359090.0,
      "step": 18475
    },
    {
      "entropy": 5.254438066482544,
      "epoch": 1.7752161383285303,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004688283664472665,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21039756387472153,
      "num_tokens": 42370947.0,
      "step": 18480
    },
    {
      "entropy": 5.292912817001342,
      "epoch": 1.7756964457252642,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046881095410141084,
      "loss": 5.001,
      "mean_token_accuracy": 0.2172191321849823,
      "num_tokens": 42381597.0,
      "step": 18485
    },
    {
      "entropy": 5.262724494934082,
      "epoch": 1.776176753121998,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046879353725584036,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2171296864748001,
      "num_tokens": 42391877.0,
      "step": 18490
    },
    {
      "entropy": 5.17707257270813,
      "epoch": 1.776657060518732,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046877611591095923,
      "loss": 4.9679,
      "mean_token_accuracy": 0.21665361076593398,
      "num_tokens": 42403916.0,
      "step": 18495
    },
    {
      "entropy": 5.3059779644012455,
      "epoch": 1.7771373679154658,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046875869006717224,
      "loss": 5.0578,
      "mean_token_accuracy": 0.2080937847495079,
      "num_tokens": 42415624.0,
      "step": 18500
    },
    {
      "entropy": 5.258943653106689,
      "epoch": 1.7776176753121997,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046874125972488375,
      "loss": 5.0017,
      "mean_token_accuracy": 0.2123072236776352,
      "num_tokens": 42427228.0,
      "step": 18505
    },
    {
      "entropy": 5.227565097808838,
      "epoch": 1.7780979827089336,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046872382488449853,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20770029127597808,
      "num_tokens": 42440023.0,
      "step": 18510
    },
    {
      "entropy": 5.265961170196533,
      "epoch": 1.7785782901056675,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046870638554642133,
      "loss": 4.9494,
      "mean_token_accuracy": 0.22162888795137406,
      "num_tokens": 42451126.0,
      "step": 18515
    },
    {
      "entropy": 5.211573934555053,
      "epoch": 1.7790585975024016,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046868894171105704,
      "loss": 5.0345,
      "mean_token_accuracy": 0.2076125144958496,
      "num_tokens": 42462975.0,
      "step": 18520
    },
    {
      "entropy": 5.197846698760986,
      "epoch": 1.7795389048991355,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004686714933788107,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2216629281640053,
      "num_tokens": 42474716.0,
      "step": 18525
    },
    {
      "entropy": 5.245202255249024,
      "epoch": 1.7800192122958693,
      "grad_norm": 1.125,
      "learning_rate": 0.0004686540405500873,
      "loss": 4.9889,
      "mean_token_accuracy": 0.2196291744709015,
      "num_tokens": 42484616.0,
      "step": 18530
    },
    {
      "entropy": 5.204309368133545,
      "epoch": 1.7804995196926032,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004686365832252922,
      "loss": 4.9695,
      "mean_token_accuracy": 0.2156553566455841,
      "num_tokens": 42494840.0,
      "step": 18535
    },
    {
      "entropy": 5.238338851928711,
      "epoch": 1.7809798270893373,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046861912140483056,
      "loss": 5.0196,
      "mean_token_accuracy": 0.2081630438566208,
      "num_tokens": 42505957.0,
      "step": 18540
    },
    {
      "entropy": 5.309097671508789,
      "epoch": 1.7814601344860712,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046860165508910787,
      "loss": 5.1717,
      "mean_token_accuracy": 0.19988202601671218,
      "num_tokens": 42517861.0,
      "step": 18545
    },
    {
      "entropy": 5.228794765472412,
      "epoch": 1.781940441882805,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004685841842785296,
      "loss": 4.9349,
      "mean_token_accuracy": 0.212996007502079,
      "num_tokens": 42529986.0,
      "step": 18550
    },
    {
      "entropy": 5.2075221061706545,
      "epoch": 1.782420749279539,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004685667089735014,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21832083761692048,
      "num_tokens": 42541778.0,
      "step": 18555
    },
    {
      "entropy": 5.299496650695801,
      "epoch": 1.7829010566762729,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046854922917442907,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20699781626462938,
      "num_tokens": 42553776.0,
      "step": 18560
    },
    {
      "entropy": 5.270418405532837,
      "epoch": 1.7833813640730067,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004685317448817182,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21399157196283342,
      "num_tokens": 42566277.0,
      "step": 18565
    },
    {
      "entropy": 5.124438762664795,
      "epoch": 1.7838616714697406,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004685142560957751,
      "loss": 4.8653,
      "mean_token_accuracy": 0.21993045806884765,
      "num_tokens": 42576598.0,
      "step": 18570
    },
    {
      "entropy": 5.198130178451538,
      "epoch": 1.7843419788664745,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004684967628170054,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2230650320649147,
      "num_tokens": 42587850.0,
      "step": 18575
    },
    {
      "entropy": 5.17558183670044,
      "epoch": 1.7848222862632084,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046847926504581553,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2166763037443161,
      "num_tokens": 42598754.0,
      "step": 18580
    },
    {
      "entropy": 5.2978668212890625,
      "epoch": 1.7853025936599423,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004684617627826116,
      "loss": 5.0268,
      "mean_token_accuracy": 0.21996672302484513,
      "num_tokens": 42609842.0,
      "step": 18585
    },
    {
      "entropy": 5.257694959640503,
      "epoch": 1.7857829010566761,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004684442560278001,
      "loss": 4.9177,
      "mean_token_accuracy": 0.223434916138649,
      "num_tokens": 42621641.0,
      "step": 18590
    },
    {
      "entropy": 5.190003442764282,
      "epoch": 1.78626320845341,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046842674478178727,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22035654187202453,
      "num_tokens": 42632523.0,
      "step": 18595
    },
    {
      "entropy": 5.288093328475952,
      "epoch": 1.7867435158501441,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004684092290449798,
      "loss": 5.045,
      "mean_token_accuracy": 0.20643867701292037,
      "num_tokens": 42643511.0,
      "step": 18600
    },
    {
      "entropy": 5.262559223175049,
      "epoch": 1.787223823246878,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004683917088177844,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22036872953176498,
      "num_tokens": 42655077.0,
      "step": 18605
    },
    {
      "entropy": 5.261846923828125,
      "epoch": 1.7877041306436119,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004683741841006077,
      "loss": 5.0979,
      "mean_token_accuracy": 0.21102109253406526,
      "num_tokens": 42667454.0,
      "step": 18610
    },
    {
      "entropy": 5.2796289920806885,
      "epoch": 1.7881844380403458,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004683566548938567,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21412864029407502,
      "num_tokens": 42678214.0,
      "step": 18615
    },
    {
      "entropy": 5.1922179698944095,
      "epoch": 1.7886647454370799,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004683391211979383,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21591276079416274,
      "num_tokens": 42689424.0,
      "step": 18620
    },
    {
      "entropy": 5.362273931503296,
      "epoch": 1.7891450528338138,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004683215830132597,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20249850451946258,
      "num_tokens": 42701679.0,
      "step": 18625
    },
    {
      "entropy": 5.245749473571777,
      "epoch": 1.7896253602305476,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046830404034022786,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2156854197382927,
      "num_tokens": 42711931.0,
      "step": 18630
    },
    {
      "entropy": 5.272154808044434,
      "epoch": 1.7901056676272815,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004682864931792502,
      "loss": 5.0744,
      "mean_token_accuracy": 0.21268342882394792,
      "num_tokens": 42723460.0,
      "step": 18635
    },
    {
      "entropy": 5.255848169326782,
      "epoch": 1.7905859750240154,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004682689415307342,
      "loss": 4.9962,
      "mean_token_accuracy": 0.21373503357172013,
      "num_tokens": 42735389.0,
      "step": 18640
    },
    {
      "entropy": 5.2207025527954105,
      "epoch": 1.7910662824207493,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004682513853950872,
      "loss": 4.9738,
      "mean_token_accuracy": 0.2162192150950432,
      "num_tokens": 42747779.0,
      "step": 18645
    },
    {
      "entropy": 5.175315904617309,
      "epoch": 1.7915465898174832,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000468233824772717,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20882033407688141,
      "num_tokens": 42758776.0,
      "step": 18650
    },
    {
      "entropy": 5.401479482650757,
      "epoch": 1.792026897214217,
      "grad_norm": 1.28125,
      "learning_rate": 0.000468216259664031,
      "loss": 5.1244,
      "mean_token_accuracy": 0.2061645820736885,
      "num_tokens": 42770234.0,
      "step": 18655
    },
    {
      "entropy": 5.258845424652099,
      "epoch": 1.792507204610951,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046819869006943727,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22286790311336518,
      "num_tokens": 42782076.0,
      "step": 18660
    },
    {
      "entropy": 5.188581275939941,
      "epoch": 1.7929875120076848,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004681811159893436,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21760196834802628,
      "num_tokens": 42793620.0,
      "step": 18665
    },
    {
      "entropy": 5.269056940078736,
      "epoch": 1.7934678194044187,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046816353742415814,
      "loss": 5.0292,
      "mean_token_accuracy": 0.21824042052030562,
      "num_tokens": 42805125.0,
      "step": 18670
    },
    {
      "entropy": 5.239722347259521,
      "epoch": 1.7939481268011528,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046814595437428885,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21907177418470383,
      "num_tokens": 42815472.0,
      "step": 18675
    },
    {
      "entropy": 5.198154354095459,
      "epoch": 1.7944284341978867,
      "grad_norm": 1.265625,
      "learning_rate": 0.000468128366840144,
      "loss": 5.0301,
      "mean_token_accuracy": 0.21130653619766235,
      "num_tokens": 42826822.0,
      "step": 18680
    },
    {
      "entropy": 5.260516691207886,
      "epoch": 1.7949087415946205,
      "grad_norm": 1.15625,
      "learning_rate": 0.000468110774822132,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2161370187997818,
      "num_tokens": 42839024.0,
      "step": 18685
    },
    {
      "entropy": 5.207973480224609,
      "epoch": 1.7953890489913544,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004680931783206612,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21794952750205993,
      "num_tokens": 42849782.0,
      "step": 18690
    },
    {
      "entropy": 5.210039901733398,
      "epoch": 1.7958693563880885,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046807557733614014,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21662873774766922,
      "num_tokens": 42859792.0,
      "step": 18695
    },
    {
      "entropy": 5.2095317363739015,
      "epoch": 1.7963496637848224,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046805797186897757,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20927114635705948,
      "num_tokens": 42872031.0,
      "step": 18700
    },
    {
      "entropy": 5.241484355926514,
      "epoch": 1.7968299711815563,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046804036191958206,
      "loss": 4.9938,
      "mean_token_accuracy": 0.22549130916595458,
      "num_tokens": 42882830.0,
      "step": 18705
    },
    {
      "entropy": 5.3077081680297855,
      "epoch": 1.7973102785782902,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046802274748836267,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21175695359706878,
      "num_tokens": 42893576.0,
      "step": 18710
    },
    {
      "entropy": 5.181085062026978,
      "epoch": 1.797790585975024,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004680051285757281,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22162724286317825,
      "num_tokens": 42904405.0,
      "step": 18715
    },
    {
      "entropy": 5.195484399795532,
      "epoch": 1.798270893371758,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004679875051820877,
      "loss": 5.0101,
      "mean_token_accuracy": 0.21413963586091994,
      "num_tokens": 42916601.0,
      "step": 18720
    },
    {
      "entropy": 5.257897520065308,
      "epoch": 1.7987512007684918,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004679698773078503,
      "loss": 5.0328,
      "mean_token_accuracy": 0.20682916790246964,
      "num_tokens": 42928977.0,
      "step": 18725
    },
    {
      "entropy": 5.280573225021362,
      "epoch": 1.7992315081652257,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046795224495342554,
      "loss": 4.9337,
      "mean_token_accuracy": 0.2159036874771118,
      "num_tokens": 42940149.0,
      "step": 18730
    },
    {
      "entropy": 5.271610689163208,
      "epoch": 1.7997118155619596,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046793460811922255,
      "loss": 5.099,
      "mean_token_accuracy": 0.20576590150594712,
      "num_tokens": 42952122.0,
      "step": 18735
    },
    {
      "entropy": 5.259262800216675,
      "epoch": 1.8001921229586935,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046791696680565075,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2174858719110489,
      "num_tokens": 42962984.0,
      "step": 18740
    },
    {
      "entropy": 5.292529726028443,
      "epoch": 1.8006724303554273,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046789932101312003,
      "loss": 5.0533,
      "mean_token_accuracy": 0.2015957474708557,
      "num_tokens": 42974409.0,
      "step": 18745
    },
    {
      "entropy": 5.320332193374634,
      "epoch": 1.8011527377521612,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004678816707420397,
      "loss": 5.0842,
      "mean_token_accuracy": 0.20993065237998962,
      "num_tokens": 42986600.0,
      "step": 18750
    },
    {
      "entropy": 5.171487855911255,
      "epoch": 1.8016330451488953,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004678640159928198,
      "loss": 4.939,
      "mean_token_accuracy": 0.2238040491938591,
      "num_tokens": 42998192.0,
      "step": 18755
    },
    {
      "entropy": 5.201213550567627,
      "epoch": 1.8021133525456292,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004678463567658701,
      "loss": 4.9509,
      "mean_token_accuracy": 0.221284182369709,
      "num_tokens": 43009680.0,
      "step": 18760
    },
    {
      "entropy": 5.210364627838135,
      "epoch": 1.802593659942363,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004678286930616006,
      "loss": 4.9681,
      "mean_token_accuracy": 0.21532466113567353,
      "num_tokens": 43021031.0,
      "step": 18765
    },
    {
      "entropy": 5.2414628028869625,
      "epoch": 1.803073967339097,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004678110248804215,
      "loss": 4.9882,
      "mean_token_accuracy": 0.22129576951265334,
      "num_tokens": 43031423.0,
      "step": 18770
    },
    {
      "entropy": 5.200354146957397,
      "epoch": 1.803554274735831,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046779335222274293,
      "loss": 5.0343,
      "mean_token_accuracy": 0.21545153856277466,
      "num_tokens": 43042895.0,
      "step": 18775
    },
    {
      "entropy": 5.263379335403442,
      "epoch": 1.804034582132565,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046777567508897515,
      "loss": 5.1006,
      "mean_token_accuracy": 0.208549465239048,
      "num_tokens": 43053052.0,
      "step": 18780
    },
    {
      "entropy": 5.352109003067016,
      "epoch": 1.8045148895292988,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046775799347952864,
      "loss": 5.0817,
      "mean_token_accuracy": 0.20444247424602507,
      "num_tokens": 43066908.0,
      "step": 18785
    },
    {
      "entropy": 5.263387680053711,
      "epoch": 1.8049951969260327,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004677403073948139,
      "loss": 5.0741,
      "mean_token_accuracy": 0.2123485505580902,
      "num_tokens": 43079736.0,
      "step": 18790
    },
    {
      "entropy": 5.248934888839722,
      "epoch": 1.8054755043227666,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004677226168352416,
      "loss": 5.0686,
      "mean_token_accuracy": 0.20914003998041153,
      "num_tokens": 43091600.0,
      "step": 18795
    },
    {
      "entropy": 5.26110429763794,
      "epoch": 1.8059558117195005,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004677049218012223,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21566279977560043,
      "num_tokens": 43103544.0,
      "step": 18800
    },
    {
      "entropy": 5.296811103820801,
      "epoch": 1.8064361191162344,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000467687222293167,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2196911409497261,
      "num_tokens": 43113533.0,
      "step": 18805
    },
    {
      "entropy": 5.199075412750244,
      "epoch": 1.8069164265129682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004676695183114866,
      "loss": 4.97,
      "mean_token_accuracy": 0.21487183570861818,
      "num_tokens": 43124917.0,
      "step": 18810
    },
    {
      "entropy": 5.26083197593689,
      "epoch": 1.8073967339097021,
      "grad_norm": 1.34375,
      "learning_rate": 0.000467651809856592,
      "loss": 5.002,
      "mean_token_accuracy": 0.2192247062921524,
      "num_tokens": 43136808.0,
      "step": 18815
    },
    {
      "entropy": 5.285592937469483,
      "epoch": 1.807877041306436,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046763409692889446,
      "loss": 5.0387,
      "mean_token_accuracy": 0.20967454016208648,
      "num_tokens": 43148567.0,
      "step": 18820
    },
    {
      "entropy": 5.19307656288147,
      "epoch": 1.8083573487031699,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046761637952880516,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21857287734746933,
      "num_tokens": 43160343.0,
      "step": 18825
    },
    {
      "entropy": 5.276287698745728,
      "epoch": 1.808837656099904,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046759865765673555,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20184392482042313,
      "num_tokens": 43173546.0,
      "step": 18830
    },
    {
      "entropy": 5.297918176651001,
      "epoch": 1.8093179634966379,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000467580931313097,
      "loss": 5.0267,
      "mean_token_accuracy": 0.2205162063241005,
      "num_tokens": 43183846.0,
      "step": 18835
    },
    {
      "entropy": 5.372235250473023,
      "epoch": 1.8097982708933718,
      "grad_norm": 1.25,
      "learning_rate": 0.00046756320049830106,
      "loss": 5.1168,
      "mean_token_accuracy": 0.2040002688765526,
      "num_tokens": 43195612.0,
      "step": 18840
    },
    {
      "entropy": 5.222968196868896,
      "epoch": 1.8102785782901056,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004675454652127594,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2085307404398918,
      "num_tokens": 43207439.0,
      "step": 18845
    },
    {
      "entropy": 5.297936725616455,
      "epoch": 1.8107588856868397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046752772545688377,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21011472940444947,
      "num_tokens": 43218066.0,
      "step": 18850
    },
    {
      "entropy": 5.246335983276367,
      "epoch": 1.8112391930835736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004675099812310861,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21124909371137618,
      "num_tokens": 43230679.0,
      "step": 18855
    },
    {
      "entropy": 5.210201263427734,
      "epoch": 1.8117195004803075,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004674922325357782,
      "loss": 5.0475,
      "mean_token_accuracy": 0.21378218829631807,
      "num_tokens": 43241566.0,
      "step": 18860
    },
    {
      "entropy": 5.3278035640716555,
      "epoch": 1.8121998078770414,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046747447937137235,
      "loss": 5.0748,
      "mean_token_accuracy": 0.2080688700079918,
      "num_tokens": 43252630.0,
      "step": 18865
    },
    {
      "entropy": 5.376592397689819,
      "epoch": 1.8126801152737753,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046745672173828057,
      "loss": 5.1889,
      "mean_token_accuracy": 0.19915415048599244,
      "num_tokens": 43264879.0,
      "step": 18870
    },
    {
      "entropy": 5.239379644393921,
      "epoch": 1.8131604226705091,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004674389596369151,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21481209248304367,
      "num_tokens": 43277321.0,
      "step": 18875
    },
    {
      "entropy": 5.284125423431396,
      "epoch": 1.813640730067243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046742119306768855,
      "loss": 5.067,
      "mean_token_accuracy": 0.21554455757141114,
      "num_tokens": 43289711.0,
      "step": 18880
    },
    {
      "entropy": 5.305082654953003,
      "epoch": 1.814121037463977,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004674034220310132,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21120877861976622,
      "num_tokens": 43301372.0,
      "step": 18885
    },
    {
      "entropy": 5.222808980941773,
      "epoch": 1.8146013448607108,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046738564652730176,
      "loss": 4.8887,
      "mean_token_accuracy": 0.22257370948791505,
      "num_tokens": 43312960.0,
      "step": 18890
    },
    {
      "entropy": 5.183613061904907,
      "epoch": 1.8150816522574447,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004673678665569669,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2142019122838974,
      "num_tokens": 43325039.0,
      "step": 18895
    },
    {
      "entropy": 5.296750402450561,
      "epoch": 1.8155619596541785,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004673500821204213,
      "loss": 5.1106,
      "mean_token_accuracy": 0.20816518664360045,
      "num_tokens": 43337064.0,
      "step": 18900
    },
    {
      "entropy": 5.331899261474609,
      "epoch": 1.8160422670509124,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000467332293218078,
      "loss": 5.0039,
      "mean_token_accuracy": 0.2080813929438591,
      "num_tokens": 43349307.0,
      "step": 18905
    },
    {
      "entropy": 5.216716718673706,
      "epoch": 1.8165225744476465,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046731449985035,
      "loss": 4.9906,
      "mean_token_accuracy": 0.2252137243747711,
      "num_tokens": 43362566.0,
      "step": 18910
    },
    {
      "entropy": 5.243274736404419,
      "epoch": 1.8170028818443804,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046729670201765036,
      "loss": 5.1127,
      "mean_token_accuracy": 0.2087276980280876,
      "num_tokens": 43374516.0,
      "step": 18915
    },
    {
      "entropy": 5.255835914611817,
      "epoch": 1.8174831892411143,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046727889972039227,
      "loss": 4.9206,
      "mean_token_accuracy": 0.2158999964594841,
      "num_tokens": 43386615.0,
      "step": 18920
    },
    {
      "entropy": 5.318445491790771,
      "epoch": 1.8179634966378482,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046726109295898904,
      "loss": 5.0325,
      "mean_token_accuracy": 0.21336304545402526,
      "num_tokens": 43396582.0,
      "step": 18925
    },
    {
      "entropy": 5.17376389503479,
      "epoch": 1.8184438040345823,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004672432817338542,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22312179654836656,
      "num_tokens": 43407617.0,
      "step": 18930
    },
    {
      "entropy": 5.277166700363159,
      "epoch": 1.8189241114313162,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046722546604540115,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20614333301782609,
      "num_tokens": 43421048.0,
      "step": 18935
    },
    {
      "entropy": 5.33627028465271,
      "epoch": 1.81940441882805,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004672076458940436,
      "loss": 5.0747,
      "mean_token_accuracy": 0.21489459574222564,
      "num_tokens": 43433325.0,
      "step": 18940
    },
    {
      "entropy": 5.2072694301605225,
      "epoch": 1.819884726224784,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046718982128019534,
      "loss": 4.9198,
      "mean_token_accuracy": 0.21766716688871385,
      "num_tokens": 43444817.0,
      "step": 18945
    },
    {
      "entropy": 5.258910512924194,
      "epoch": 1.8203650336215178,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046717199220427003,
      "loss": 5.0518,
      "mean_token_accuracy": 0.2105592593550682,
      "num_tokens": 43456295.0,
      "step": 18950
    },
    {
      "entropy": 5.282863903045654,
      "epoch": 1.8208453410182517,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046715415866668163,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2135201781988144,
      "num_tokens": 43468218.0,
      "step": 18955
    },
    {
      "entropy": 5.236268424987793,
      "epoch": 1.8213256484149856,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004671363206678443,
      "loss": 5.0288,
      "mean_token_accuracy": 0.2119890883564949,
      "num_tokens": 43480182.0,
      "step": 18960
    },
    {
      "entropy": 5.269964265823364,
      "epoch": 1.8218059558117194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046711847820817215,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20965287536382676,
      "num_tokens": 43491108.0,
      "step": 18965
    },
    {
      "entropy": 5.235532903671265,
      "epoch": 1.8222862632084533,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004671006312880794,
      "loss": 4.9777,
      "mean_token_accuracy": 0.22017362713813782,
      "num_tokens": 43503530.0,
      "step": 18970
    },
    {
      "entropy": 5.323285627365112,
      "epoch": 1.8227665706051872,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004670827799079805,
      "loss": 5.0661,
      "mean_token_accuracy": 0.21732708364725112,
      "num_tokens": 43514805.0,
      "step": 18975
    },
    {
      "entropy": 5.328000545501709,
      "epoch": 1.823246878001921,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046706492406828966,
      "loss": 5.1453,
      "mean_token_accuracy": 0.20621414482593536,
      "num_tokens": 43526027.0,
      "step": 18980
    },
    {
      "entropy": 5.284997749328613,
      "epoch": 1.8237271853986552,
      "grad_norm": 1.25,
      "learning_rate": 0.0004670470637694217,
      "loss": 5.0736,
      "mean_token_accuracy": 0.21315695643424987,
      "num_tokens": 43537191.0,
      "step": 18985
    },
    {
      "entropy": 5.274760770797729,
      "epoch": 1.824207492795389,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004670291990117912,
      "loss": 5.009,
      "mean_token_accuracy": 0.21176569163799286,
      "num_tokens": 43547494.0,
      "step": 18990
    },
    {
      "entropy": 5.234680080413819,
      "epoch": 1.824687800192123,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004670113297958128,
      "loss": 4.9759,
      "mean_token_accuracy": 0.22401323318481445,
      "num_tokens": 43559048.0,
      "step": 18995
    },
    {
      "entropy": 5.325394105911255,
      "epoch": 1.8251681075888568,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046699345612190155,
      "loss": 5.0385,
      "mean_token_accuracy": 0.20949212163686753,
      "num_tokens": 43569571.0,
      "step": 19000
    },
    {
      "entropy": 5.235664558410645,
      "epoch": 1.825648414985591,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046697557799047233,
      "loss": 4.963,
      "mean_token_accuracy": 0.21767594963312148,
      "num_tokens": 43580092.0,
      "step": 19005
    },
    {
      "entropy": 5.189020395278931,
      "epoch": 1.8261287223823248,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004669576954019403,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21412672996520996,
      "num_tokens": 43592867.0,
      "step": 19010
    },
    {
      "entropy": 5.218458604812622,
      "epoch": 1.8266090297790587,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004669398083567205,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2149222269654274,
      "num_tokens": 43602618.0,
      "step": 19015
    },
    {
      "entropy": 5.26710000038147,
      "epoch": 1.8270893371757926,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004669219168552284,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21357909590005875,
      "num_tokens": 43614425.0,
      "step": 19020
    },
    {
      "entropy": 5.17151985168457,
      "epoch": 1.8275696445725265,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046690402089787916,
      "loss": 4.8826,
      "mean_token_accuracy": 0.21835967004299164,
      "num_tokens": 43627334.0,
      "step": 19025
    },
    {
      "entropy": 5.201187467575073,
      "epoch": 1.8280499519692603,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004668861204850884,
      "loss": 4.9768,
      "mean_token_accuracy": 0.218691848218441,
      "num_tokens": 43638509.0,
      "step": 19030
    },
    {
      "entropy": 5.202214908599854,
      "epoch": 1.8285302593659942,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046686821561727176,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22288002967834472,
      "num_tokens": 43649633.0,
      "step": 19035
    },
    {
      "entropy": 5.298175239562989,
      "epoch": 1.829010566762728,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004668503062948449,
      "loss": 5.0182,
      "mean_token_accuracy": 0.21366020888090134,
      "num_tokens": 43661175.0,
      "step": 19040
    },
    {
      "entropy": 5.219539022445678,
      "epoch": 1.829490874159462,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004668323925182236,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2159034565091133,
      "num_tokens": 43672000.0,
      "step": 19045
    },
    {
      "entropy": 5.187828779220581,
      "epoch": 1.8299711815561959,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046681447428782377,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21802874952554702,
      "num_tokens": 43682574.0,
      "step": 19050
    },
    {
      "entropy": 5.262164115905762,
      "epoch": 1.8304514889529298,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004667965516040613,
      "loss": 5.0534,
      "mean_token_accuracy": 0.21243723928928376,
      "num_tokens": 43693051.0,
      "step": 19055
    },
    {
      "entropy": 5.298206758499146,
      "epoch": 1.8309317963496636,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004667786244673526,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21041272729635238,
      "num_tokens": 43705471.0,
      "step": 19060
    },
    {
      "entropy": 5.197599983215332,
      "epoch": 1.8314121037463977,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046676069287811365,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22551905661821364,
      "num_tokens": 43716676.0,
      "step": 19065
    },
    {
      "entropy": 5.302572250366211,
      "epoch": 1.8318924111431316,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004667427568367607,
      "loss": 5.0486,
      "mean_token_accuracy": 0.21318131685256958,
      "num_tokens": 43727997.0,
      "step": 19070
    },
    {
      "entropy": 5.275035285949707,
      "epoch": 1.8323727185398655,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046672481634371047,
      "loss": 5.0384,
      "mean_token_accuracy": 0.2127968117594719,
      "num_tokens": 43739163.0,
      "step": 19075
    },
    {
      "entropy": 5.275390863418579,
      "epoch": 1.8328530259365994,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046670687139937925,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21151025593280792,
      "num_tokens": 43751442.0,
      "step": 19080
    },
    {
      "entropy": 5.230707359313965,
      "epoch": 1.8333333333333335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004666889220041837,
      "loss": 5.054,
      "mean_token_accuracy": 0.21026014238595964,
      "num_tokens": 43763430.0,
      "step": 19085
    },
    {
      "entropy": 5.2473915100097654,
      "epoch": 1.8338136407300674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046667096815854056,
      "loss": 4.9451,
      "mean_token_accuracy": 0.2247275784611702,
      "num_tokens": 43774660.0,
      "step": 19090
    },
    {
      "entropy": 5.2131236553192135,
      "epoch": 1.8342939481268012,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004666530098628667,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22274540811777116,
      "num_tokens": 43785997.0,
      "step": 19095
    },
    {
      "entropy": 5.255900621414185,
      "epoch": 1.8347742555235351,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004666350471175791,
      "loss": 5.0072,
      "mean_token_accuracy": 0.21260422468185425,
      "num_tokens": 43797744.0,
      "step": 19100
    },
    {
      "entropy": 5.305290174484253,
      "epoch": 1.835254562920269,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004666170799230947,
      "loss": 5.0852,
      "mean_token_accuracy": 0.20951036661863326,
      "num_tokens": 43810415.0,
      "step": 19105
    },
    {
      "entropy": 5.153095388412476,
      "epoch": 1.8357348703170029,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004665991082798307,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21873484253883363,
      "num_tokens": 43822652.0,
      "step": 19110
    },
    {
      "entropy": 5.311591768264771,
      "epoch": 1.8362151777137368,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004665811321882043,
      "loss": 5.057,
      "mean_token_accuracy": 0.209820257127285,
      "num_tokens": 43834607.0,
      "step": 19115
    },
    {
      "entropy": 5.196281814575196,
      "epoch": 1.8366954851104706,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046656315164863297,
      "loss": 5.0049,
      "mean_token_accuracy": 0.21712229102849961,
      "num_tokens": 43845124.0,
      "step": 19120
    },
    {
      "entropy": 5.261275434494019,
      "epoch": 1.8371757925072045,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046654516666153403,
      "loss": 5.1002,
      "mean_token_accuracy": 0.20942269265651703,
      "num_tokens": 43855964.0,
      "step": 19125
    },
    {
      "entropy": 5.264906883239746,
      "epoch": 1.8376560999039384,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004665271772273251,
      "loss": 4.9388,
      "mean_token_accuracy": 0.2267317071557045,
      "num_tokens": 43866659.0,
      "step": 19130
    },
    {
      "entropy": 5.260662364959717,
      "epoch": 1.8381364073006723,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004665091833464239,
      "loss": 5.0622,
      "mean_token_accuracy": 0.21081708520650863,
      "num_tokens": 43877410.0,
      "step": 19135
    },
    {
      "entropy": 5.256427669525147,
      "epoch": 1.8386167146974062,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046649118501924805,
      "loss": 5.0782,
      "mean_token_accuracy": 0.21080237179994582,
      "num_tokens": 43889672.0,
      "step": 19140
    },
    {
      "entropy": 5.251862668991089,
      "epoch": 1.8390970220941403,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004664731822462154,
      "loss": 5.044,
      "mean_token_accuracy": 0.212780424952507,
      "num_tokens": 43900876.0,
      "step": 19145
    },
    {
      "entropy": 5.225858306884765,
      "epoch": 1.8395773294908742,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046645517502774415,
      "loss": 4.8963,
      "mean_token_accuracy": 0.22224969416856766,
      "num_tokens": 43911460.0,
      "step": 19150
    },
    {
      "entropy": 5.21096363067627,
      "epoch": 1.840057636887608,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046643716336425224,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21370268762111663,
      "num_tokens": 43922344.0,
      "step": 19155
    },
    {
      "entropy": 5.225855541229248,
      "epoch": 1.8405379442843421,
      "grad_norm": 1.25,
      "learning_rate": 0.0004664191472561578,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22057160437107087,
      "num_tokens": 43933960.0,
      "step": 19160
    },
    {
      "entropy": 5.2479558944702145,
      "epoch": 1.841018251681076,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004664011267038792,
      "loss": 5.1389,
      "mean_token_accuracy": 0.210689041018486,
      "num_tokens": 43945805.0,
      "step": 19165
    },
    {
      "entropy": 5.287203454971314,
      "epoch": 1.84149855907781,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046638310170783476,
      "loss": 5.0534,
      "mean_token_accuracy": 0.2150415986776352,
      "num_tokens": 43958318.0,
      "step": 19170
    },
    {
      "entropy": 5.253131008148193,
      "epoch": 1.8419788664745438,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000466365072268443,
      "loss": 4.9088,
      "mean_token_accuracy": 0.22445641309022904,
      "num_tokens": 43969385.0,
      "step": 19175
    },
    {
      "entropy": 5.181105995178223,
      "epoch": 1.8424591738712777,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004663470383861225,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2198967903852463,
      "num_tokens": 43980383.0,
      "step": 19180
    },
    {
      "entropy": 5.253467559814453,
      "epoch": 1.8429394812680115,
      "grad_norm": 1.109375,
      "learning_rate": 0.000466329000061292,
      "loss": 5.0331,
      "mean_token_accuracy": 0.21567924320697784,
      "num_tokens": 43992900.0,
      "step": 19185
    },
    {
      "entropy": 5.29862790107727,
      "epoch": 1.8434197886647454,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004663109572943702,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22889216095209122,
      "num_tokens": 44001823.0,
      "step": 19190
    },
    {
      "entropy": 5.277613925933838,
      "epoch": 1.8439000960614793,
      "grad_norm": 1.125,
      "learning_rate": 0.000466292910085776,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20697865188121795,
      "num_tokens": 44014851.0,
      "step": 19195
    },
    {
      "entropy": 5.2534040927886965,
      "epoch": 1.8443804034582132,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046627485843592854,
      "loss": 5.0467,
      "mean_token_accuracy": 0.2088354453444481,
      "num_tokens": 44025191.0,
      "step": 19200
    },
    {
      "entropy": 5.3285150051116945,
      "epoch": 1.844860710854947,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046625680234524674,
      "loss": 5.0654,
      "mean_token_accuracy": 0.21145121455192567,
      "num_tokens": 44036319.0,
      "step": 19205
    },
    {
      "entropy": 5.351266717910766,
      "epoch": 1.845341018251681,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046623874181414993,
      "loss": 5.0986,
      "mean_token_accuracy": 0.2089647889137268,
      "num_tokens": 44048618.0,
      "step": 19210
    },
    {
      "entropy": 5.270941877365113,
      "epoch": 1.8458213256484148,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004662206768430574,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21442210376262666,
      "num_tokens": 44060754.0,
      "step": 19215
    },
    {
      "entropy": 5.200630235671997,
      "epoch": 1.846301633045149,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004662026074323886,
      "loss": 4.9586,
      "mean_token_accuracy": 0.21486074030399321,
      "num_tokens": 44071239.0,
      "step": 19220
    },
    {
      "entropy": 5.3543799877166744,
      "epoch": 1.8467819404418828,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046618453358256303,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20405264794826508,
      "num_tokens": 44082635.0,
      "step": 19225
    },
    {
      "entropy": 5.301238775253296,
      "epoch": 1.8472622478386167,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046616645529400026,
      "loss": 4.9881,
      "mean_token_accuracy": 0.2151183992624283,
      "num_tokens": 44093680.0,
      "step": 19230
    },
    {
      "entropy": 5.249710750579834,
      "epoch": 1.8477425552353506,
      "grad_norm": 1.125,
      "learning_rate": 0.0004661483725671201,
      "loss": 4.9907,
      "mean_token_accuracy": 0.21503675132989883,
      "num_tokens": 44104490.0,
      "step": 19235
    },
    {
      "entropy": 5.153631067276001,
      "epoch": 1.8482228626320847,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046613028540234226,
      "loss": 4.9328,
      "mean_token_accuracy": 0.2183071717619896,
      "num_tokens": 44116961.0,
      "step": 19240
    },
    {
      "entropy": 5.226397752761841,
      "epoch": 1.8487031700288186,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004661121938000867,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22256339490413665,
      "num_tokens": 44127264.0,
      "step": 19245
    },
    {
      "entropy": 5.237424659729004,
      "epoch": 1.8491834774255524,
      "grad_norm": 1.25,
      "learning_rate": 0.0004660940977607736,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2198490396142006,
      "num_tokens": 44139374.0,
      "step": 19250
    },
    {
      "entropy": 5.151271963119507,
      "epoch": 1.8496637848222863,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046607599728482285,
      "loss": 4.8809,
      "mean_token_accuracy": 0.21850163042545317,
      "num_tokens": 44150416.0,
      "step": 19255
    },
    {
      "entropy": 5.282458734512329,
      "epoch": 1.8501440922190202,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046605789237265496,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21628472208976746,
      "num_tokens": 44162094.0,
      "step": 19260
    },
    {
      "entropy": 5.2536633014678955,
      "epoch": 1.850624399615754,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046603978302469,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20829408764839172,
      "num_tokens": 44174526.0,
      "step": 19265
    },
    {
      "entropy": 5.250750875473022,
      "epoch": 1.851104707012488,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004660216692413486,
      "loss": 4.9783,
      "mean_token_accuracy": 0.21716944575309755,
      "num_tokens": 44185961.0,
      "step": 19270
    },
    {
      "entropy": 5.153362131118774,
      "epoch": 1.8515850144092219,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004660035510230513,
      "loss": 4.966,
      "mean_token_accuracy": 0.21736075282096862,
      "num_tokens": 44198431.0,
      "step": 19275
    },
    {
      "entropy": 5.283346462249756,
      "epoch": 1.8520653218059557,
      "grad_norm": 1.125,
      "learning_rate": 0.0004659854283702186,
      "loss": 5.0296,
      "mean_token_accuracy": 0.21926084607839585,
      "num_tokens": 44210518.0,
      "step": 19280
    },
    {
      "entropy": 5.169636344909668,
      "epoch": 1.8525456292026896,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004659673012832715,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21551107466220856,
      "num_tokens": 44222300.0,
      "step": 19285
    },
    {
      "entropy": 5.214459705352783,
      "epoch": 1.8530259365994235,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004659491697626306,
      "loss": 5.0386,
      "mean_token_accuracy": 0.21116111427545547,
      "num_tokens": 44233142.0,
      "step": 19290
    },
    {
      "entropy": 5.229175043106079,
      "epoch": 1.8535062439961574,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046593103380871705,
      "loss": 5.043,
      "mean_token_accuracy": 0.21388214379549025,
      "num_tokens": 44246507.0,
      "step": 19295
    },
    {
      "entropy": 5.295354413986206,
      "epoch": 1.8539865513928915,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046591289342195184,
      "loss": 5.0353,
      "mean_token_accuracy": 0.2086443156003952,
      "num_tokens": 44257451.0,
      "step": 19300
    },
    {
      "entropy": 5.186765146255493,
      "epoch": 1.8544668587896254,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004658947486027562,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22254096567630768,
      "num_tokens": 44269157.0,
      "step": 19305
    },
    {
      "entropy": 5.1699975490570065,
      "epoch": 1.8549471661863592,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046587659935155124,
      "loss": 4.943,
      "mean_token_accuracy": 0.2166450098156929,
      "num_tokens": 44279890.0,
      "step": 19310
    },
    {
      "entropy": 5.19454779624939,
      "epoch": 1.8554274735830933,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046585844566875845,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2237894669175148,
      "num_tokens": 44291457.0,
      "step": 19315
    },
    {
      "entropy": 5.179895067214966,
      "epoch": 1.8559077809798272,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004658402875547993,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21959411948919297,
      "num_tokens": 44301692.0,
      "step": 19320
    },
    {
      "entropy": 5.248722887039184,
      "epoch": 1.856388088376561,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046582212501009533,
      "loss": 5.0317,
      "mean_token_accuracy": 0.20764374732971191,
      "num_tokens": 44313540.0,
      "step": 19325
    },
    {
      "entropy": 5.279478120803833,
      "epoch": 1.856868395773295,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046580395803506825,
      "loss": 5.0878,
      "mean_token_accuracy": 0.21689770817756654,
      "num_tokens": 44324970.0,
      "step": 19330
    },
    {
      "entropy": 5.271679067611695,
      "epoch": 1.8573487031700289,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004657857866301399,
      "loss": 5.0116,
      "mean_token_accuracy": 0.212438403069973,
      "num_tokens": 44336521.0,
      "step": 19335
    },
    {
      "entropy": 5.157049465179443,
      "epoch": 1.8578290105667628,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046576761079573204,
      "loss": 4.898,
      "mean_token_accuracy": 0.22655290514230728,
      "num_tokens": 44347128.0,
      "step": 19340
    },
    {
      "entropy": 5.240991830825806,
      "epoch": 1.8583093179634966,
      "grad_norm": 1.125,
      "learning_rate": 0.0004657494305322667,
      "loss": 5.0737,
      "mean_token_accuracy": 0.2119502142071724,
      "num_tokens": 44358503.0,
      "step": 19345
    },
    {
      "entropy": 5.294353437423706,
      "epoch": 1.8587896253602305,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004657312458401661,
      "loss": 5.0357,
      "mean_token_accuracy": 0.21674038618803024,
      "num_tokens": 44370849.0,
      "step": 19350
    },
    {
      "entropy": 5.327580690383911,
      "epoch": 1.8592699327569644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004657130567198522,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21758226901292801,
      "num_tokens": 44383806.0,
      "step": 19355
    },
    {
      "entropy": 5.1128240585327145,
      "epoch": 1.8597502401536983,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046569486317174746,
      "loss": 4.7969,
      "mean_token_accuracy": 0.22238929569721222,
      "num_tokens": 44393937.0,
      "step": 19360
    },
    {
      "entropy": 5.165859127044678,
      "epoch": 1.8602305475504322,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004656766651962742,
      "loss": 5.0369,
      "mean_token_accuracy": 0.2191337153315544,
      "num_tokens": 44404967.0,
      "step": 19365
    },
    {
      "entropy": 5.259901523590088,
      "epoch": 1.860710854947166,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004656584627938551,
      "loss": 5.0115,
      "mean_token_accuracy": 0.22085045278072357,
      "num_tokens": 44416421.0,
      "step": 19370
    },
    {
      "entropy": 5.209828662872314,
      "epoch": 1.8611911623439001,
      "grad_norm": 1.125,
      "learning_rate": 0.00046564025596491254,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2200235053896904,
      "num_tokens": 44427145.0,
      "step": 19375
    },
    {
      "entropy": 5.217675018310547,
      "epoch": 1.861671469740634,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004656220447098693,
      "loss": 4.9562,
      "mean_token_accuracy": 0.22361350506544114,
      "num_tokens": 44438456.0,
      "step": 19380
    },
    {
      "entropy": 5.188925790786743,
      "epoch": 1.862151777137368,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004656038290291483,
      "loss": 4.9135,
      "mean_token_accuracy": 0.21427663266658784,
      "num_tokens": 44450366.0,
      "step": 19385
    },
    {
      "entropy": 5.166432428359985,
      "epoch": 1.8626320845341018,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004655856089231723,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22123729288578034,
      "num_tokens": 44461246.0,
      "step": 19390
    },
    {
      "entropy": 5.253466606140137,
      "epoch": 1.8631123919308359,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004655673843923644,
      "loss": 4.9901,
      "mean_token_accuracy": 0.2147843897342682,
      "num_tokens": 44473099.0,
      "step": 19395
    },
    {
      "entropy": 5.252869558334351,
      "epoch": 1.8635926993275698,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004655491554371477,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21053396910429,
      "num_tokens": 44484516.0,
      "step": 19400
    },
    {
      "entropy": 5.12810869216919,
      "epoch": 1.8640730067243036,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046553092205794543,
      "loss": 4.8784,
      "mean_token_accuracy": 0.2256895512342453,
      "num_tokens": 44495500.0,
      "step": 19405
    },
    {
      "entropy": 5.164000844955444,
      "epoch": 1.8645533141210375,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046551268425518096,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21758059710264205,
      "num_tokens": 44506346.0,
      "step": 19410
    },
    {
      "entropy": 5.268248987197876,
      "epoch": 1.8650336215177714,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004654944420292776,
      "loss": 5.0661,
      "mean_token_accuracy": 0.2116922691464424,
      "num_tokens": 44517234.0,
      "step": 19415
    },
    {
      "entropy": 5.19019775390625,
      "epoch": 1.8655139289145053,
      "grad_norm": 1.171875,
      "learning_rate": 0.000465476195380659,
      "loss": 4.8895,
      "mean_token_accuracy": 0.22848164141178132,
      "num_tokens": 44528543.0,
      "step": 19420
    },
    {
      "entropy": 5.253494119644165,
      "epoch": 1.8659942363112392,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004654579443097487,
      "loss": 5.0417,
      "mean_token_accuracy": 0.21326085180044174,
      "num_tokens": 44539385.0,
      "step": 19425
    },
    {
      "entropy": 5.125870037078857,
      "epoch": 1.866474543707973,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004654396888169705,
      "loss": 4.8645,
      "mean_token_accuracy": 0.22228908091783522,
      "num_tokens": 44550013.0,
      "step": 19430
    },
    {
      "entropy": 5.232476997375488,
      "epoch": 1.866954851104707,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046542142890274816,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21201082915067673,
      "num_tokens": 44561926.0,
      "step": 19435
    },
    {
      "entropy": 5.156580829620362,
      "epoch": 1.8674351585014408,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004654031645675057,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2183776706457138,
      "num_tokens": 44574589.0,
      "step": 19440
    },
    {
      "entropy": 5.244059228897095,
      "epoch": 1.8679154658981747,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004653848958116672,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21688778549432755,
      "num_tokens": 44586531.0,
      "step": 19445
    },
    {
      "entropy": 5.246910047531128,
      "epoch": 1.8683957732949086,
      "grad_norm": 1.125,
      "learning_rate": 0.00046536662263565667,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22195253223180772,
      "num_tokens": 44597432.0,
      "step": 19450
    },
    {
      "entropy": 5.291172027587891,
      "epoch": 1.8688760806916427,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004653483450398985,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20669280290603637,
      "num_tokens": 44609264.0,
      "step": 19455
    },
    {
      "entropy": 5.137449645996094,
      "epoch": 1.8693563880883766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046533006302481694,
      "loss": 4.8906,
      "mean_token_accuracy": 0.22691554576158524,
      "num_tokens": 44620410.0,
      "step": 19460
    },
    {
      "entropy": 5.222999238967896,
      "epoch": 1.8698366954851104,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004653117765908365,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21418838798999787,
      "num_tokens": 44633123.0,
      "step": 19465
    },
    {
      "entropy": 5.285911703109742,
      "epoch": 1.8703170028818443,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004652934857383816,
      "loss": 5.0353,
      "mean_token_accuracy": 0.20569598227739333,
      "num_tokens": 44644693.0,
      "step": 19470
    },
    {
      "entropy": 5.194276523590088,
      "epoch": 1.8707973102785784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004652751904678772,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22709491699934006,
      "num_tokens": 44656669.0,
      "step": 19475
    },
    {
      "entropy": 5.2133385181427006,
      "epoch": 1.8712776176753123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046525689077974775,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22596708089113235,
      "num_tokens": 44667938.0,
      "step": 19480
    },
    {
      "entropy": 5.2457986831665036,
      "epoch": 1.8717579250720462,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046523858667441834,
      "loss": 5.035,
      "mean_token_accuracy": 0.20877998769283296,
      "num_tokens": 44678967.0,
      "step": 19485
    },
    {
      "entropy": 5.253645229339599,
      "epoch": 1.87223823246878,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004652202781523138,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22498586922883987,
      "num_tokens": 44689583.0,
      "step": 19490
    },
    {
      "entropy": 5.175134563446045,
      "epoch": 1.872718539865514,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004652019652138592,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22119050472974777,
      "num_tokens": 44700942.0,
      "step": 19495
    },
    {
      "entropy": 5.203688621520996,
      "epoch": 1.8731988472622478,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004651836478594798,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22072943598031997,
      "num_tokens": 44712323.0,
      "step": 19500
    },
    {
      "entropy": 5.303492021560669,
      "epoch": 1.8736791546589817,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004651653260896008,
      "loss": 5.1226,
      "mean_token_accuracy": 0.21065306961536406,
      "num_tokens": 44724033.0,
      "step": 19505
    },
    {
      "entropy": 5.274795150756836,
      "epoch": 1.8741594620557156,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046514699990464763,
      "loss": 5.0374,
      "mean_token_accuracy": 0.21364638358354568,
      "num_tokens": 44735388.0,
      "step": 19510
    },
    {
      "entropy": 5.23540620803833,
      "epoch": 1.8746397694524495,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004651286693050458,
      "loss": 4.9632,
      "mean_token_accuracy": 0.2217638984322548,
      "num_tokens": 44748165.0,
      "step": 19515
    },
    {
      "entropy": 5.241716194152832,
      "epoch": 1.8751200768491834,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004651103342912207,
      "loss": 4.967,
      "mean_token_accuracy": 0.22017327696084976,
      "num_tokens": 44758500.0,
      "step": 19520
    },
    {
      "entropy": 5.285635375976563,
      "epoch": 1.8756003842459172,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046509199486359824,
      "loss": 5.1232,
      "mean_token_accuracy": 0.20430054068565368,
      "num_tokens": 44770347.0,
      "step": 19525
    },
    {
      "entropy": 5.277915668487549,
      "epoch": 1.8760806916426513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046507365102260403,
      "loss": 4.9741,
      "mean_token_accuracy": 0.21804948449134826,
      "num_tokens": 44782365.0,
      "step": 19530
    },
    {
      "entropy": 5.188326120376587,
      "epoch": 1.8765609990393852,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046505530276866417,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22620483934879304,
      "num_tokens": 44793128.0,
      "step": 19535
    },
    {
      "entropy": 5.185820531845093,
      "epoch": 1.877041306436119,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046503695010220443,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21976949274539948,
      "num_tokens": 44804236.0,
      "step": 19540
    },
    {
      "entropy": 5.161147832870483,
      "epoch": 1.877521613832853,
      "grad_norm": 1.15625,
      "learning_rate": 0.000465018593023651,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21554794460535048,
      "num_tokens": 44815983.0,
      "step": 19545
    },
    {
      "entropy": 5.268780183792114,
      "epoch": 1.878001921229587,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004650002315334302,
      "loss": 5.1124,
      "mean_token_accuracy": 0.20502481162548064,
      "num_tokens": 44828267.0,
      "step": 19550
    },
    {
      "entropy": 5.181786584854126,
      "epoch": 1.878482228626321,
      "grad_norm": 1.203125,
      "learning_rate": 0.000464981865631968,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22299952507019044,
      "num_tokens": 44838668.0,
      "step": 19555
    },
    {
      "entropy": 5.2035966396331785,
      "epoch": 1.8789625360230549,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004649634953196912,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21632902026176454,
      "num_tokens": 44849492.0,
      "step": 19560
    },
    {
      "entropy": 5.403333902359009,
      "epoch": 1.8794428434197887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046494512059702605,
      "loss": 5.1513,
      "mean_token_accuracy": 0.20506762862205505,
      "num_tokens": 44861388.0,
      "step": 19565
    },
    {
      "entropy": 5.206557369232177,
      "epoch": 1.8799231508165226,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004649267414643992,
      "loss": 4.9708,
      "mean_token_accuracy": 0.2163632556796074,
      "num_tokens": 44873353.0,
      "step": 19570
    },
    {
      "entropy": 5.2025166034698485,
      "epoch": 1.8804034582132565,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004649083579222374,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21468825042247772,
      "num_tokens": 44885020.0,
      "step": 19575
    },
    {
      "entropy": 5.355774974822998,
      "epoch": 1.8808837656099904,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046488996997096744,
      "loss": 5.1145,
      "mean_token_accuracy": 0.2059138000011444,
      "num_tokens": 44896743.0,
      "step": 19580
    },
    {
      "entropy": 5.249156332015991,
      "epoch": 1.8813640730067243,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004648715776110162,
      "loss": 4.9561,
      "mean_token_accuracy": 0.22051790952682496,
      "num_tokens": 44909138.0,
      "step": 19585
    },
    {
      "entropy": 5.204092931747437,
      "epoch": 1.8818443804034581,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004648531808428108,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21704428344964982,
      "num_tokens": 44921139.0,
      "step": 19590
    },
    {
      "entropy": 5.151141214370727,
      "epoch": 1.882324687800192,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004648347796667782,
      "loss": 4.8581,
      "mean_token_accuracy": 0.21930660605430602,
      "num_tokens": 44933282.0,
      "step": 19595
    },
    {
      "entropy": 5.258844614028931,
      "epoch": 1.882804995196926,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004648163740833458,
      "loss": 5.0472,
      "mean_token_accuracy": 0.2039278954267502,
      "num_tokens": 44946350.0,
      "step": 19600
    },
    {
      "entropy": 5.204915142059326,
      "epoch": 1.8832853025936598,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046479796409294076,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2116349831223488,
      "num_tokens": 44958348.0,
      "step": 19605
    },
    {
      "entropy": 5.204988098144531,
      "epoch": 1.8837656099903939,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004647795496959907,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22756162732839585,
      "num_tokens": 44970206.0,
      "step": 19610
    },
    {
      "entropy": 5.298921871185303,
      "epoch": 1.8842459173871278,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046476113089292286,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2106211394071579,
      "num_tokens": 44980814.0,
      "step": 19615
    },
    {
      "entropy": 5.1849854469299315,
      "epoch": 1.8847262247838616,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004647427076841651,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2235767051577568,
      "num_tokens": 44991148.0,
      "step": 19620
    },
    {
      "entropy": 5.117211246490479,
      "epoch": 1.8852065321805955,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046472428007014515,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22996314615011215,
      "num_tokens": 45003070.0,
      "step": 19625
    },
    {
      "entropy": 5.194110012054443,
      "epoch": 1.8856868395773296,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004647058480512907,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2154890462756157,
      "num_tokens": 45015760.0,
      "step": 19630
    },
    {
      "entropy": 5.261440658569336,
      "epoch": 1.8861671469740635,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046468741162802987,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21402243226766587,
      "num_tokens": 45026297.0,
      "step": 19635
    },
    {
      "entropy": 5.198152494430542,
      "epoch": 1.8866474543707974,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004646689708007905,
      "loss": 4.8901,
      "mean_token_accuracy": 0.2198719009757042,
      "num_tokens": 45037814.0,
      "step": 19640
    },
    {
      "entropy": 5.194475746154785,
      "epoch": 1.8871277617675313,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046465052557000087,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22716270983219147,
      "num_tokens": 45050632.0,
      "step": 19645
    },
    {
      "entropy": 5.272702217102051,
      "epoch": 1.8876080691642652,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046463207593608916,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21338137686252595,
      "num_tokens": 45061866.0,
      "step": 19650
    },
    {
      "entropy": 5.219325923919678,
      "epoch": 1.888088376560999,
      "grad_norm": 1.125,
      "learning_rate": 0.0004646136218994837,
      "loss": 5.065,
      "mean_token_accuracy": 0.20938050597906113,
      "num_tokens": 45074550.0,
      "step": 19655
    },
    {
      "entropy": 5.292240571975708,
      "epoch": 1.888568683957733,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046459516346061304,
      "loss": 5.0253,
      "mean_token_accuracy": 0.21347078531980515,
      "num_tokens": 45085710.0,
      "step": 19660
    },
    {
      "entropy": 5.229112672805786,
      "epoch": 1.8890489913544668,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046457670061990564,
      "loss": 4.9882,
      "mean_token_accuracy": 0.21825681179761885,
      "num_tokens": 45097917.0,
      "step": 19665
    },
    {
      "entropy": 5.16017484664917,
      "epoch": 1.8895292987512007,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046455823337779024,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22570082694292068,
      "num_tokens": 45108455.0,
      "step": 19670
    },
    {
      "entropy": 5.3094648838043215,
      "epoch": 1.8900096061479346,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004645397617346954,
      "loss": 5.139,
      "mean_token_accuracy": 0.2093791350722313,
      "num_tokens": 45120313.0,
      "step": 19675
    },
    {
      "entropy": 5.228615379333496,
      "epoch": 1.8904899135446684,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004645212856910502,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2183237835764885,
      "num_tokens": 45131673.0,
      "step": 19680
    },
    {
      "entropy": 5.173794984817505,
      "epoch": 1.8909702209414025,
      "grad_norm": 1.25,
      "learning_rate": 0.0004645028052472835,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21703631728887557,
      "num_tokens": 45143589.0,
      "step": 19685
    },
    {
      "entropy": 5.183329772949219,
      "epoch": 1.8914505283381364,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046448432040382444,
      "loss": 4.94,
      "mean_token_accuracy": 0.22232582569122314,
      "num_tokens": 45155904.0,
      "step": 19690
    },
    {
      "entropy": 5.20554552078247,
      "epoch": 1.8919308357348703,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000464465831161102,
      "loss": 4.9386,
      "mean_token_accuracy": 0.22366637289524077,
      "num_tokens": 45166802.0,
      "step": 19695
    },
    {
      "entropy": 5.278332281112671,
      "epoch": 1.8924111431316042,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004644473375195456,
      "loss": 5.0304,
      "mean_token_accuracy": 0.2092037796974182,
      "num_tokens": 45178986.0,
      "step": 19700
    },
    {
      "entropy": 5.202651023864746,
      "epoch": 1.8928914505283383,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046442883947958466,
      "loss": 4.9905,
      "mean_token_accuracy": 0.22424103766679765,
      "num_tokens": 45190814.0,
      "step": 19705
    },
    {
      "entropy": 5.180655002593994,
      "epoch": 1.8933717579250722,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046441033704164845,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2252107799053192,
      "num_tokens": 45202553.0,
      "step": 19710
    },
    {
      "entropy": 5.265337705612183,
      "epoch": 1.893852065321806,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004643918302061666,
      "loss": 5.0597,
      "mean_token_accuracy": 0.21409657895565032,
      "num_tokens": 45214482.0,
      "step": 19715
    },
    {
      "entropy": 5.187909412384033,
      "epoch": 1.89433237271854,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004643733189735689,
      "loss": 4.9551,
      "mean_token_accuracy": 0.21984284073114396,
      "num_tokens": 45226675.0,
      "step": 19720
    },
    {
      "entropy": 5.240240430831909,
      "epoch": 1.8948126801152738,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000464354803344285,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21885981559753417,
      "num_tokens": 45238995.0,
      "step": 19725
    },
    {
      "entropy": 5.196890258789063,
      "epoch": 1.8952929875120077,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046433628331874496,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2164921522140503,
      "num_tokens": 45250696.0,
      "step": 19730
    },
    {
      "entropy": 5.2248022079467775,
      "epoch": 1.8957732949087416,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004643177588973785,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22189972400665284,
      "num_tokens": 45262434.0,
      "step": 19735
    },
    {
      "entropy": 5.213742017745972,
      "epoch": 1.8962536023054755,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004642992300806159,
      "loss": 5.039,
      "mean_token_accuracy": 0.21464217454195023,
      "num_tokens": 45273267.0,
      "step": 19740
    },
    {
      "entropy": 5.196971654891968,
      "epoch": 1.8967339097022093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004642806968688873,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21888742595911026,
      "num_tokens": 45284601.0,
      "step": 19745
    },
    {
      "entropy": 5.152501916885376,
      "epoch": 1.8972142170989432,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046426215926262295,
      "loss": 4.9693,
      "mean_token_accuracy": 0.21905821859836577,
      "num_tokens": 45296504.0,
      "step": 19750
    },
    {
      "entropy": 5.14156928062439,
      "epoch": 1.897694524495677,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004642436172622532,
      "loss": 4.8978,
      "mean_token_accuracy": 0.225938580930233,
      "num_tokens": 45307232.0,
      "step": 19755
    },
    {
      "entropy": 5.113257360458374,
      "epoch": 1.898174831892411,
      "grad_norm": 1.25,
      "learning_rate": 0.0004642250708682086,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2196539208292961,
      "num_tokens": 45317768.0,
      "step": 19760
    },
    {
      "entropy": 5.255691003799439,
      "epoch": 1.898655139289145,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046420652008091984,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21009523421525955,
      "num_tokens": 45329172.0,
      "step": 19765
    },
    {
      "entropy": 5.279930925369262,
      "epoch": 1.899135446685879,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004641879649008174,
      "loss": 5.0577,
      "mean_token_accuracy": 0.21355075985193253,
      "num_tokens": 45339840.0,
      "step": 19770
    },
    {
      "entropy": 5.268459844589233,
      "epoch": 1.8996157540826129,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004641694053283323,
      "loss": 5.0521,
      "mean_token_accuracy": 0.21217428743839264,
      "num_tokens": 45351532.0,
      "step": 19775
    },
    {
      "entropy": 5.250380659103394,
      "epoch": 1.9000960614793467,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046415084136389525,
      "loss": 5.068,
      "mean_token_accuracy": 0.21027124524116517,
      "num_tokens": 45362872.0,
      "step": 19780
    },
    {
      "entropy": 5.429735708236694,
      "epoch": 1.9005763688760808,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004641322730079374,
      "loss": 5.1282,
      "mean_token_accuracy": 0.20690999627113343,
      "num_tokens": 45374877.0,
      "step": 19785
    },
    {
      "entropy": 5.240040588378906,
      "epoch": 1.9010566762728147,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004641137002608897,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21904985904693602,
      "num_tokens": 45387136.0,
      "step": 19790
    },
    {
      "entropy": 5.181788110733033,
      "epoch": 1.9015369836695486,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046409512312318345,
      "loss": 4.9339,
      "mean_token_accuracy": 0.22222308814525604,
      "num_tokens": 45398310.0,
      "step": 19795
    },
    {
      "entropy": 5.21841025352478,
      "epoch": 1.9020172910662825,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00046407654159524994,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21402820497751235,
      "num_tokens": 45410456.0,
      "step": 19800
    },
    {
      "entropy": 5.228430986404419,
      "epoch": 1.9024975984630164,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046405795567752055,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2190715953707695,
      "num_tokens": 45423292.0,
      "step": 19805
    },
    {
      "entropy": 5.156324005126953,
      "epoch": 1.9029779058597502,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046403936537042686,
      "loss": 4.9695,
      "mean_token_accuracy": 0.2253912255167961,
      "num_tokens": 45434699.0,
      "step": 19810
    },
    {
      "entropy": 5.235641956329346,
      "epoch": 1.9034582132564841,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046402077067440043,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21601807475090026,
      "num_tokens": 45446849.0,
      "step": 19815
    },
    {
      "entropy": 5.259928846359253,
      "epoch": 1.903938520653218,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046400217158987293,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2160150796175003,
      "num_tokens": 45458070.0,
      "step": 19820
    },
    {
      "entropy": 5.235132503509521,
      "epoch": 1.9044188280499519,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046398356811727626,
      "loss": 5.021,
      "mean_token_accuracy": 0.21625799983739852,
      "num_tokens": 45470829.0,
      "step": 19825
    },
    {
      "entropy": 5.226069593429566,
      "epoch": 1.9048991354466858,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004639649602570423,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22681839466094972,
      "num_tokens": 45482354.0,
      "step": 19830
    },
    {
      "entropy": 5.180108070373535,
      "epoch": 1.9053794428434196,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046394634800960314,
      "loss": 4.9047,
      "mean_token_accuracy": 0.2193937510251999,
      "num_tokens": 45493819.0,
      "step": 19835
    },
    {
      "entropy": 5.223860359191894,
      "epoch": 1.9058597502401537,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046392773137539074,
      "loss": 5.0297,
      "mean_token_accuracy": 0.21463452130556107,
      "num_tokens": 45504877.0,
      "step": 19840
    },
    {
      "entropy": 5.231951522827148,
      "epoch": 1.9063400576368876,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046390911035483744,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21593422144651414,
      "num_tokens": 45516160.0,
      "step": 19845
    },
    {
      "entropy": 5.248753881454467,
      "epoch": 1.9068203650336215,
      "grad_norm": 1.25,
      "learning_rate": 0.0004638904849483756,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21566216945648192,
      "num_tokens": 45529028.0,
      "step": 19850
    },
    {
      "entropy": 5.208504915237427,
      "epoch": 1.9073006724303554,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046387185515643756,
      "loss": 4.956,
      "mean_token_accuracy": 0.22161496728658675,
      "num_tokens": 45541507.0,
      "step": 19855
    },
    {
      "entropy": 5.32581901550293,
      "epoch": 1.9077809798270895,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004638532209794559,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20723584592342376,
      "num_tokens": 45553379.0,
      "step": 19860
    },
    {
      "entropy": 5.2136204719543455,
      "epoch": 1.9082612872238234,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004638345824178631,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2232545793056488,
      "num_tokens": 45564962.0,
      "step": 19865
    },
    {
      "entropy": 5.2839936256408695,
      "epoch": 1.9087415946205573,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046381593947209215,
      "loss": 5.0679,
      "mean_token_accuracy": 0.21257580071687698,
      "num_tokens": 45575270.0,
      "step": 19870
    },
    {
      "entropy": 5.239506244659424,
      "epoch": 1.9092219020172911,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004637972921425757,
      "loss": 4.9312,
      "mean_token_accuracy": 0.21831659525632857,
      "num_tokens": 45587446.0,
      "step": 19875
    },
    {
      "entropy": 5.306182241439819,
      "epoch": 1.909702209414025,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046377864042974675,
      "loss": 5.1012,
      "mean_token_accuracy": 0.22032327800989152,
      "num_tokens": 45599598.0,
      "step": 19880
    },
    {
      "entropy": 5.194127464294434,
      "epoch": 1.910182516810759,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004637599843340384,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21984664350748062,
      "num_tokens": 45610862.0,
      "step": 19885
    },
    {
      "entropy": 5.1968120574951175,
      "epoch": 1.9106628242074928,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046374132385588356,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22774270921945572,
      "num_tokens": 45621398.0,
      "step": 19890
    },
    {
      "entropy": 5.224905157089234,
      "epoch": 1.9111431316042267,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046372265899571576,
      "loss": 4.938,
      "mean_token_accuracy": 0.22123683094978333,
      "num_tokens": 45631139.0,
      "step": 19895
    },
    {
      "entropy": 5.242269086837768,
      "epoch": 1.9116234390009605,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046370398975396817,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21626681238412857,
      "num_tokens": 45642439.0,
      "step": 19900
    },
    {
      "entropy": 5.2423638820648195,
      "epoch": 1.9121037463976944,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004636853161310743,
      "loss": 4.9079,
      "mean_token_accuracy": 0.22206637561321257,
      "num_tokens": 45653701.0,
      "step": 19905
    },
    {
      "entropy": 5.225725173950195,
      "epoch": 1.9125840537944283,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046366663812746764,
      "loss": 4.9495,
      "mean_token_accuracy": 0.225691057741642,
      "num_tokens": 45663870.0,
      "step": 19910
    },
    {
      "entropy": 5.2265135765075685,
      "epoch": 1.9130643611911622,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004636479557435818,
      "loss": 4.962,
      "mean_token_accuracy": 0.21748601496219636,
      "num_tokens": 45674597.0,
      "step": 19915
    },
    {
      "entropy": 5.275295257568359,
      "epoch": 1.9135446685878963,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046362926897985067,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21360148191452027,
      "num_tokens": 45685871.0,
      "step": 19920
    },
    {
      "entropy": 5.230063629150391,
      "epoch": 1.9140249759846302,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000463610577836708,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21834159642457962,
      "num_tokens": 45696742.0,
      "step": 19925
    },
    {
      "entropy": 5.289859437942505,
      "epoch": 1.914505283381364,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046359188231458783,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2236787423491478,
      "num_tokens": 45708398.0,
      "step": 19930
    },
    {
      "entropy": 5.2601769924163815,
      "epoch": 1.914985590778098,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046357318241392414,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21354379504919052,
      "num_tokens": 45721525.0,
      "step": 19935
    },
    {
      "entropy": 5.142284107208252,
      "epoch": 1.915465898174832,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000463554478135151,
      "loss": 4.895,
      "mean_token_accuracy": 0.2256608635187149,
      "num_tokens": 45732089.0,
      "step": 19940
    },
    {
      "entropy": 5.240254735946655,
      "epoch": 1.915946205571566,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004635357694787029,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22077442854642867,
      "num_tokens": 45743243.0,
      "step": 19945
    },
    {
      "entropy": 5.215694808959961,
      "epoch": 1.9164265129682998,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000463517056445014,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22036272585391997,
      "num_tokens": 45754113.0,
      "step": 19950
    },
    {
      "entropy": 5.150321054458618,
      "epoch": 1.9169068203650337,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046349833903451884,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22519267350435257,
      "num_tokens": 45766190.0,
      "step": 19955
    },
    {
      "entropy": 5.260282135009765,
      "epoch": 1.9173871277617676,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046347961724765196,
      "loss": 5.0721,
      "mean_token_accuracy": 0.21615226119756697,
      "num_tokens": 45777736.0,
      "step": 19960
    },
    {
      "entropy": 5.280118942260742,
      "epoch": 1.9178674351585014,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046346089108484806,
      "loss": 5.0184,
      "mean_token_accuracy": 0.2155522570014,
      "num_tokens": 45788813.0,
      "step": 19965
    },
    {
      "entropy": 5.24834041595459,
      "epoch": 1.9183477425552353,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046344216054654193,
      "loss": 5.029,
      "mean_token_accuracy": 0.21547577232122422,
      "num_tokens": 45800671.0,
      "step": 19970
    },
    {
      "entropy": 5.217158937454224,
      "epoch": 1.9188280499519692,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046342342563316833,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21697375923395157,
      "num_tokens": 45811587.0,
      "step": 19975
    },
    {
      "entropy": 5.162120676040649,
      "epoch": 1.919308357348703,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046340468634516223,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22510828375816344,
      "num_tokens": 45823381.0,
      "step": 19980
    },
    {
      "entropy": 5.253180122375488,
      "epoch": 1.919788664745437,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046338594268295884,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21300329267978668,
      "num_tokens": 45833560.0,
      "step": 19985
    },
    {
      "entropy": 5.244262218475342,
      "epoch": 1.9202689721421708,
      "grad_norm": 1.125,
      "learning_rate": 0.0004633671946469933,
      "loss": 4.957,
      "mean_token_accuracy": 0.2232781931757927,
      "num_tokens": 45843823.0,
      "step": 19990
    },
    {
      "entropy": 5.269139957427979,
      "epoch": 1.920749279538905,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046334844223770076,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2119199424982071,
      "num_tokens": 45855122.0,
      "step": 19995
    },
    {
      "entropy": 5.236318731307984,
      "epoch": 1.9212295869356388,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046332968545551674,
      "loss": 4.9934,
      "mean_token_accuracy": 0.222648923099041,
      "num_tokens": 45866638.0,
      "step": 20000
    },
    {
      "entropy": 5.290158987045288,
      "epoch": 1.9217098943323727,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004633109243008765,
      "loss": 5.0766,
      "mean_token_accuracy": 0.20551694482564925,
      "num_tokens": 45878654.0,
      "step": 20005
    },
    {
      "entropy": 5.257811880111694,
      "epoch": 1.9221902017291066,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004632921587742159,
      "loss": 4.9742,
      "mean_token_accuracy": 0.2164493516087532,
      "num_tokens": 45889466.0,
      "step": 20010
    },
    {
      "entropy": 5.215721273422242,
      "epoch": 1.9226705091258407,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046327338887597043,
      "loss": 4.971,
      "mean_token_accuracy": 0.21727113127708436,
      "num_tokens": 45900750.0,
      "step": 20015
    },
    {
      "entropy": 5.2928773880004885,
      "epoch": 1.9231508165225746,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000463254614606576,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21045851409435273,
      "num_tokens": 45912351.0,
      "step": 20020
    },
    {
      "entropy": 5.204592275619507,
      "epoch": 1.9236311239193085,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004632358359664683,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22264819592237473,
      "num_tokens": 45923484.0,
      "step": 20025
    },
    {
      "entropy": 5.172143268585205,
      "epoch": 1.9241114313160423,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046321705295608356,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2192937910556793,
      "num_tokens": 45935512.0,
      "step": 20030
    },
    {
      "entropy": 5.186205816268921,
      "epoch": 1.9245917387127762,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046319826557585764,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21843493431806565,
      "num_tokens": 45946353.0,
      "step": 20035
    },
    {
      "entropy": 5.261382389068603,
      "epoch": 1.92507204610951,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004631794738262269,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21624341905117034,
      "num_tokens": 45958281.0,
      "step": 20040
    },
    {
      "entropy": 5.202132654190064,
      "epoch": 1.925552353506244,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004631606777076275,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21799016743898392,
      "num_tokens": 45969548.0,
      "step": 20045
    },
    {
      "entropy": 5.225641489028931,
      "epoch": 1.9260326609029779,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046314187722049587,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21340711116790773,
      "num_tokens": 45980300.0,
      "step": 20050
    },
    {
      "entropy": 5.188181304931641,
      "epoch": 1.9265129682997117,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046312307236526863,
      "loss": 4.9192,
      "mean_token_accuracy": 0.2197803482413292,
      "num_tokens": 45991280.0,
      "step": 20055
    },
    {
      "entropy": 5.333532094955444,
      "epoch": 1.9269932756964456,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046310426314238217,
      "loss": 5.0357,
      "mean_token_accuracy": 0.2104990378022194,
      "num_tokens": 46003074.0,
      "step": 20060
    },
    {
      "entropy": 5.1787947654724125,
      "epoch": 1.9274735830931795,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004630854495522733,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22118938267230986,
      "num_tokens": 46014784.0,
      "step": 20065
    },
    {
      "entropy": 5.138860416412354,
      "epoch": 1.9279538904899134,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046306663159537874,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22388557195663453,
      "num_tokens": 46025403.0,
      "step": 20070
    },
    {
      "entropy": 5.280381488800049,
      "epoch": 1.9284341978866475,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046304780927213554,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21679102778434753,
      "num_tokens": 46037171.0,
      "step": 20075
    },
    {
      "entropy": 5.312437868118286,
      "epoch": 1.9289145052833814,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046302898258298046,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21162158399820327,
      "num_tokens": 46049219.0,
      "step": 20080
    },
    {
      "entropy": 5.243570566177368,
      "epoch": 1.9293948126801153,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004630101515283509,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22686096429824829,
      "num_tokens": 46061068.0,
      "step": 20085
    },
    {
      "entropy": 5.227107334136963,
      "epoch": 1.9298751200768491,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046299131610868377,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22316648066043854,
      "num_tokens": 46073752.0,
      "step": 20090
    },
    {
      "entropy": 5.244417190551758,
      "epoch": 1.9303554274735832,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004629724763244165,
      "loss": 5.0428,
      "mean_token_accuracy": 0.2146281599998474,
      "num_tokens": 46085908.0,
      "step": 20095
    },
    {
      "entropy": 5.212504434585571,
      "epoch": 1.9308357348703171,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004629536321759866,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22637156993150712,
      "num_tokens": 46097467.0,
      "step": 20100
    },
    {
      "entropy": 5.230949926376343,
      "epoch": 1.931316042267051,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046293478366383133,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22313885539770126,
      "num_tokens": 46107951.0,
      "step": 20105
    },
    {
      "entropy": 5.2354960441589355,
      "epoch": 1.9317963496637849,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004629159307883885,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2117501378059387,
      "num_tokens": 46119862.0,
      "step": 20110
    },
    {
      "entropy": 5.193958282470703,
      "epoch": 1.9322766570605188,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004628970735500958,
      "loss": 4.9311,
      "mean_token_accuracy": 0.21831852793693543,
      "num_tokens": 46130625.0,
      "step": 20115
    },
    {
      "entropy": 5.254614400863647,
      "epoch": 1.9327569644572526,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046287821194939094,
      "loss": 4.9654,
      "mean_token_accuracy": 0.21740808933973313,
      "num_tokens": 46141819.0,
      "step": 20120
    },
    {
      "entropy": 5.2839525699615475,
      "epoch": 1.9332372718539865,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046285934598671186,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20602949410676957,
      "num_tokens": 46154474.0,
      "step": 20125
    },
    {
      "entropy": 5.282279348373413,
      "epoch": 1.9337175792507204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046284047566249665,
      "loss": 4.9658,
      "mean_token_accuracy": 0.212571220099926,
      "num_tokens": 46165397.0,
      "step": 20130
    },
    {
      "entropy": 5.167595911026001,
      "epoch": 1.9341978866474543,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046282160097718336,
      "loss": 4.939,
      "mean_token_accuracy": 0.221282397210598,
      "num_tokens": 46177202.0,
      "step": 20135
    },
    {
      "entropy": 5.147082805633545,
      "epoch": 1.9346781940441882,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004628027219312102,
      "loss": 4.969,
      "mean_token_accuracy": 0.21722161918878555,
      "num_tokens": 46189879.0,
      "step": 20140
    },
    {
      "entropy": 5.3384874820709225,
      "epoch": 1.935158501440922,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004627838385250155,
      "loss": 5.0665,
      "mean_token_accuracy": 0.21013033986091614,
      "num_tokens": 46201788.0,
      "step": 20145
    },
    {
      "entropy": 5.284536361694336,
      "epoch": 1.9356388088376562,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046276495075903764,
      "loss": 4.9815,
      "mean_token_accuracy": 0.22138524949550628,
      "num_tokens": 46212316.0,
      "step": 20150
    },
    {
      "entropy": 5.193808126449585,
      "epoch": 1.93611911623439,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046274605863371517,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22330798506736754,
      "num_tokens": 46223192.0,
      "step": 20155
    },
    {
      "entropy": 5.269413328170776,
      "epoch": 1.936599423631124,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004627271621494868,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21411036252975463,
      "num_tokens": 46234404.0,
      "step": 20160
    },
    {
      "entropy": 5.266096305847168,
      "epoch": 1.9370797310278578,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000462708261306791,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21357710659503937,
      "num_tokens": 46246653.0,
      "step": 20165
    },
    {
      "entropy": 5.239640474319458,
      "epoch": 1.937560038424592,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004626893561060669,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2110196530818939,
      "num_tokens": 46258145.0,
      "step": 20170
    },
    {
      "entropy": 5.231673812866211,
      "epoch": 1.9380403458213258,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046267044654775324,
      "loss": 4.9337,
      "mean_token_accuracy": 0.221611824631691,
      "num_tokens": 46269021.0,
      "step": 20175
    },
    {
      "entropy": 5.139617204666138,
      "epoch": 1.9385206532180597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000462651532632289,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21917923241853715,
      "num_tokens": 46280729.0,
      "step": 20180
    },
    {
      "entropy": 5.204895162582398,
      "epoch": 1.9390009606147935,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046263261436011344,
      "loss": 4.9998,
      "mean_token_accuracy": 0.21340069323778152,
      "num_tokens": 46291999.0,
      "step": 20185
    },
    {
      "entropy": 5.260077476501465,
      "epoch": 1.9394812680115274,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004626136917316657,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21250567585229874,
      "num_tokens": 46302740.0,
      "step": 20190
    },
    {
      "entropy": 5.2593145847320555,
      "epoch": 1.9399615754082613,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046259476474738514,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22178202122449875,
      "num_tokens": 46313700.0,
      "step": 20195
    },
    {
      "entropy": 5.220777273178101,
      "epoch": 1.9404418828049952,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046257583340771123,
      "loss": 4.9851,
      "mean_token_accuracy": 0.21660949736833573,
      "num_tokens": 46324445.0,
      "step": 20200
    },
    {
      "entropy": 5.098165607452392,
      "epoch": 1.940922190201729,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004625568977130835,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2166528195142746,
      "num_tokens": 46335192.0,
      "step": 20205
    },
    {
      "entropy": 5.2699487686157225,
      "epoch": 1.941402497598463,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004625379576639414,
      "loss": 5.0009,
      "mean_token_accuracy": 0.21568366140127182,
      "num_tokens": 46346067.0,
      "step": 20210
    },
    {
      "entropy": 5.228092908859253,
      "epoch": 1.9418828049951968,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046251901326072487,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22167645394802094,
      "num_tokens": 46356736.0,
      "step": 20215
    },
    {
      "entropy": 5.206680345535278,
      "epoch": 1.9423631123919307,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046250006450387367,
      "loss": 4.9791,
      "mean_token_accuracy": 0.22064134925603868,
      "num_tokens": 46368250.0,
      "step": 20220
    },
    {
      "entropy": 5.20731372833252,
      "epoch": 1.9428434197886646,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004624811113938277,
      "loss": 5.0705,
      "mean_token_accuracy": 0.21059169471263886,
      "num_tokens": 46379662.0,
      "step": 20225
    },
    {
      "entropy": 5.212683582305909,
      "epoch": 1.9433237271853987,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004624621539310271,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2216317966580391,
      "num_tokens": 46391485.0,
      "step": 20230
    },
    {
      "entropy": 5.29567518234253,
      "epoch": 1.9438040345821326,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046244319211591193,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20684416145086287,
      "num_tokens": 46403608.0,
      "step": 20235
    },
    {
      "entropy": 5.230904674530029,
      "epoch": 1.9442843419788665,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004624242259489223,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2222265288233757,
      "num_tokens": 46416154.0,
      "step": 20240
    },
    {
      "entropy": 5.270880174636841,
      "epoch": 1.9447646493756003,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046240525543049884,
      "loss": 4.9672,
      "mean_token_accuracy": 0.2181414470076561,
      "num_tokens": 46427654.0,
      "step": 20245
    },
    {
      "entropy": 5.232746648788452,
      "epoch": 1.9452449567723344,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046238628056108176,
      "loss": 4.941,
      "mean_token_accuracy": 0.21832856237888337,
      "num_tokens": 46438074.0,
      "step": 20250
    },
    {
      "entropy": 5.144287347793579,
      "epoch": 1.9457252641690683,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046236730134111166,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22703658044338226,
      "num_tokens": 46449002.0,
      "step": 20255
    },
    {
      "entropy": 5.222492122650147,
      "epoch": 1.9462055715658022,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004623483177710291,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21719553619623183,
      "num_tokens": 46460576.0,
      "step": 20260
    },
    {
      "entropy": 5.305343151092529,
      "epoch": 1.946685878962536,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004623293298512751,
      "loss": 5.0059,
      "mean_token_accuracy": 0.217582044005394,
      "num_tokens": 46471636.0,
      "step": 20265
    },
    {
      "entropy": 5.1289817810058596,
      "epoch": 1.94716618635927,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046231033758229026,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22269021719694138,
      "num_tokens": 46482364.0,
      "step": 20270
    },
    {
      "entropy": 5.227625894546509,
      "epoch": 1.9476464937560038,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004622913409645154,
      "loss": 5.0209,
      "mean_token_accuracy": 0.21284470707178116,
      "num_tokens": 46493612.0,
      "step": 20275
    },
    {
      "entropy": 5.176160478591919,
      "epoch": 1.9481268011527377,
      "grad_norm": 1.0,
      "learning_rate": 0.0004622723399983919,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21819289177656173,
      "num_tokens": 46504825.0,
      "step": 20280
    },
    {
      "entropy": 5.215923738479614,
      "epoch": 1.9486071085494716,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046225333468436077,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2214005023241043,
      "num_tokens": 46515094.0,
      "step": 20285
    },
    {
      "entropy": 5.261217164993286,
      "epoch": 1.9490874159462055,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046223432502286323,
      "loss": 4.9966,
      "mean_token_accuracy": 0.20903264880180358,
      "num_tokens": 46527956.0,
      "step": 20290
    },
    {
      "entropy": 5.183873271942138,
      "epoch": 1.9495677233429394,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046221531101434056,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2210679292678833,
      "num_tokens": 46539446.0,
      "step": 20295
    },
    {
      "entropy": 5.301405906677246,
      "epoch": 1.9500480307396733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004621962926592343,
      "loss": 5.0162,
      "mean_token_accuracy": 0.22003831714391708,
      "num_tokens": 46551008.0,
      "step": 20300
    },
    {
      "entropy": 5.114235162734985,
      "epoch": 1.9505283381364071,
      "grad_norm": 1.109375,
      "learning_rate": 0.000462177269957986,
      "loss": 4.8535,
      "mean_token_accuracy": 0.2182070106267929,
      "num_tokens": 46563904.0,
      "step": 20305
    },
    {
      "entropy": 5.22625150680542,
      "epoch": 1.9510086455331412,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004621582429110373,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21297204941511155,
      "num_tokens": 46575065.0,
      "step": 20310
    },
    {
      "entropy": 5.174101877212524,
      "epoch": 1.9514889529298751,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046213921151883,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2228015273809433,
      "num_tokens": 46587683.0,
      "step": 20315
    },
    {
      "entropy": 5.196189498901367,
      "epoch": 1.951969260326609,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004621201757818059,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22439981997013092,
      "num_tokens": 46600060.0,
      "step": 20320
    },
    {
      "entropy": 5.236166524887085,
      "epoch": 1.952449567723343,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046210113570040683,
      "loss": 5.0113,
      "mean_token_accuracy": 0.21538367867469788,
      "num_tokens": 46611619.0,
      "step": 20325
    },
    {
      "entropy": 5.151376247406006,
      "epoch": 1.952929875120077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000462082091275075,
      "loss": 4.8687,
      "mean_token_accuracy": 0.2240465000271797,
      "num_tokens": 46622286.0,
      "step": 20330
    },
    {
      "entropy": 5.24801778793335,
      "epoch": 1.9534101825168109,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004620630425062526,
      "loss": 5.0344,
      "mean_token_accuracy": 0.20863564908504487,
      "num_tokens": 46633402.0,
      "step": 20335
    },
    {
      "entropy": 5.128415727615357,
      "epoch": 1.9538904899135447,
      "grad_norm": 1.375,
      "learning_rate": 0.0004620439893943817,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22660980522632598,
      "num_tokens": 46644889.0,
      "step": 20340
    },
    {
      "entropy": 5.254168319702148,
      "epoch": 1.9543707973102786,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004620249319399049,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22341285943984984,
      "num_tokens": 46656355.0,
      "step": 20345
    },
    {
      "entropy": 5.280762672424316,
      "epoch": 1.9548511047070125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046200587014326455,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21311995536088943,
      "num_tokens": 46668570.0,
      "step": 20350
    },
    {
      "entropy": 5.203006553649902,
      "epoch": 1.9553314121037464,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004619868040049031,
      "loss": 4.9448,
      "mean_token_accuracy": 0.21671674102544786,
      "num_tokens": 46680015.0,
      "step": 20355
    },
    {
      "entropy": 5.2284801483154295,
      "epoch": 1.9558117195004803,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004619677335252633,
      "loss": 4.9961,
      "mean_token_accuracy": 0.2137531265616417,
      "num_tokens": 46690872.0,
      "step": 20360
    },
    {
      "entropy": 5.28906364440918,
      "epoch": 1.9562920268972142,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046194865870478793,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21636375933885574,
      "num_tokens": 46702760.0,
      "step": 20365
    },
    {
      "entropy": 5.207029581069946,
      "epoch": 1.956772334293948,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046192957954391983,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20954471677541733,
      "num_tokens": 46715038.0,
      "step": 20370
    },
    {
      "entropy": 5.232690286636353,
      "epoch": 1.957252641690682,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000461910496043102,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2192747637629509,
      "num_tokens": 46725851.0,
      "step": 20375
    },
    {
      "entropy": 5.1992024898529055,
      "epoch": 1.9577329490874158,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004618914082027773,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2236976593732834,
      "num_tokens": 46736380.0,
      "step": 20380
    },
    {
      "entropy": 5.145764255523682,
      "epoch": 1.95821325648415,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046187231602338926,
      "loss": 4.8714,
      "mean_token_accuracy": 0.22642964422702788,
      "num_tokens": 46747872.0,
      "step": 20385
    },
    {
      "entropy": 5.19242582321167,
      "epoch": 1.9586935638808838,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046185321950538086,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2163282200694084,
      "num_tokens": 46759704.0,
      "step": 20390
    },
    {
      "entropy": 5.244603681564331,
      "epoch": 1.9591738712776177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004618341186491955,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22220176905393602,
      "num_tokens": 46770895.0,
      "step": 20395
    },
    {
      "entropy": 5.247511625289917,
      "epoch": 1.9596541786743515,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004618150134552768,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21735584884881973,
      "num_tokens": 46781190.0,
      "step": 20400
    },
    {
      "entropy": 5.166856193542481,
      "epoch": 1.9601344860710856,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000461795903924068,
      "loss": 4.9482,
      "mean_token_accuracy": 0.22068553119897844,
      "num_tokens": 46792172.0,
      "step": 20405
    },
    {
      "entropy": 5.232017898559571,
      "epoch": 1.9606147934678195,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046177679005601313,
      "loss": 5.0399,
      "mean_token_accuracy": 0.20908942371606826,
      "num_tokens": 46804189.0,
      "step": 20410
    },
    {
      "entropy": 5.3066754817962645,
      "epoch": 1.9610951008645534,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004617576718515558,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21398459672927855,
      "num_tokens": 46815127.0,
      "step": 20415
    },
    {
      "entropy": 5.198750305175781,
      "epoch": 1.9615754082612873,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004617385493111399,
      "loss": 4.9796,
      "mean_token_accuracy": 0.21238848865032195,
      "num_tokens": 46825927.0,
      "step": 20420
    },
    {
      "entropy": 5.260508251190186,
      "epoch": 1.9620557156580212,
      "grad_norm": 1.125,
      "learning_rate": 0.0004617194224352093,
      "loss": 5.0586,
      "mean_token_accuracy": 0.21188971847295762,
      "num_tokens": 46836689.0,
      "step": 20425
    },
    {
      "entropy": 5.192009449005127,
      "epoch": 1.962536023054755,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004617002912242083,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22751171737909318,
      "num_tokens": 46847650.0,
      "step": 20430
    },
    {
      "entropy": 5.155014514923096,
      "epoch": 1.963016330451489,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046168115567858084,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2239232048392296,
      "num_tokens": 46859469.0,
      "step": 20435
    },
    {
      "entropy": 5.197812843322754,
      "epoch": 1.9634966378482228,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046166201579877125,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2259680837392807,
      "num_tokens": 46869583.0,
      "step": 20440
    },
    {
      "entropy": 5.271603965759278,
      "epoch": 1.9639769452449567,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004616428715852239,
      "loss": 4.9413,
      "mean_token_accuracy": 0.2225704088807106,
      "num_tokens": 46881357.0,
      "step": 20445
    },
    {
      "entropy": 5.139634609222412,
      "epoch": 1.9644572526416906,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004616237230383833,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21683990359306335,
      "num_tokens": 46893627.0,
      "step": 20450
    },
    {
      "entropy": 5.222458982467652,
      "epoch": 1.9649375600384245,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046160457015869414,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20801270008087158,
      "num_tokens": 46904621.0,
      "step": 20455
    },
    {
      "entropy": 5.308166980743408,
      "epoch": 1.9654178674351583,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046158541294660083,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21600136607885362,
      "num_tokens": 46914381.0,
      "step": 20460
    },
    {
      "entropy": 5.260594320297241,
      "epoch": 1.9658981748318924,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004615662514025483,
      "loss": 4.934,
      "mean_token_accuracy": 0.22015719562768937,
      "num_tokens": 46925139.0,
      "step": 20465
    },
    {
      "entropy": 5.213877439498901,
      "epoch": 1.9663784822286263,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046154708552698147,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21449769139289857,
      "num_tokens": 46937638.0,
      "step": 20470
    },
    {
      "entropy": 5.1656208515167235,
      "epoch": 1.9668587896253602,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046152791532034517,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21671010106801986,
      "num_tokens": 46949460.0,
      "step": 20475
    },
    {
      "entropy": 5.297852087020874,
      "epoch": 1.967339097022094,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046150874078308463,
      "loss": 5.0187,
      "mean_token_accuracy": 0.21506672203540803,
      "num_tokens": 46961047.0,
      "step": 20480
    },
    {
      "entropy": 5.20629997253418,
      "epoch": 1.9678194044188282,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004614895619156449,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22329482287168503,
      "num_tokens": 46972009.0,
      "step": 20485
    },
    {
      "entropy": 5.081929922103882,
      "epoch": 1.968299711815562,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046147037871847125,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22288259714841843,
      "num_tokens": 46982911.0,
      "step": 20490
    },
    {
      "entropy": 5.247358798980713,
      "epoch": 1.968780019212296,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004614511911920092,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21515191048383714,
      "num_tokens": 46995096.0,
      "step": 20495
    },
    {
      "entropy": 5.34033169746399,
      "epoch": 1.9692603266090298,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046143199933670404,
      "loss": 5.0406,
      "mean_token_accuracy": 0.2163153275847435,
      "num_tokens": 47006814.0,
      "step": 20500
    },
    {
      "entropy": 5.153030776977539,
      "epoch": 1.9697406340057637,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004614128031530015,
      "loss": 4.864,
      "mean_token_accuracy": 0.22274877279996871,
      "num_tokens": 47018204.0,
      "step": 20505
    },
    {
      "entropy": 5.143307733535766,
      "epoch": 1.9702209414024976,
      "grad_norm": 1.25,
      "learning_rate": 0.00046139360264134724,
      "loss": 4.9746,
      "mean_token_accuracy": 0.22338173687458038,
      "num_tokens": 47029655.0,
      "step": 20510
    },
    {
      "entropy": 5.267326736450196,
      "epoch": 1.9707012487992315,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004613743978021869,
      "loss": 5.0555,
      "mean_token_accuracy": 0.21485082805156708,
      "num_tokens": 47041830.0,
      "step": 20515
    },
    {
      "entropy": 5.259827470779419,
      "epoch": 1.9711815561959654,
      "grad_norm": 1.125,
      "learning_rate": 0.00046135518863596654,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22404208779335022,
      "num_tokens": 47052817.0,
      "step": 20520
    },
    {
      "entropy": 5.225026941299438,
      "epoch": 1.9716618635926992,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046133597514313204,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2164018616080284,
      "num_tokens": 47064452.0,
      "step": 20525
    },
    {
      "entropy": 5.249282932281494,
      "epoch": 1.9721421709894331,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004613167573241295,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21002217680215834,
      "num_tokens": 47074952.0,
      "step": 20530
    },
    {
      "entropy": 5.265222644805908,
      "epoch": 1.972622478386167,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004612975351794051,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21633655428886414,
      "num_tokens": 47085496.0,
      "step": 20535
    },
    {
      "entropy": 5.237248182296753,
      "epoch": 1.973102785782901,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004612783087094051,
      "loss": 5.0149,
      "mean_token_accuracy": 0.2143534928560257,
      "num_tokens": 47097798.0,
      "step": 20540
    },
    {
      "entropy": 5.299124765396118,
      "epoch": 1.973583093179635,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046125907791457594,
      "loss": 5.0607,
      "mean_token_accuracy": 0.21407242268323898,
      "num_tokens": 47109798.0,
      "step": 20545
    },
    {
      "entropy": 5.285663080215454,
      "epoch": 1.9740634005763689,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046123984279536405,
      "loss": 4.9764,
      "mean_token_accuracy": 0.22138723433017732,
      "num_tokens": 47120073.0,
      "step": 20550
    },
    {
      "entropy": 5.210918378829956,
      "epoch": 1.9745437079731027,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046122060335221604,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21661639213562012,
      "num_tokens": 47130596.0,
      "step": 20555
    },
    {
      "entropy": 5.125567150115967,
      "epoch": 1.9750240153698368,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046120135958557855,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23451912999153138,
      "num_tokens": 47141660.0,
      "step": 20560
    },
    {
      "entropy": 5.379381418228149,
      "epoch": 1.9755043227665707,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046118211149589843,
      "loss": 5.1202,
      "mean_token_accuracy": 0.20049034357070922,
      "num_tokens": 47153046.0,
      "step": 20565
    },
    {
      "entropy": 5.325334739685059,
      "epoch": 1.9759846301633046,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004611628590836225,
      "loss": 5.0752,
      "mean_token_accuracy": 0.2129594087600708,
      "num_tokens": 47162706.0,
      "step": 20570
    },
    {
      "entropy": 5.221619033813477,
      "epoch": 1.9764649375600385,
      "grad_norm": 1.25,
      "learning_rate": 0.0004611436023491979,
      "loss": 4.9842,
      "mean_token_accuracy": 0.2195766717195511,
      "num_tokens": 47174508.0,
      "step": 20575
    },
    {
      "entropy": 5.215278387069702,
      "epoch": 1.9769452449567724,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004611243412930714,
      "loss": 4.8873,
      "mean_token_accuracy": 0.22413036674261094,
      "num_tokens": 47185019.0,
      "step": 20580
    },
    {
      "entropy": 5.199540710449218,
      "epoch": 1.9774255523535063,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046110507591569047,
      "loss": 4.928,
      "mean_token_accuracy": 0.22358032464981079,
      "num_tokens": 47195538.0,
      "step": 20585
    },
    {
      "entropy": 5.264446973800659,
      "epoch": 1.9779058597502401,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004610858062175023,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22034751921892165,
      "num_tokens": 47206398.0,
      "step": 20590
    },
    {
      "entropy": 5.215377712249756,
      "epoch": 1.978386167146974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046106653219895417,
      "loss": 5.0416,
      "mean_token_accuracy": 0.2078413888812065,
      "num_tokens": 47217775.0,
      "step": 20595
    },
    {
      "entropy": 5.252941083908081,
      "epoch": 1.978866474543708,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004610472538604938,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21355969607830047,
      "num_tokens": 47230179.0,
      "step": 20600
    },
    {
      "entropy": 5.196918201446533,
      "epoch": 1.9793467819404418,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046102797120256854,
      "loss": 4.8686,
      "mean_token_accuracy": 0.23061528503894807,
      "num_tokens": 47241416.0,
      "step": 20605
    },
    {
      "entropy": 5.147827625274658,
      "epoch": 1.9798270893371757,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004610086842256262,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2199065774679184,
      "num_tokens": 47253916.0,
      "step": 20610
    },
    {
      "entropy": 5.275148630142212,
      "epoch": 1.9803073967339095,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004609893929301146,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21208497732877732,
      "num_tokens": 47265176.0,
      "step": 20615
    },
    {
      "entropy": 5.273549270629883,
      "epoch": 1.9807877041306436,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004609700973164816,
      "loss": 4.919,
      "mean_token_accuracy": 0.2220047265291214,
      "num_tokens": 47276178.0,
      "step": 20620
    },
    {
      "entropy": 5.139065599441528,
      "epoch": 1.9812680115273775,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004609507973851751,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2157987102866173,
      "num_tokens": 47287607.0,
      "step": 20625
    },
    {
      "entropy": 5.245670461654663,
      "epoch": 1.9817483189241114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046093149313664316,
      "loss": 5.0847,
      "mean_token_accuracy": 0.21333062052726745,
      "num_tokens": 47300069.0,
      "step": 20630
    },
    {
      "entropy": 5.264751720428467,
      "epoch": 1.9822286263208453,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004609121845713342,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22011056244373323,
      "num_tokens": 47312349.0,
      "step": 20635
    },
    {
      "entropy": 5.156843996047973,
      "epoch": 1.9827089337175794,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004608928716896963,
      "loss": 4.9381,
      "mean_token_accuracy": 0.22272602915763856,
      "num_tokens": 47323171.0,
      "step": 20640
    },
    {
      "entropy": 5.245606708526611,
      "epoch": 1.9831892411143133,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004608735544921778,
      "loss": 5.005,
      "mean_token_accuracy": 0.21766031235456468,
      "num_tokens": 47333990.0,
      "step": 20645
    },
    {
      "entropy": 5.171510219573975,
      "epoch": 1.9836695485110472,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046085423297922745,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21861768513917923,
      "num_tokens": 47346081.0,
      "step": 20650
    },
    {
      "entropy": 5.3294299125671385,
      "epoch": 1.984149855907781,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046083490715129367,
      "loss": 5.1431,
      "mean_token_accuracy": 0.2091592252254486,
      "num_tokens": 47357979.0,
      "step": 20655
    },
    {
      "entropy": 5.3960693836212155,
      "epoch": 1.984630163304515,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004608155770088251,
      "loss": 5.0449,
      "mean_token_accuracy": 0.20639441311359405,
      "num_tokens": 47369066.0,
      "step": 20660
    },
    {
      "entropy": 5.236590623855591,
      "epoch": 1.9851104707012488,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046079624255227066,
      "loss": 4.9812,
      "mean_token_accuracy": 0.21744994670152665,
      "num_tokens": 47380316.0,
      "step": 20665
    },
    {
      "entropy": 5.180174732208252,
      "epoch": 1.9855907780979827,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004607769037820791,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21243722438812257,
      "num_tokens": 47392040.0,
      "step": 20670
    },
    {
      "entropy": 5.2457887649536135,
      "epoch": 1.9860710854947166,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004607575606986995,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21983893513679503,
      "num_tokens": 47404114.0,
      "step": 20675
    },
    {
      "entropy": 5.211176919937134,
      "epoch": 1.9865513928914504,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004607382133025809,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2170942336320877,
      "num_tokens": 47416581.0,
      "step": 20680
    },
    {
      "entropy": 5.188728141784668,
      "epoch": 1.9870317002881843,
      "grad_norm": 1.125,
      "learning_rate": 0.00046071886159417257,
      "loss": 4.8463,
      "mean_token_accuracy": 0.22798649966716766,
      "num_tokens": 47427314.0,
      "step": 20685
    },
    {
      "entropy": 5.244807147979737,
      "epoch": 1.9875120076849182,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004606995055739238,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20856667757034303,
      "num_tokens": 47438672.0,
      "step": 20690
    },
    {
      "entropy": 5.270337772369385,
      "epoch": 1.9879923150816523,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046068014524228374,
      "loss": 5.043,
      "mean_token_accuracy": 0.21193131804466248,
      "num_tokens": 47449901.0,
      "step": 20695
    },
    {
      "entropy": 5.243450164794922,
      "epoch": 1.9884726224783862,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046066078059970217,
      "loss": 5.0454,
      "mean_token_accuracy": 0.20830730795860292,
      "num_tokens": 47462831.0,
      "step": 20700
    },
    {
      "entropy": 5.217765283584595,
      "epoch": 1.98895292987512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004606414116466286,
      "loss": 4.9817,
      "mean_token_accuracy": 0.22152943760156632,
      "num_tokens": 47474075.0,
      "step": 20705
    },
    {
      "entropy": 5.255944490432739,
      "epoch": 1.989433237271854,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046062203838351267,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20782520323991777,
      "num_tokens": 47486893.0,
      "step": 20710
    },
    {
      "entropy": 5.249710369110107,
      "epoch": 1.989913544668588,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046060266081080414,
      "loss": 4.9858,
      "mean_token_accuracy": 0.21065388470888138,
      "num_tokens": 47497969.0,
      "step": 20715
    },
    {
      "entropy": 5.344042730331421,
      "epoch": 1.990393852065322,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004605832789289531,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21749683618545532,
      "num_tokens": 47509476.0,
      "step": 20720
    },
    {
      "entropy": 5.191856575012207,
      "epoch": 1.9908741594620558,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004605638927384093,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22442605793476106,
      "num_tokens": 47520124.0,
      "step": 20725
    },
    {
      "entropy": 5.189966869354248,
      "epoch": 1.9913544668587897,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046054450223962284,
      "loss": 4.917,
      "mean_token_accuracy": 0.22332502007484437,
      "num_tokens": 47532079.0,
      "step": 20730
    },
    {
      "entropy": 5.277091789245605,
      "epoch": 1.9918347742555236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046052510743304405,
      "loss": 4.9663,
      "mean_token_accuracy": 0.2243661642074585,
      "num_tokens": 47544142.0,
      "step": 20735
    },
    {
      "entropy": 5.231726169586182,
      "epoch": 1.9923150816522575,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004605057083191232,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21689167618751526,
      "num_tokens": 47555688.0,
      "step": 20740
    },
    {
      "entropy": 5.203685617446899,
      "epoch": 1.9927953890489913,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004604863048983106,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22286435812711716,
      "num_tokens": 47566590.0,
      "step": 20745
    },
    {
      "entropy": 5.236692667007446,
      "epoch": 1.9932756964457252,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004604668971710568,
      "loss": 4.987,
      "mean_token_accuracy": 0.22076989114284515,
      "num_tokens": 47577781.0,
      "step": 20750
    },
    {
      "entropy": 5.2157630443573,
      "epoch": 1.993756003842459,
      "grad_norm": 1.25,
      "learning_rate": 0.0004604474851378124,
      "loss": 4.9088,
      "mean_token_accuracy": 0.23151296228170395,
      "num_tokens": 47589074.0,
      "step": 20755
    },
    {
      "entropy": 5.194499540328979,
      "epoch": 1.994236311239193,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046042806879902803,
      "loss": 4.9502,
      "mean_token_accuracy": 0.22196974456310273,
      "num_tokens": 47600841.0,
      "step": 20760
    },
    {
      "entropy": 5.272843742370606,
      "epoch": 1.9947166186359269,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004604086481551546,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21806015074253082,
      "num_tokens": 47612332.0,
      "step": 20765
    },
    {
      "entropy": 5.21380877494812,
      "epoch": 1.9951969260326607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004603892232066428,
      "loss": 4.9486,
      "mean_token_accuracy": 0.216632941365242,
      "num_tokens": 47623773.0,
      "step": 20770
    },
    {
      "entropy": 5.1869062900543215,
      "epoch": 1.9956772334293948,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046036979395394374,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22383994460105897,
      "num_tokens": 47634802.0,
      "step": 20775
    },
    {
      "entropy": 5.276925277709961,
      "epoch": 1.9961575408261287,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004603503603975085,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20876845866441726,
      "num_tokens": 47646772.0,
      "step": 20780
    },
    {
      "entropy": 5.306274271011352,
      "epoch": 1.9966378482228626,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004603309225377883,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21634097248315812,
      "num_tokens": 47658321.0,
      "step": 20785
    },
    {
      "entropy": 5.135401010513306,
      "epoch": 1.9971181556195965,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004603114803752344,
      "loss": 4.8964,
      "mean_token_accuracy": 0.2289227217435837,
      "num_tokens": 47668996.0,
      "step": 20790
    },
    {
      "entropy": 5.165858888626099,
      "epoch": 1.9975984630163306,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046029203391029813,
      "loss": 4.935,
      "mean_token_accuracy": 0.22385275214910508,
      "num_tokens": 47679091.0,
      "step": 20795
    },
    {
      "entropy": 5.220847845077515,
      "epoch": 1.9980787704130645,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046027258314343107,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22134607136249543,
      "num_tokens": 47690740.0,
      "step": 20800
    },
    {
      "entropy": 5.188599634170532,
      "epoch": 1.9985590778097984,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046025312807508487,
      "loss": 4.9328,
      "mean_token_accuracy": 0.22074706554412843,
      "num_tokens": 47703514.0,
      "step": 20805
    },
    {
      "entropy": 5.199264478683472,
      "epoch": 1.9990393852065322,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046023366870571097,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21952833831310273,
      "num_tokens": 47715220.0,
      "step": 20810
    },
    {
      "entropy": 5.234171533584595,
      "epoch": 1.9995196926032661,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046021420503576145,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21852201074361802,
      "num_tokens": 47727387.0,
      "step": 20815
    },
    {
      "entropy": 5.257473945617676,
      "epoch": 2.0,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004601947370656879,
      "loss": 5.0447,
      "mean_token_accuracy": 0.2101148009300232,
      "num_tokens": 47737072.0,
      "step": 20820
    },
    {
      "entropy": 5.207997894287109,
      "epoch": 2.000480307396734,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004601752647959426,
      "loss": 4.8623,
      "mean_token_accuracy": 0.23113146871328355,
      "num_tokens": 47748516.0,
      "step": 20825
    },
    {
      "entropy": 5.197933101654053,
      "epoch": 2.0009606147934678,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004601557882269775,
      "loss": 4.8328,
      "mean_token_accuracy": 0.22704352885484697,
      "num_tokens": 47760482.0,
      "step": 20830
    },
    {
      "entropy": 5.219780445098877,
      "epoch": 2.0014409221902016,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004601363073592447,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2278415709733963,
      "num_tokens": 47771204.0,
      "step": 20835
    },
    {
      "entropy": 5.13938045501709,
      "epoch": 2.0019212295869355,
      "grad_norm": 1.125,
      "learning_rate": 0.0004601168221931967,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22653318345546722,
      "num_tokens": 47782404.0,
      "step": 20840
    },
    {
      "entropy": 5.207552242279053,
      "epoch": 2.0024015369836694,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004600973327292857,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2204338029026985,
      "num_tokens": 47793612.0,
      "step": 20845
    },
    {
      "entropy": 5.23489236831665,
      "epoch": 2.0028818443804033,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046007783896796436,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22741975635290146,
      "num_tokens": 47804467.0,
      "step": 20850
    },
    {
      "entropy": 5.301383066177368,
      "epoch": 2.003362151777137,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004600583409096851,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21438181400299072,
      "num_tokens": 47815654.0,
      "step": 20855
    },
    {
      "entropy": 5.248008298873901,
      "epoch": 2.0038424591738715,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046003883855490066,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21798405796289444,
      "num_tokens": 47827630.0,
      "step": 20860
    },
    {
      "entropy": 5.1771644115447994,
      "epoch": 2.0043227665706054,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004600193319040638,
      "loss": 4.861,
      "mean_token_accuracy": 0.22447020262479783,
      "num_tokens": 47838046.0,
      "step": 20865
    },
    {
      "entropy": 5.24453649520874,
      "epoch": 2.0048030739673393,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045999982095762756,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2169887378811836,
      "num_tokens": 47849506.0,
      "step": 20870
    },
    {
      "entropy": 5.143292188644409,
      "epoch": 2.005283381364073,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045998030571604473,
      "loss": 4.7993,
      "mean_token_accuracy": 0.23068008273839952,
      "num_tokens": 47860689.0,
      "step": 20875
    },
    {
      "entropy": 5.1915336608886715,
      "epoch": 2.005763688760807,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004599607861797685,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22547078877687454,
      "num_tokens": 47872215.0,
      "step": 20880
    },
    {
      "entropy": 5.226893568038941,
      "epoch": 2.006243996157541,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045994126234925203,
      "loss": 5.0161,
      "mean_token_accuracy": 0.213435760140419,
      "num_tokens": 47883726.0,
      "step": 20885
    },
    {
      "entropy": 5.251782083511353,
      "epoch": 2.006724303554275,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045992173422494865,
      "loss": 4.8882,
      "mean_token_accuracy": 0.22214985638856888,
      "num_tokens": 47893915.0,
      "step": 20890
    },
    {
      "entropy": 5.216142845153809,
      "epoch": 2.0072046109510087,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004599022018073117,
      "loss": 4.976,
      "mean_token_accuracy": 0.22134243547916413,
      "num_tokens": 47905026.0,
      "step": 20895
    },
    {
      "entropy": 5.281191873550415,
      "epoch": 2.0076849183477425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004598826650967946,
      "loss": 4.97,
      "mean_token_accuracy": 0.2161965399980545,
      "num_tokens": 47918410.0,
      "step": 20900
    },
    {
      "entropy": 5.303548383712768,
      "epoch": 2.0081652257444764,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045986312409385105,
      "loss": 4.9481,
      "mean_token_accuracy": 0.21744633316993714,
      "num_tokens": 47929584.0,
      "step": 20905
    },
    {
      "entropy": 5.1761274337768555,
      "epoch": 2.0086455331412103,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004598435787989347,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2186829552054405,
      "num_tokens": 47940585.0,
      "step": 20910
    },
    {
      "entropy": 5.242428588867187,
      "epoch": 2.009125840537944,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045982402921249934,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2119331955909729,
      "num_tokens": 47953951.0,
      "step": 20915
    },
    {
      "entropy": 5.152158927917481,
      "epoch": 2.009606147934678,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004598044753349988,
      "loss": 4.8275,
      "mean_token_accuracy": 0.22737876176834107,
      "num_tokens": 47965783.0,
      "step": 20920
    },
    {
      "entropy": 5.149895095825196,
      "epoch": 2.010086455331412,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045978491716688706,
      "loss": 4.8678,
      "mean_token_accuracy": 0.22113073617219925,
      "num_tokens": 47978509.0,
      "step": 20925
    },
    {
      "entropy": 5.303455400466919,
      "epoch": 2.010566762728146,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004597653547086184,
      "loss": 4.9717,
      "mean_token_accuracy": 0.22181342989206315,
      "num_tokens": 47989828.0,
      "step": 20930
    },
    {
      "entropy": 5.196484088897705,
      "epoch": 2.0110470701248797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004597457879606467,
      "loss": 4.8303,
      "mean_token_accuracy": 0.22119663208723067,
      "num_tokens": 47999683.0,
      "step": 20935
    },
    {
      "entropy": 5.085349369049072,
      "epoch": 2.011527377521614,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045972621692342636,
      "loss": 4.7571,
      "mean_token_accuracy": 0.22735804617404937,
      "num_tokens": 48010526.0,
      "step": 20940
    },
    {
      "entropy": 5.116644239425659,
      "epoch": 2.012007684918348,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045970664159741186,
      "loss": 4.7949,
      "mean_token_accuracy": 0.23714192658662797,
      "num_tokens": 48021383.0,
      "step": 20945
    },
    {
      "entropy": 5.124062442779541,
      "epoch": 2.012487992315082,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045968706198305765,
      "loss": 4.7235,
      "mean_token_accuracy": 0.23369904458522797,
      "num_tokens": 48031459.0,
      "step": 20950
    },
    {
      "entropy": 5.282777738571167,
      "epoch": 2.0129682997118157,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045966747808081824,
      "loss": 4.9903,
      "mean_token_accuracy": 0.21157298386096954,
      "num_tokens": 48043590.0,
      "step": 20955
    },
    {
      "entropy": 5.154452800750732,
      "epoch": 2.0134486071085496,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004596478898911483,
      "loss": 4.7862,
      "mean_token_accuracy": 0.2291048511862755,
      "num_tokens": 48054715.0,
      "step": 20960
    },
    {
      "entropy": 5.160197448730469,
      "epoch": 2.0139289145052834,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045962829741450265,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22040279507637023,
      "num_tokens": 48065606.0,
      "step": 20965
    },
    {
      "entropy": 5.237830448150635,
      "epoch": 2.0144092219020173,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004596087006513361,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22034614235162736,
      "num_tokens": 48076097.0,
      "step": 20970
    },
    {
      "entropy": 5.275104522705078,
      "epoch": 2.014889529298751,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045958909960210385,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21535389572381974,
      "num_tokens": 48087429.0,
      "step": 20975
    },
    {
      "entropy": 5.076815891265869,
      "epoch": 2.015369836695485,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045956949426726075,
      "loss": 4.8157,
      "mean_token_accuracy": 0.2302336350083351,
      "num_tokens": 48098347.0,
      "step": 20980
    },
    {
      "entropy": 5.073042535781861,
      "epoch": 2.015850144092219,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045954988464726203,
      "loss": 4.7324,
      "mean_token_accuracy": 0.23198778331279754,
      "num_tokens": 48109355.0,
      "step": 20985
    },
    {
      "entropy": 5.201717948913574,
      "epoch": 2.016330451488953,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000459530270742563,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21356847286224365,
      "num_tokens": 48121875.0,
      "step": 20990
    },
    {
      "entropy": 5.263304948806763,
      "epoch": 2.0168107588856867,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045951065255361905,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21643733680248262,
      "num_tokens": 48133192.0,
      "step": 20995
    },
    {
      "entropy": 5.151610898971557,
      "epoch": 2.0172910662824206,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004594910300808856,
      "loss": 4.8249,
      "mean_token_accuracy": 0.22777229994535447,
      "num_tokens": 48143621.0,
      "step": 21000
    },
    {
      "epoch": 2.0172910662824206,
      "eval_entropy": 4.941178742529182,
      "eval_loss": 5.033307075500488,
      "eval_mean_token_accuracy": 0.22388243330401916,
      "eval_num_tokens": 48143621.0,
      "eval_runtime": 26.6805,
      "eval_samples_per_second": 1229.926,
      "eval_steps_per_second": 153.745,
      "step": 21000
    },
    {
      "entropy": 5.093407440185547,
      "epoch": 2.0177713736791545,
      "grad_norm": 1.125,
      "learning_rate": 0.0004594714033248183,
      "loss": 4.7582,
      "mean_token_accuracy": 0.23399291634559632,
      "num_tokens": 48153803.0,
      "step": 21005
    },
    {
      "entropy": 5.191311264038086,
      "epoch": 2.0182516810758884,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004594517722858728,
      "loss": 4.8619,
      "mean_token_accuracy": 0.21755242496728897,
      "num_tokens": 48165742.0,
      "step": 21010
    },
    {
      "entropy": 5.150562000274658,
      "epoch": 2.0187319884726227,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045943213696450475,
      "loss": 4.8556,
      "mean_token_accuracy": 0.2285227119922638,
      "num_tokens": 48177894.0,
      "step": 21015
    },
    {
      "entropy": 5.187864685058594,
      "epoch": 2.0192122958693566,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045941249736117023,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22163758128881456,
      "num_tokens": 48188536.0,
      "step": 21020
    },
    {
      "entropy": 5.196171712875366,
      "epoch": 2.0196926032660905,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004593928534763251,
      "loss": 4.8303,
      "mean_token_accuracy": 0.22475417852401733,
      "num_tokens": 48199276.0,
      "step": 21025
    },
    {
      "entropy": 5.273113870620728,
      "epoch": 2.0201729106628243,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004593732053104254,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21574490815401076,
      "num_tokens": 48212479.0,
      "step": 21030
    },
    {
      "entropy": 5.143319940567016,
      "epoch": 2.020653218059558,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045935355286392735,
      "loss": 4.7939,
      "mean_token_accuracy": 0.22910930812358857,
      "num_tokens": 48224279.0,
      "step": 21035
    },
    {
      "entropy": 5.197102355957031,
      "epoch": 2.021133525456292,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004593338961372873,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21459668278694152,
      "num_tokens": 48235943.0,
      "step": 21040
    },
    {
      "entropy": 5.126576519012451,
      "epoch": 2.021613832853026,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004593142351309614,
      "loss": 4.7467,
      "mean_token_accuracy": 0.2361180603504181,
      "num_tokens": 48246407.0,
      "step": 21045
    },
    {
      "entropy": 5.263282251358032,
      "epoch": 2.02209414024976,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004592945698454064,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21927241384983062,
      "num_tokens": 48256957.0,
      "step": 21050
    },
    {
      "entropy": 5.121550512313843,
      "epoch": 2.0225744476464937,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045927490028107866,
      "loss": 4.8111,
      "mean_token_accuracy": 0.22994543612003326,
      "num_tokens": 48268253.0,
      "step": 21055
    },
    {
      "entropy": 5.263857412338257,
      "epoch": 2.0230547550432276,
      "grad_norm": 1.15625,
      "learning_rate": 0.000459255226438435,
      "loss": 5.0115,
      "mean_token_accuracy": 0.2233571782708168,
      "num_tokens": 48279980.0,
      "step": 21060
    },
    {
      "entropy": 5.290223455429077,
      "epoch": 2.0235350624399615,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000459235548317932,
      "loss": 4.9654,
      "mean_token_accuracy": 0.21738289594650267,
      "num_tokens": 48291199.0,
      "step": 21065
    },
    {
      "entropy": 5.256168794631958,
      "epoch": 2.0240153698366954,
      "grad_norm": 1.25,
      "learning_rate": 0.00045921586592002667,
      "loss": 4.9299,
      "mean_token_accuracy": 0.21773264855146407,
      "num_tokens": 48301610.0,
      "step": 21070
    },
    {
      "entropy": 5.202200078964234,
      "epoch": 2.0244956772334293,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000459196179245176,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21690093874931335,
      "num_tokens": 48313655.0,
      "step": 21075
    },
    {
      "entropy": 5.205820751190186,
      "epoch": 2.024975984630163,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004591764882938369,
      "loss": 4.8479,
      "mean_token_accuracy": 0.2270449846982956,
      "num_tokens": 48325385.0,
      "step": 21080
    },
    {
      "entropy": 5.183474493026734,
      "epoch": 2.025456292026897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004591567930664667,
      "loss": 4.8034,
      "mean_token_accuracy": 0.22712614387273788,
      "num_tokens": 48335986.0,
      "step": 21085
    },
    {
      "entropy": 5.18653302192688,
      "epoch": 2.025936599423631,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004591370935635226,
      "loss": 4.9417,
      "mean_token_accuracy": 0.22101778239011766,
      "num_tokens": 48347196.0,
      "step": 21090
    },
    {
      "entropy": 5.284462642669678,
      "epoch": 2.0264169068203652,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004591173897854619,
      "loss": 4.9804,
      "mean_token_accuracy": 0.21575525850057603,
      "num_tokens": 48357944.0,
      "step": 21095
    },
    {
      "entropy": 5.289441680908203,
      "epoch": 2.026897214217099,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004590976817327422,
      "loss": 5.0734,
      "mean_token_accuracy": 0.20921075344085693,
      "num_tokens": 48369838.0,
      "step": 21100
    },
    {
      "entropy": 5.2147520065307615,
      "epoch": 2.027377521613833,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004590779694058209,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22528356015682222,
      "num_tokens": 48381107.0,
      "step": 21105
    },
    {
      "entropy": 5.147636032104492,
      "epoch": 2.027857829010567,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045905825280515586,
      "loss": 4.9186,
      "mean_token_accuracy": 0.22786805927753448,
      "num_tokens": 48392699.0,
      "step": 21110
    },
    {
      "entropy": 5.2457359790802,
      "epoch": 2.0283381364073008,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045903853193120464,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21879518628120423,
      "num_tokens": 48403836.0,
      "step": 21115
    },
    {
      "entropy": 5.19785737991333,
      "epoch": 2.0288184438040346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045901880678442524,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2257927194237709,
      "num_tokens": 48417373.0,
      "step": 21120
    },
    {
      "entropy": 5.234004831314087,
      "epoch": 2.0292987512007685,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045899907736527556,
      "loss": 4.8902,
      "mean_token_accuracy": 0.2240306630730629,
      "num_tokens": 48428033.0,
      "step": 21125
    },
    {
      "entropy": 5.155744504928589,
      "epoch": 2.0297790585975024,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045897934367421364,
      "loss": 4.8409,
      "mean_token_accuracy": 0.22534074187278746,
      "num_tokens": 48440147.0,
      "step": 21130
    },
    {
      "entropy": 5.084714651107788,
      "epoch": 2.0302593659942363,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004589596057116977,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22626282423734664,
      "num_tokens": 48451406.0,
      "step": 21135
    },
    {
      "entropy": 5.251463842391968,
      "epoch": 2.03073967339097,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045893986347818593,
      "loss": 4.8753,
      "mean_token_accuracy": 0.22858646810054778,
      "num_tokens": 48464393.0,
      "step": 21140
    },
    {
      "entropy": 5.259707593917847,
      "epoch": 2.031219980787704,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004589201169741368,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21791695356369017,
      "num_tokens": 48476433.0,
      "step": 21145
    },
    {
      "entropy": 5.190305376052857,
      "epoch": 2.031700288184438,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045890036620000856,
      "loss": 4.871,
      "mean_token_accuracy": 0.21782172620296478,
      "num_tokens": 48488372.0,
      "step": 21150
    },
    {
      "entropy": 5.142861080169678,
      "epoch": 2.032180595581172,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045888061115626,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2249121308326721,
      "num_tokens": 48498834.0,
      "step": 21155
    },
    {
      "entropy": 5.242824363708496,
      "epoch": 2.0326609029779057,
      "grad_norm": 1.125,
      "learning_rate": 0.0004588608518433496,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22475893348455428,
      "num_tokens": 48509160.0,
      "step": 21160
    },
    {
      "entropy": 5.184771156311035,
      "epoch": 2.0331412103746396,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004588410882617362,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22207292467355727,
      "num_tokens": 48522005.0,
      "step": 21165
    },
    {
      "entropy": 5.146971416473389,
      "epoch": 2.0336215177713735,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004588213204118786,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22238440811634064,
      "num_tokens": 48534238.0,
      "step": 21170
    },
    {
      "entropy": 5.210536193847656,
      "epoch": 2.034101825168108,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045880154829423586,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22164306789636612,
      "num_tokens": 48546440.0,
      "step": 21175
    },
    {
      "entropy": 5.298585271835327,
      "epoch": 2.0345821325648417,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004587817719092668,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22575015872716903,
      "num_tokens": 48556893.0,
      "step": 21180
    },
    {
      "entropy": 5.166137027740478,
      "epoch": 2.0350624399615755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045876199125743087,
      "loss": 4.8674,
      "mean_token_accuracy": 0.22293417751789094,
      "num_tokens": 48568483.0,
      "step": 21185
    },
    {
      "entropy": 5.221321678161621,
      "epoch": 2.0355427473583094,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004587422063391871,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21707276850938798,
      "num_tokens": 48579352.0,
      "step": 21190
    },
    {
      "entropy": 5.2925567626953125,
      "epoch": 2.0360230547550433,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004587224171549949,
      "loss": 4.958,
      "mean_token_accuracy": 0.20912449061870575,
      "num_tokens": 48591447.0,
      "step": 21195
    },
    {
      "entropy": 5.239304733276367,
      "epoch": 2.036503362151777,
      "grad_norm": 1.125,
      "learning_rate": 0.00045870262370531376,
      "loss": 4.9512,
      "mean_token_accuracy": 0.2115371584892273,
      "num_tokens": 48603724.0,
      "step": 21200
    },
    {
      "entropy": 5.262460088729858,
      "epoch": 2.036983669548511,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045868282599060314,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21064842641353607,
      "num_tokens": 48616319.0,
      "step": 21205
    },
    {
      "entropy": 5.208098697662353,
      "epoch": 2.037463976945245,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004586630240113227,
      "loss": 4.9082,
      "mean_token_accuracy": 0.22649887800216675,
      "num_tokens": 48627121.0,
      "step": 21210
    },
    {
      "entropy": 5.123235893249512,
      "epoch": 2.037944284341979,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004586432177679322,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23300582319498062,
      "num_tokens": 48638687.0,
      "step": 21215
    },
    {
      "entropy": 5.193862819671631,
      "epoch": 2.0384245917387127,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045862340726089153,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2156683847308159,
      "num_tokens": 48649480.0,
      "step": 21220
    },
    {
      "entropy": 5.334590101242066,
      "epoch": 2.0389048991354466,
      "grad_norm": 1.125,
      "learning_rate": 0.0004586035924906606,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21198081225156784,
      "num_tokens": 48660197.0,
      "step": 21225
    },
    {
      "entropy": 5.164762496948242,
      "epoch": 2.0393852065321805,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045858377345769946,
      "loss": 4.8192,
      "mean_token_accuracy": 0.22521817833185195,
      "num_tokens": 48671711.0,
      "step": 21230
    },
    {
      "entropy": 5.158647251129151,
      "epoch": 2.0398655139289144,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004585639501624682,
      "loss": 4.8149,
      "mean_token_accuracy": 0.22675239741802217,
      "num_tokens": 48683285.0,
      "step": 21235
    },
    {
      "entropy": 5.235030651092529,
      "epoch": 2.0403458213256482,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000458544122605427,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22100878953933717,
      "num_tokens": 48696308.0,
      "step": 21240
    },
    {
      "entropy": 5.146478700637817,
      "epoch": 2.040826128722382,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045852429078703646,
      "loss": 4.7327,
      "mean_token_accuracy": 0.23324364572763442,
      "num_tokens": 48706999.0,
      "step": 21245
    },
    {
      "entropy": 5.145619964599609,
      "epoch": 2.0413064361191164,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045850445470775673,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22685497850179673,
      "num_tokens": 48717248.0,
      "step": 21250
    },
    {
      "entropy": 5.212717580795288,
      "epoch": 2.0417867435158503,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004584846143680485,
      "loss": 4.9719,
      "mean_token_accuracy": 0.2198764756321907,
      "num_tokens": 48727948.0,
      "step": 21255
    },
    {
      "entropy": 5.254075813293457,
      "epoch": 2.042267050912584,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004584647697683723,
      "loss": 4.909,
      "mean_token_accuracy": 0.22531513571739198,
      "num_tokens": 48740474.0,
      "step": 21260
    },
    {
      "entropy": 5.248530626296997,
      "epoch": 2.042747358309318,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045844492090918904,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21490535140037537,
      "num_tokens": 48753415.0,
      "step": 21265
    },
    {
      "entropy": 5.1680028438568115,
      "epoch": 2.043227665706052,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045842506779095936,
      "loss": 4.8358,
      "mean_token_accuracy": 0.22975049316883087,
      "num_tokens": 48764466.0,
      "step": 21270
    },
    {
      "entropy": 5.123490571975708,
      "epoch": 2.043707973102786,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004584052104141443,
      "loss": 4.8535,
      "mean_token_accuracy": 0.22468803226947784,
      "num_tokens": 48777513.0,
      "step": 21275
    },
    {
      "entropy": 5.249536848068237,
      "epoch": 2.0441882804995197,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004583853487792048,
      "loss": 4.964,
      "mean_token_accuracy": 0.22335170954465866,
      "num_tokens": 48788732.0,
      "step": 21280
    },
    {
      "entropy": 5.195466470718384,
      "epoch": 2.0446685878962536,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004583654828866021,
      "loss": 4.8234,
      "mean_token_accuracy": 0.2228946715593338,
      "num_tokens": 48799620.0,
      "step": 21285
    },
    {
      "entropy": 5.121470212936401,
      "epoch": 2.0451488952929875,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004583456127367973,
      "loss": 4.8794,
      "mean_token_accuracy": 0.2222111463546753,
      "num_tokens": 48811914.0,
      "step": 21290
    },
    {
      "entropy": 5.165575933456421,
      "epoch": 2.0456292026897214,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004583257383302519,
      "loss": 4.8537,
      "mean_token_accuracy": 0.23043118715286254,
      "num_tokens": 48822610.0,
      "step": 21295
    },
    {
      "entropy": 5.2476400375366214,
      "epoch": 2.0461095100864553,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004583058596674271,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21359362602233886,
      "num_tokens": 48834296.0,
      "step": 21300
    },
    {
      "entropy": 5.136348247528076,
      "epoch": 2.046589817483189,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004582859767487846,
      "loss": 4.7837,
      "mean_token_accuracy": 0.23469096720218657,
      "num_tokens": 48846455.0,
      "step": 21305
    },
    {
      "entropy": 5.246692276000976,
      "epoch": 2.047070124879923,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045826608957478604,
      "loss": 4.9954,
      "mean_token_accuracy": 0.21699339747428895,
      "num_tokens": 48858364.0,
      "step": 21310
    },
    {
      "entropy": 5.223354196548462,
      "epoch": 2.047550432276657,
      "grad_norm": 1.125,
      "learning_rate": 0.00045824619814589297,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21964579373598098,
      "num_tokens": 48869829.0,
      "step": 21315
    },
    {
      "entropy": 5.186509704589843,
      "epoch": 2.0480307396733908,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004582263024625674,
      "loss": 4.8878,
      "mean_token_accuracy": 0.2251600056886673,
      "num_tokens": 48880951.0,
      "step": 21320
    },
    {
      "entropy": 5.210753297805786,
      "epoch": 2.048511047070125,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004582064025252711,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22490002065896988,
      "num_tokens": 48893367.0,
      "step": 21325
    },
    {
      "entropy": 5.2239336490631105,
      "epoch": 2.048991354466859,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004581864983344661,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2211512565612793,
      "num_tokens": 48904400.0,
      "step": 21330
    },
    {
      "entropy": 5.218151807785034,
      "epoch": 2.049471661863593,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004581665898906147,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2193092957139015,
      "num_tokens": 48916305.0,
      "step": 21335
    },
    {
      "entropy": 5.198601961135864,
      "epoch": 2.0499519692603267,
      "grad_norm": 1.125,
      "learning_rate": 0.00045814667719417887,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21167248040437697,
      "num_tokens": 48927160.0,
      "step": 21340
    },
    {
      "entropy": 5.203508710861206,
      "epoch": 2.0504322766570606,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000458126760245621,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21260669827461243,
      "num_tokens": 48940776.0,
      "step": 21345
    },
    {
      "entropy": 5.2547690868377686,
      "epoch": 2.0509125840537945,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004581068390454036,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22185130119323732,
      "num_tokens": 48952375.0,
      "step": 21350
    },
    {
      "entropy": 5.200519037246704,
      "epoch": 2.0513928914505284,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045808691359398905,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2220393106341362,
      "num_tokens": 48964684.0,
      "step": 21355
    },
    {
      "entropy": 5.251200389862061,
      "epoch": 2.0518731988472623,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004580669838918401,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22279408276081086,
      "num_tokens": 48975654.0,
      "step": 21360
    },
    {
      "entropy": 5.200453805923462,
      "epoch": 2.052353506243996,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045804704993941935,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2207440346479416,
      "num_tokens": 48987983.0,
      "step": 21365
    },
    {
      "entropy": 5.099079704284668,
      "epoch": 2.05283381364073,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045802711173718966,
      "loss": 4.7995,
      "mean_token_accuracy": 0.23463097512722014,
      "num_tokens": 48998728.0,
      "step": 21370
    },
    {
      "entropy": 5.217243957519531,
      "epoch": 2.053314121037464,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004580071692856138,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2186826303601265,
      "num_tokens": 49009177.0,
      "step": 21375
    },
    {
      "entropy": 5.171085262298584,
      "epoch": 2.053794428434198,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045798722258515504,
      "loss": 4.8537,
      "mean_token_accuracy": 0.23012643456459045,
      "num_tokens": 49021550.0,
      "step": 21380
    },
    {
      "entropy": 5.225291299819946,
      "epoch": 2.0542747358309317,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045796727163627623,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22325250655412673,
      "num_tokens": 49032349.0,
      "step": 21385
    },
    {
      "entropy": 5.137249612808228,
      "epoch": 2.0547550432276656,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004579473164394408,
      "loss": 4.8261,
      "mean_token_accuracy": 0.23287673890590668,
      "num_tokens": 49044829.0,
      "step": 21390
    },
    {
      "entropy": 5.169116258621216,
      "epoch": 2.0552353506243994,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045792735699511176,
      "loss": 4.9547,
      "mean_token_accuracy": 0.221877521276474,
      "num_tokens": 49057319.0,
      "step": 21395
    },
    {
      "entropy": 5.23916277885437,
      "epoch": 2.0557156580211333,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045790739330375276,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2258367583155632,
      "num_tokens": 49068872.0,
      "step": 21400
    },
    {
      "entropy": 5.147930049896241,
      "epoch": 2.0561959654178676,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045788742536582717,
      "loss": 4.8206,
      "mean_token_accuracy": 0.22233800888061522,
      "num_tokens": 49081080.0,
      "step": 21405
    },
    {
      "entropy": 5.088253688812256,
      "epoch": 2.0566762728146015,
      "grad_norm": 1.46875,
      "learning_rate": 0.00045786745318179866,
      "loss": 4.768,
      "mean_token_accuracy": 0.23265078961849212,
      "num_tokens": 49092535.0,
      "step": 21410
    },
    {
      "entropy": 5.203385925292968,
      "epoch": 2.0571565802113354,
      "grad_norm": 1.125,
      "learning_rate": 0.0004578474767521309,
      "loss": 4.9674,
      "mean_token_accuracy": 0.21658169329166413,
      "num_tokens": 49103743.0,
      "step": 21415
    },
    {
      "entropy": 5.25092306137085,
      "epoch": 2.0576368876080693,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045782749607728765,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22512667775154113,
      "num_tokens": 49115346.0,
      "step": 21420
    },
    {
      "entropy": 5.162568235397339,
      "epoch": 2.058117195004803,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045780751115773286,
      "loss": 4.8668,
      "mean_token_accuracy": 0.23150654733181,
      "num_tokens": 49126000.0,
      "step": 21425
    },
    {
      "entropy": 5.126068830490112,
      "epoch": 2.058597502401537,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004577875219939304,
      "loss": 4.8167,
      "mean_token_accuracy": 0.2281006693840027,
      "num_tokens": 49136617.0,
      "step": 21430
    },
    {
      "entropy": 5.191619062423706,
      "epoch": 2.059077809798271,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004577675285863446,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22351066172122955,
      "num_tokens": 49147804.0,
      "step": 21435
    },
    {
      "entropy": 5.181457138061523,
      "epoch": 2.059558117195005,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045774753093543943,
      "loss": 4.8163,
      "mean_token_accuracy": 0.2315751552581787,
      "num_tokens": 49158367.0,
      "step": 21440
    },
    {
      "entropy": 5.208706712722778,
      "epoch": 2.0600384245917387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004577275290416791,
      "loss": 4.9182,
      "mean_token_accuracy": 0.22565925419330596,
      "num_tokens": 49170236.0,
      "step": 21445
    },
    {
      "entropy": 5.183334112167358,
      "epoch": 2.0605187319884726,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004577075229055283,
      "loss": 4.8984,
      "mean_token_accuracy": 0.224626886844635,
      "num_tokens": 49183379.0,
      "step": 21450
    },
    {
      "entropy": 5.188776922225952,
      "epoch": 2.0609990393852065,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045768751252745133,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21921564042568206,
      "num_tokens": 49194649.0,
      "step": 21455
    },
    {
      "entropy": 5.210404682159424,
      "epoch": 2.0614793467819403,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045766749790791274,
      "loss": 4.9125,
      "mean_token_accuracy": 0.22198745906352996,
      "num_tokens": 49206523.0,
      "step": 21460
    },
    {
      "entropy": 5.205079698562622,
      "epoch": 2.061959654178674,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004576474790473773,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21463411450386047,
      "num_tokens": 49219020.0,
      "step": 21465
    },
    {
      "entropy": 5.1811188697814945,
      "epoch": 2.062439961575408,
      "grad_norm": 1.125,
      "learning_rate": 0.00045762745594630973,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2247908443212509,
      "num_tokens": 49230751.0,
      "step": 21470
    },
    {
      "entropy": 5.277703475952149,
      "epoch": 2.062920268972142,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004576074286051749,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22306783348321915,
      "num_tokens": 49241758.0,
      "step": 21475
    },
    {
      "entropy": 5.13308629989624,
      "epoch": 2.063400576368876,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045758739702443787,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22889921367168425,
      "num_tokens": 49253404.0,
      "step": 21480
    },
    {
      "entropy": 5.242451763153076,
      "epoch": 2.06388088376561,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004575673612045636,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21508880853652954,
      "num_tokens": 49264000.0,
      "step": 21485
    },
    {
      "entropy": 5.229710435867309,
      "epoch": 2.064361191162344,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004575473211460173,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21860820353031157,
      "num_tokens": 49275657.0,
      "step": 21490
    },
    {
      "entropy": 5.228364181518555,
      "epoch": 2.064841498559078,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004575272768492644,
      "loss": 5.0324,
      "mean_token_accuracy": 0.21086435765028,
      "num_tokens": 49288570.0,
      "step": 21495
    },
    {
      "entropy": 5.225734376907349,
      "epoch": 2.065321805955812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045750722831476993,
      "loss": 4.9008,
      "mean_token_accuracy": 0.21867357939481735,
      "num_tokens": 49300982.0,
      "step": 21500
    },
    {
      "entropy": 5.22962589263916,
      "epoch": 2.0658021133525457,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045748717554299964,
      "loss": 4.9564,
      "mean_token_accuracy": 0.22012604027986526,
      "num_tokens": 49313141.0,
      "step": 21505
    },
    {
      "entropy": 5.2185704708099365,
      "epoch": 2.0662824207492796,
      "grad_norm": 1.0,
      "learning_rate": 0.000457467118534419,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22198057174682617,
      "num_tokens": 49324726.0,
      "step": 21510
    },
    {
      "entropy": 5.271744346618652,
      "epoch": 2.0667627281460135,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004574470572894938,
      "loss": 4.9709,
      "mean_token_accuracy": 0.21429276168346406,
      "num_tokens": 49336187.0,
      "step": 21515
    },
    {
      "entropy": 5.275170803070068,
      "epoch": 2.0672430355427474,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004574269918086895,
      "loss": 4.9311,
      "mean_token_accuracy": 0.223145492374897,
      "num_tokens": 49348075.0,
      "step": 21520
    },
    {
      "entropy": 5.185292768478393,
      "epoch": 2.0677233429394812,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004574069220924722,
      "loss": 4.8822,
      "mean_token_accuracy": 0.2236448898911476,
      "num_tokens": 49358339.0,
      "step": 21525
    },
    {
      "entropy": 5.215222311019898,
      "epoch": 2.068203650336215,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004573868481413079,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22415236085653306,
      "num_tokens": 49369650.0,
      "step": 21530
    },
    {
      "entropy": 5.169541645050049,
      "epoch": 2.068683957732949,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045736676995566244,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22616811692714692,
      "num_tokens": 49381982.0,
      "step": 21535
    },
    {
      "entropy": 5.2314229011535645,
      "epoch": 2.069164265129683,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045734668753600217,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21810407489538192,
      "num_tokens": 49393442.0,
      "step": 21540
    },
    {
      "entropy": 5.2392706871032715,
      "epoch": 2.0696445725264168,
      "grad_norm": 1.125,
      "learning_rate": 0.00045732660088279326,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22833432704210282,
      "num_tokens": 49404311.0,
      "step": 21545
    },
    {
      "entropy": 5.248991441726685,
      "epoch": 2.0701248799231506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045730650999650216,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21662491112947463,
      "num_tokens": 49417587.0,
      "step": 21550
    },
    {
      "entropy": 5.249261283874512,
      "epoch": 2.0706051873198845,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045728641487759506,
      "loss": 4.947,
      "mean_token_accuracy": 0.21923352032899857,
      "num_tokens": 49429119.0,
      "step": 21555
    },
    {
      "entropy": 5.169037389755249,
      "epoch": 2.071085494716619,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004572663155265388,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2205416291952133,
      "num_tokens": 49440734.0,
      "step": 21560
    },
    {
      "entropy": 5.196160507202149,
      "epoch": 2.0715658021133527,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004572462119437999,
      "loss": 4.8472,
      "mean_token_accuracy": 0.21969019621610641,
      "num_tokens": 49452830.0,
      "step": 21565
    },
    {
      "entropy": 5.208309125900269,
      "epoch": 2.0720461095100866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045722610412984513,
      "loss": 4.8721,
      "mean_token_accuracy": 0.220079205930233,
      "num_tokens": 49464380.0,
      "step": 21570
    },
    {
      "entropy": 5.0763853073120115,
      "epoch": 2.0725264169068205,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004572059920851412,
      "loss": 4.8483,
      "mean_token_accuracy": 0.2316389188170433,
      "num_tokens": 49475436.0,
      "step": 21575
    },
    {
      "entropy": 5.18530158996582,
      "epoch": 2.0730067243035544,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045718587581015534,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21858569979667664,
      "num_tokens": 49487015.0,
      "step": 21580
    },
    {
      "entropy": 5.2257678508758545,
      "epoch": 2.0734870317002883,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004571657553053544,
      "loss": 4.792,
      "mean_token_accuracy": 0.22679604440927506,
      "num_tokens": 49498485.0,
      "step": 21585
    },
    {
      "entropy": 5.261003541946411,
      "epoch": 2.073967339097022,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004571456305712055,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2220904305577278,
      "num_tokens": 49508899.0,
      "step": 21590
    },
    {
      "entropy": 5.1665746688842775,
      "epoch": 2.074447646493756,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004571255016081759,
      "loss": 4.8175,
      "mean_token_accuracy": 0.22671875059604646,
      "num_tokens": 49520956.0,
      "step": 21595
    },
    {
      "entropy": 5.262011432647705,
      "epoch": 2.07492795389049,
      "grad_norm": 1.34375,
      "learning_rate": 0.000457105368416733,
      "loss": 5.0095,
      "mean_token_accuracy": 0.20953049063682555,
      "num_tokens": 49532726.0,
      "step": 21600
    },
    {
      "entropy": 5.257593870162964,
      "epoch": 2.0754082612872238,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045708523099734417,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21485285758972167,
      "num_tokens": 49545156.0,
      "step": 21605
    },
    {
      "entropy": 5.184183120727539,
      "epoch": 2.0758885686839577,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045706508935047693,
      "loss": 4.9047,
      "mean_token_accuracy": 0.22326010167598725,
      "num_tokens": 49557552.0,
      "step": 21610
    },
    {
      "entropy": 5.188806200027466,
      "epoch": 2.0763688760806915,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000457044943476599,
      "loss": 4.916,
      "mean_token_accuracy": 0.21916239112615585,
      "num_tokens": 49567906.0,
      "step": 21615
    },
    {
      "entropy": 5.188045644760132,
      "epoch": 2.0768491834774254,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045702479337617795,
      "loss": 4.8232,
      "mean_token_accuracy": 0.22927991896867753,
      "num_tokens": 49578641.0,
      "step": 21620
    },
    {
      "entropy": 5.2214438915252686,
      "epoch": 2.0773294908741593,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004570046390496818,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22721950262784957,
      "num_tokens": 49590449.0,
      "step": 21625
    },
    {
      "entropy": 5.207981967926026,
      "epoch": 2.077809798270893,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004569844804975783,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2221996769309044,
      "num_tokens": 49600877.0,
      "step": 21630
    },
    {
      "entropy": 5.20042200088501,
      "epoch": 2.0782901056676275,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004569643177203356,
      "loss": 4.9194,
      "mean_token_accuracy": 0.21821517497301102,
      "num_tokens": 49612310.0,
      "step": 21635
    },
    {
      "entropy": 5.196003055572509,
      "epoch": 2.0787704130643614,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004569441507184218,
      "loss": 4.8892,
      "mean_token_accuracy": 0.21416153907775878,
      "num_tokens": 49625347.0,
      "step": 21640
    },
    {
      "entropy": 5.25957932472229,
      "epoch": 2.0792507204610953,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045692397949230495,
      "loss": 4.9864,
      "mean_token_accuracy": 0.21552609354257585,
      "num_tokens": 49637943.0,
      "step": 21645
    },
    {
      "entropy": 5.184257221221924,
      "epoch": 2.079731027857829,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045690380404245364,
      "loss": 4.7988,
      "mean_token_accuracy": 0.23304454237222672,
      "num_tokens": 49649318.0,
      "step": 21650
    },
    {
      "entropy": 5.156260299682617,
      "epoch": 2.080211335254563,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045688362436933607,
      "loss": 4.8796,
      "mean_token_accuracy": 0.2269807815551758,
      "num_tokens": 49661368.0,
      "step": 21655
    },
    {
      "entropy": 5.1872039318084715,
      "epoch": 2.080691642651297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004568634404734208,
      "loss": 4.83,
      "mean_token_accuracy": 0.22870273888111115,
      "num_tokens": 49673566.0,
      "step": 21660
    },
    {
      "entropy": 5.1330241680145265,
      "epoch": 2.081171950048031,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004568432523551765,
      "loss": 4.8568,
      "mean_token_accuracy": 0.22635709792375563,
      "num_tokens": 49685151.0,
      "step": 21665
    },
    {
      "entropy": 5.256837320327759,
      "epoch": 2.0816522574447647,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004568230600150718,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22683397978544234,
      "num_tokens": 49696409.0,
      "step": 21670
    },
    {
      "entropy": 5.1499443531036375,
      "epoch": 2.0821325648414986,
      "grad_norm": 1.125,
      "learning_rate": 0.0004568028634535757,
      "loss": 4.8704,
      "mean_token_accuracy": 0.22599780410528184,
      "num_tokens": 49709031.0,
      "step": 21675
    },
    {
      "entropy": 5.057080602645874,
      "epoch": 2.0826128722382324,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004567826626711568,
      "loss": 4.7978,
      "mean_token_accuracy": 0.23206383883953094,
      "num_tokens": 49719646.0,
      "step": 21680
    },
    {
      "entropy": 5.109302568435669,
      "epoch": 2.0830931796349663,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004567624576682843,
      "loss": 4.8291,
      "mean_token_accuracy": 0.23019110709428786,
      "num_tokens": 49730384.0,
      "step": 21685
    },
    {
      "entropy": 5.233508586883545,
      "epoch": 2.0835734870317,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004567422484454273,
      "loss": 4.959,
      "mean_token_accuracy": 0.22004858106374742,
      "num_tokens": 49742256.0,
      "step": 21690
    },
    {
      "entropy": 5.199009275436401,
      "epoch": 2.084053794428434,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045672203500305493,
      "loss": 4.8313,
      "mean_token_accuracy": 0.22851186692714692,
      "num_tokens": 49753547.0,
      "step": 21695
    },
    {
      "entropy": 5.165274333953858,
      "epoch": 2.084534101825168,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045670181734163654,
      "loss": 4.9864,
      "mean_token_accuracy": 0.21861637085676194,
      "num_tokens": 49765472.0,
      "step": 21700
    },
    {
      "entropy": 5.1561150550842285,
      "epoch": 2.085014409221902,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004566815954616414,
      "loss": 4.7827,
      "mean_token_accuracy": 0.2253882646560669,
      "num_tokens": 49777031.0,
      "step": 21705
    },
    {
      "entropy": 5.239693737030029,
      "epoch": 2.0854947166186357,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045666136936353913,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22766790688037872,
      "num_tokens": 49787923.0,
      "step": 21710
    },
    {
      "entropy": 5.151098155975342,
      "epoch": 2.08597502401537,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004566411390477993,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22014029175043107,
      "num_tokens": 49799652.0,
      "step": 21715
    },
    {
      "entropy": 5.106624603271484,
      "epoch": 2.086455331412104,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045662090451489156,
      "loss": 4.8791,
      "mean_token_accuracy": 0.2241659790277481,
      "num_tokens": 49811278.0,
      "step": 21720
    },
    {
      "entropy": 5.259225177764892,
      "epoch": 2.086935638808838,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045660066576528577,
      "loss": 4.8991,
      "mean_token_accuracy": 0.22872586399316788,
      "num_tokens": 49823560.0,
      "step": 21725
    },
    {
      "entropy": 5.168730211257935,
      "epoch": 2.0874159462055717,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004565804227994518,
      "loss": 4.7895,
      "mean_token_accuracy": 0.22872886806726456,
      "num_tokens": 49834662.0,
      "step": 21730
    },
    {
      "entropy": 5.138026714324951,
      "epoch": 2.0878962536023056,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004565601756178595,
      "loss": 4.8708,
      "mean_token_accuracy": 0.22781936973333358,
      "num_tokens": 49845832.0,
      "step": 21735
    },
    {
      "entropy": 5.210244131088257,
      "epoch": 2.0883765609990395,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004565399242209791,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2231581002473831,
      "num_tokens": 49856420.0,
      "step": 21740
    },
    {
      "entropy": 5.116448974609375,
      "epoch": 2.0888568683957733,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004565196686092807,
      "loss": 4.7757,
      "mean_token_accuracy": 0.23266566246747972,
      "num_tokens": 49867680.0,
      "step": 21745
    },
    {
      "entropy": 5.124364042282105,
      "epoch": 2.089337175792507,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004564994087832346,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2177980661392212,
      "num_tokens": 49879246.0,
      "step": 21750
    },
    {
      "entropy": 5.152933740615845,
      "epoch": 2.089817483189241,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045647914474331123,
      "loss": 4.8273,
      "mean_token_accuracy": 0.23337583839893342,
      "num_tokens": 49890536.0,
      "step": 21755
    },
    {
      "entropy": 5.270107412338257,
      "epoch": 2.090297790585975,
      "grad_norm": 1.25,
      "learning_rate": 0.00045645887648998094,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21734450459480287,
      "num_tokens": 49902331.0,
      "step": 21760
    },
    {
      "entropy": 5.19177212715149,
      "epoch": 2.090778097982709,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045643860402371433,
      "loss": 4.8211,
      "mean_token_accuracy": 0.23614693880081178,
      "num_tokens": 49913814.0,
      "step": 21765
    },
    {
      "entropy": 5.121470212936401,
      "epoch": 2.0912584053794427,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004564183273449821,
      "loss": 4.8037,
      "mean_token_accuracy": 0.2256082057952881,
      "num_tokens": 49924592.0,
      "step": 21770
    },
    {
      "entropy": 5.154105997085571,
      "epoch": 2.0917387127761766,
      "grad_norm": 1.125,
      "learning_rate": 0.0004563980464542551,
      "loss": 4.9102,
      "mean_token_accuracy": 0.21861217468976973,
      "num_tokens": 49936526.0,
      "step": 21775
    },
    {
      "entropy": 5.226053762435913,
      "epoch": 2.0922190201729105,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045637776135200406,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22775020599365234,
      "num_tokens": 49947655.0,
      "step": 21780
    },
    {
      "entropy": 5.190950727462768,
      "epoch": 2.0926993275696444,
      "grad_norm": 1.125,
      "learning_rate": 0.0004563574720386999,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22483016699552535,
      "num_tokens": 49959251.0,
      "step": 21785
    },
    {
      "entropy": 5.213126802444458,
      "epoch": 2.0931796349663783,
      "grad_norm": 1.125,
      "learning_rate": 0.0004563371785148139,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22027482688426972,
      "num_tokens": 49970411.0,
      "step": 21790
    },
    {
      "entropy": 5.216670417785645,
      "epoch": 2.0936599423631126,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045631688078081695,
      "loss": 5.0372,
      "mean_token_accuracy": 0.21725357472896575,
      "num_tokens": 49983935.0,
      "step": 21795
    },
    {
      "entropy": 5.191430950164795,
      "epoch": 2.0941402497598465,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004562965788371805,
      "loss": 4.8462,
      "mean_token_accuracy": 0.2293910101056099,
      "num_tokens": 49995278.0,
      "step": 21800
    },
    {
      "entropy": 5.1762079238891605,
      "epoch": 2.0946205571565804,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004562762726843758,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22537946999073027,
      "num_tokens": 50005534.0,
      "step": 21805
    },
    {
      "entropy": 5.1693638324737545,
      "epoch": 2.0951008645533142,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045625596232287436,
      "loss": 4.8967,
      "mean_token_accuracy": 0.22431332916021346,
      "num_tokens": 50016583.0,
      "step": 21810
    },
    {
      "entropy": 5.12965145111084,
      "epoch": 2.095581171950048,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004562356477531477,
      "loss": 4.8282,
      "mean_token_accuracy": 0.22288532108068465,
      "num_tokens": 50027290.0,
      "step": 21815
    },
    {
      "entropy": 5.275524377822876,
      "epoch": 2.096061479346782,
      "grad_norm": 1.25,
      "learning_rate": 0.0004562153289756674,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22302284836769104,
      "num_tokens": 50039671.0,
      "step": 21820
    },
    {
      "entropy": 5.2127727508544925,
      "epoch": 2.096541786743516,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004561950059909053,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22877921313047409,
      "num_tokens": 50051639.0,
      "step": 21825
    },
    {
      "entropy": 5.155646514892578,
      "epoch": 2.0970220941402498,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004561746787993332,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21504004001617433,
      "num_tokens": 50064475.0,
      "step": 21830
    },
    {
      "entropy": 5.206075620651245,
      "epoch": 2.0975024015369836,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045615434740142307,
      "loss": 4.8491,
      "mean_token_accuracy": 0.22428507059812547,
      "num_tokens": 50076038.0,
      "step": 21835
    },
    {
      "entropy": 5.2276746273040775,
      "epoch": 2.0979827089337175,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045613401179764686,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22757401466369628,
      "num_tokens": 50086425.0,
      "step": 21840
    },
    {
      "entropy": 5.2021668434143065,
      "epoch": 2.0984630163304514,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045611367198847676,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22466616332530975,
      "num_tokens": 50098320.0,
      "step": 21845
    },
    {
      "entropy": 5.118850135803223,
      "epoch": 2.0989433237271853,
      "grad_norm": 1.09375,
      "learning_rate": 0.000456093327974385,
      "loss": 4.8291,
      "mean_token_accuracy": 0.225565804541111,
      "num_tokens": 50109619.0,
      "step": 21850
    },
    {
      "entropy": 5.1862670421600345,
      "epoch": 2.099423631123919,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004560729797558438,
      "loss": 4.838,
      "mean_token_accuracy": 0.23262035250663757,
      "num_tokens": 50121043.0,
      "step": 21855
    },
    {
      "entropy": 5.23045768737793,
      "epoch": 2.099903938520653,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004560526273333259,
      "loss": 4.9526,
      "mean_token_accuracy": 0.21904795467853547,
      "num_tokens": 50132673.0,
      "step": 21860
    },
    {
      "entropy": 5.103979682922363,
      "epoch": 2.100384245917387,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045603227070730346,
      "loss": 4.7993,
      "mean_token_accuracy": 0.22576977163553238,
      "num_tokens": 50144393.0,
      "step": 21865
    },
    {
      "entropy": 5.162962150573731,
      "epoch": 2.1008645533141213,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045601190987824933,
      "loss": 4.8623,
      "mean_token_accuracy": 0.21882206499576567,
      "num_tokens": 50156278.0,
      "step": 21870
    },
    {
      "entropy": 5.171366596221924,
      "epoch": 2.101344860710855,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045599154484663606,
      "loss": 4.8841,
      "mean_token_accuracy": 0.22640691250562667,
      "num_tokens": 50167652.0,
      "step": 21875
    },
    {
      "entropy": 5.178159856796265,
      "epoch": 2.101825168107589,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045597117561293663,
      "loss": 4.9466,
      "mean_token_accuracy": 0.22815511226654053,
      "num_tokens": 50179482.0,
      "step": 21880
    },
    {
      "entropy": 5.1986102104187015,
      "epoch": 2.102305475504323,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004559508021776238,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22812045961618424,
      "num_tokens": 50190137.0,
      "step": 21885
    },
    {
      "entropy": 5.377681541442871,
      "epoch": 2.1027857829010568,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004559304245411707,
      "loss": 5.053,
      "mean_token_accuracy": 0.21378545165061952,
      "num_tokens": 50201853.0,
      "step": 21890
    },
    {
      "entropy": 5.196824789047241,
      "epoch": 2.1032660902977907,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045591004270405044,
      "loss": 4.8534,
      "mean_token_accuracy": 0.227129702270031,
      "num_tokens": 50212635.0,
      "step": 21895
    },
    {
      "entropy": 5.143590879440308,
      "epoch": 2.1037463976945245,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004558896566667361,
      "loss": 4.8048,
      "mean_token_accuracy": 0.2245868682861328,
      "num_tokens": 50224842.0,
      "step": 21900
    },
    {
      "entropy": 5.2247498512268065,
      "epoch": 2.1042267050912584,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045586926642970113,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22170940935611724,
      "num_tokens": 50236741.0,
      "step": 21905
    },
    {
      "entropy": 5.230319261550903,
      "epoch": 2.1047070124879923,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004558488719934188,
      "loss": 4.8712,
      "mean_token_accuracy": 0.2145049601793289,
      "num_tokens": 50247572.0,
      "step": 21910
    },
    {
      "entropy": 5.258041667938232,
      "epoch": 2.105187319884726,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004558284733583627,
      "loss": 4.972,
      "mean_token_accuracy": 0.2251180276274681,
      "num_tokens": 50258566.0,
      "step": 21915
    },
    {
      "entropy": 5.176934766769409,
      "epoch": 2.10566762728146,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045580807052500645,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22604912221431733,
      "num_tokens": 50270577.0,
      "step": 21920
    },
    {
      "entropy": 5.1981360912323,
      "epoch": 2.106147934678194,
      "grad_norm": 1.25,
      "learning_rate": 0.0004557876634938236,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22409000247716904,
      "num_tokens": 50281124.0,
      "step": 21925
    },
    {
      "entropy": 5.183554697036743,
      "epoch": 2.106628242074928,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004557672522652881,
      "loss": 4.8058,
      "mean_token_accuracy": 0.23224017173051834,
      "num_tokens": 50291945.0,
      "step": 21930
    },
    {
      "entropy": 5.163864183425903,
      "epoch": 2.1071085494716617,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004557468368398738,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22895077615976334,
      "num_tokens": 50303312.0,
      "step": 21935
    },
    {
      "entropy": 5.197466230392456,
      "epoch": 2.1075888568683956,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004557264172180546,
      "loss": 4.8749,
      "mean_token_accuracy": 0.2196645498275757,
      "num_tokens": 50314791.0,
      "step": 21940
    },
    {
      "entropy": 5.280777549743652,
      "epoch": 2.1080691642651295,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004557059934003046,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2190120831131935,
      "num_tokens": 50326438.0,
      "step": 21945
    },
    {
      "entropy": 5.085853910446167,
      "epoch": 2.108549471661864,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004556855653870981,
      "loss": 4.7195,
      "mean_token_accuracy": 0.2359825611114502,
      "num_tokens": 50338091.0,
      "step": 21950
    },
    {
      "entropy": 5.115252542495727,
      "epoch": 2.1090297790585977,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004556651331789092,
      "loss": 4.8292,
      "mean_token_accuracy": 0.2290794774889946,
      "num_tokens": 50350387.0,
      "step": 21955
    },
    {
      "entropy": 5.187832880020141,
      "epoch": 2.1095100864553316,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004556446967762125,
      "loss": 4.8689,
      "mean_token_accuracy": 0.2235697716474533,
      "num_tokens": 50362689.0,
      "step": 21960
    },
    {
      "entropy": 5.203122234344482,
      "epoch": 2.1099903938520654,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045562425617948226,
      "loss": 4.8291,
      "mean_token_accuracy": 0.22703847885131836,
      "num_tokens": 50373290.0,
      "step": 21965
    },
    {
      "entropy": 5.155915927886963,
      "epoch": 2.1104707012487993,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045560381138919315,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2229043334722519,
      "num_tokens": 50385267.0,
      "step": 21970
    },
    {
      "entropy": 5.203227949142456,
      "epoch": 2.110951008645533,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045558336240581984,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21787139028310776,
      "num_tokens": 50398014.0,
      "step": 21975
    },
    {
      "entropy": 5.248868417739868,
      "epoch": 2.111431316042267,
      "grad_norm": 1.125,
      "learning_rate": 0.00045556290922983705,
      "loss": 4.8803,
      "mean_token_accuracy": 0.22950793355703353,
      "num_tokens": 50408661.0,
      "step": 21980
    },
    {
      "entropy": 5.171831512451172,
      "epoch": 2.111911623439001,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004555424518617197,
      "loss": 4.875,
      "mean_token_accuracy": 0.22743822932243346,
      "num_tokens": 50419231.0,
      "step": 21985
    },
    {
      "entropy": 5.1759045124053955,
      "epoch": 2.112391930835735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045552199030194274,
      "loss": 4.8734,
      "mean_token_accuracy": 0.2240717440843582,
      "num_tokens": 50430412.0,
      "step": 21990
    },
    {
      "entropy": 5.313192892074585,
      "epoch": 2.1128722382324687,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045550152455098113,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21500394344329835,
      "num_tokens": 50442431.0,
      "step": 21995
    },
    {
      "entropy": 5.2053369045257565,
      "epoch": 2.1133525456292026,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004554810546093102,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22319127768278121,
      "num_tokens": 50454417.0,
      "step": 22000
    },
    {
      "entropy": 5.152405071258545,
      "epoch": 2.1138328530259365,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000455460580477405,
      "loss": 4.8927,
      "mean_token_accuracy": 0.22537903785705565,
      "num_tokens": 50466447.0,
      "step": 22005
    },
    {
      "entropy": 5.142823696136475,
      "epoch": 2.1143131604226704,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000455440102155741,
      "loss": 4.8487,
      "mean_token_accuracy": 0.21924786418676376,
      "num_tokens": 50477843.0,
      "step": 22010
    },
    {
      "entropy": 5.22438907623291,
      "epoch": 2.1147934678194042,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004554196196447937,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2206392228603363,
      "num_tokens": 50489407.0,
      "step": 22015
    },
    {
      "entropy": 5.210554409027099,
      "epoch": 2.115273775216138,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004553991329450385,
      "loss": 4.8552,
      "mean_token_accuracy": 0.21922307461500168,
      "num_tokens": 50499492.0,
      "step": 22020
    },
    {
      "entropy": 5.110989904403686,
      "epoch": 2.115754082612872,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045537864205695116,
      "loss": 4.9152,
      "mean_token_accuracy": 0.22032220363616944,
      "num_tokens": 50511741.0,
      "step": 22025
    },
    {
      "entropy": 5.155436325073242,
      "epoch": 2.1162343900096063,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004553581469810073,
      "loss": 4.8205,
      "mean_token_accuracy": 0.2268645077943802,
      "num_tokens": 50523238.0,
      "step": 22030
    },
    {
      "entropy": 5.170282363891602,
      "epoch": 2.11671469740634,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045533764771768287,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22948864996433258,
      "num_tokens": 50534420.0,
      "step": 22035
    },
    {
      "entropy": 5.150433778762817,
      "epoch": 2.117195004803074,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045531714426745373,
      "loss": 4.8045,
      "mean_token_accuracy": 0.23159960210323333,
      "num_tokens": 50545785.0,
      "step": 22040
    },
    {
      "entropy": 5.153758192062378,
      "epoch": 2.117675312199808,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004552966366307959,
      "loss": 4.8771,
      "mean_token_accuracy": 0.21929933577775956,
      "num_tokens": 50556441.0,
      "step": 22045
    },
    {
      "entropy": 5.11735429763794,
      "epoch": 2.118155619596542,
      "grad_norm": 1.125,
      "learning_rate": 0.0004552761248081856,
      "loss": 4.8188,
      "mean_token_accuracy": 0.2272941514849663,
      "num_tokens": 50568519.0,
      "step": 22050
    },
    {
      "entropy": 5.2307037830352785,
      "epoch": 2.1186359269932757,
      "grad_norm": 1.046875,
      "learning_rate": 0.000455255608800099,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22845635265111924,
      "num_tokens": 50579469.0,
      "step": 22055
    },
    {
      "entropy": 5.011847925186157,
      "epoch": 2.1191162343900096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045523508860701237,
      "loss": 4.6836,
      "mean_token_accuracy": 0.23924311995506287,
      "num_tokens": 50591405.0,
      "step": 22060
    },
    {
      "entropy": 5.165399217605591,
      "epoch": 2.1195965417867435,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004552145642294021,
      "loss": 4.9033,
      "mean_token_accuracy": 0.23318920731544496,
      "num_tokens": 50603742.0,
      "step": 22065
    },
    {
      "entropy": 5.185004663467407,
      "epoch": 2.1200768491834774,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045519403566774493,
      "loss": 4.8802,
      "mean_token_accuracy": 0.2285549208521843,
      "num_tokens": 50615753.0,
      "step": 22070
    },
    {
      "entropy": 5.143928480148316,
      "epoch": 2.1205571565802113,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004551735029225172,
      "loss": 4.7701,
      "mean_token_accuracy": 0.23006146252155305,
      "num_tokens": 50626715.0,
      "step": 22075
    },
    {
      "entropy": 5.147879219055175,
      "epoch": 2.121037463976945,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045515296599419583,
      "loss": 4.8137,
      "mean_token_accuracy": 0.22976673841476442,
      "num_tokens": 50637552.0,
      "step": 22080
    },
    {
      "entropy": 5.186726999282837,
      "epoch": 2.121517771373679,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004551324248832574,
      "loss": 4.9443,
      "mean_token_accuracy": 0.21983531564474107,
      "num_tokens": 50648143.0,
      "step": 22085
    },
    {
      "entropy": 5.197938776016235,
      "epoch": 2.121998078770413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004551118795901791,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21496883332729338,
      "num_tokens": 50658878.0,
      "step": 22090
    },
    {
      "entropy": 5.1873246192932125,
      "epoch": 2.122478386167147,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004550913301154376,
      "loss": 4.8777,
      "mean_token_accuracy": 0.2279421254992485,
      "num_tokens": 50670912.0,
      "step": 22095
    },
    {
      "entropy": 5.206608295440674,
      "epoch": 2.1229586935638807,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004550707764595103,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21867920011281966,
      "num_tokens": 50683030.0,
      "step": 22100
    },
    {
      "entropy": 5.242287015914917,
      "epoch": 2.123439000960615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045505021862287434,
      "loss": 4.9283,
      "mean_token_accuracy": 0.21250441968441008,
      "num_tokens": 50694654.0,
      "step": 22105
    },
    {
      "entropy": 5.205548286437988,
      "epoch": 2.123919308357349,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045502965660600684,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2183799996972084,
      "num_tokens": 50706277.0,
      "step": 22110
    },
    {
      "entropy": 5.206652736663818,
      "epoch": 2.1243996157540828,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004550090904093853,
      "loss": 4.8479,
      "mean_token_accuracy": 0.228007735311985,
      "num_tokens": 50717610.0,
      "step": 22115
    },
    {
      "entropy": 5.242623424530029,
      "epoch": 2.1248799231508166,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004549885200334872,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21569137424230575,
      "num_tokens": 50728827.0,
      "step": 22120
    },
    {
      "entropy": 5.131516981124878,
      "epoch": 2.1253602305475505,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004549679454787901,
      "loss": 4.8248,
      "mean_token_accuracy": 0.23280414342880248,
      "num_tokens": 50740098.0,
      "step": 22125
    },
    {
      "entropy": 5.196243381500244,
      "epoch": 2.1258405379442844,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045494736674577175,
      "loss": 4.902,
      "mean_token_accuracy": 0.2179240569472313,
      "num_tokens": 50752166.0,
      "step": 22130
    },
    {
      "entropy": 5.181368017196656,
      "epoch": 2.1263208453410183,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004549267838349099,
      "loss": 4.8904,
      "mean_token_accuracy": 0.22681136131286622,
      "num_tokens": 50763561.0,
      "step": 22135
    },
    {
      "entropy": 5.215662336349487,
      "epoch": 2.126801152737752,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004549061967466823,
      "loss": 4.9416,
      "mean_token_accuracy": 0.22608317732810973,
      "num_tokens": 50774194.0,
      "step": 22140
    },
    {
      "entropy": 5.127267599105835,
      "epoch": 2.127281460134486,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004548856054815671,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22481022626161576,
      "num_tokens": 50784873.0,
      "step": 22145
    },
    {
      "entropy": 5.2286396503448485,
      "epoch": 2.12776176753122,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045486501004004225,
      "loss": 4.9739,
      "mean_token_accuracy": 0.21994537562131883,
      "num_tokens": 50795687.0,
      "step": 22150
    },
    {
      "entropy": 5.274788236618042,
      "epoch": 2.128242074927954,
      "grad_norm": 1.0625,
      "learning_rate": 0.000454844410422586,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2203521251678467,
      "num_tokens": 50806304.0,
      "step": 22155
    },
    {
      "entropy": 5.050665950775146,
      "epoch": 2.1287223823246877,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045482380662967655,
      "loss": 4.7806,
      "mean_token_accuracy": 0.2251899302005768,
      "num_tokens": 50816775.0,
      "step": 22160
    },
    {
      "entropy": 5.233350563049316,
      "epoch": 2.1292026897214216,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004548031986617923,
      "loss": 4.9511,
      "mean_token_accuracy": 0.213375324010849,
      "num_tokens": 50828266.0,
      "step": 22165
    },
    {
      "entropy": 5.208167695999146,
      "epoch": 2.1296829971181555,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004547825865194117,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2201148435473442,
      "num_tokens": 50840140.0,
      "step": 22170
    },
    {
      "entropy": 5.265772962570191,
      "epoch": 2.1301633045148893,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045476197020301323,
      "loss": 4.942,
      "mean_token_accuracy": 0.22228063493967057,
      "num_tokens": 50850415.0,
      "step": 22175
    },
    {
      "entropy": 5.16298131942749,
      "epoch": 2.1306436119116237,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045474134971307554,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22116966992616655,
      "num_tokens": 50861396.0,
      "step": 22180
    },
    {
      "entropy": 5.201243305206299,
      "epoch": 2.1311239193083575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004547207250500775,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2181514471769333,
      "num_tokens": 50873931.0,
      "step": 22185
    },
    {
      "entropy": 5.185925436019898,
      "epoch": 2.1316042267050914,
      "grad_norm": 1.796875,
      "learning_rate": 0.000454700096214498,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2209986761212349,
      "num_tokens": 50884557.0,
      "step": 22190
    },
    {
      "entropy": 5.125216054916382,
      "epoch": 2.1320845341018253,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045467946320681567,
      "loss": 4.7819,
      "mean_token_accuracy": 0.22966799437999724,
      "num_tokens": 50895757.0,
      "step": 22195
    },
    {
      "entropy": 5.182810735702515,
      "epoch": 2.132564841498559,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004546588260275098,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22748427838087082,
      "num_tokens": 50906437.0,
      "step": 22200
    },
    {
      "entropy": 5.139232778549195,
      "epoch": 2.133045148895293,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045463818467705955,
      "loss": 4.8457,
      "mean_token_accuracy": 0.2311835765838623,
      "num_tokens": 50916689.0,
      "step": 22205
    },
    {
      "entropy": 5.254155969619751,
      "epoch": 2.133525456292027,
      "grad_norm": 1.203125,
      "learning_rate": 0.000454617539155944,
      "loss": 4.9309,
      "mean_token_accuracy": 0.2202681615948677,
      "num_tokens": 50928266.0,
      "step": 22210
    },
    {
      "entropy": 5.266765975952149,
      "epoch": 2.134005763688761,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045459688946464255,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21727835536003112,
      "num_tokens": 50940220.0,
      "step": 22215
    },
    {
      "entropy": 5.124820709228516,
      "epoch": 2.1344860710854947,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004545762356036346,
      "loss": 4.8366,
      "mean_token_accuracy": 0.23560182899236679,
      "num_tokens": 50952668.0,
      "step": 22220
    },
    {
      "entropy": 5.149776601791382,
      "epoch": 2.1349663784822286,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004545555775733998,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22512850016355515,
      "num_tokens": 50964624.0,
      "step": 22225
    },
    {
      "entropy": 5.217778301239013,
      "epoch": 2.1354466858789625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045453491537441747,
      "loss": 4.9516,
      "mean_token_accuracy": 0.22014205455780028,
      "num_tokens": 50976704.0,
      "step": 22230
    },
    {
      "entropy": 5.1184944152832035,
      "epoch": 2.1359269932756964,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045451424900716763,
      "loss": 4.7913,
      "mean_token_accuracy": 0.22769148647785187,
      "num_tokens": 50987986.0,
      "step": 22235
    },
    {
      "entropy": 5.136394834518432,
      "epoch": 2.1364073006724302,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045449357847212994,
      "loss": 4.8332,
      "mean_token_accuracy": 0.22452268302440642,
      "num_tokens": 50999624.0,
      "step": 22240
    },
    {
      "entropy": 5.192212057113648,
      "epoch": 2.136887608069164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004544729037697844,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22724328190088272,
      "num_tokens": 51011800.0,
      "step": 22245
    },
    {
      "entropy": 5.1243151187896725,
      "epoch": 2.137367915465898,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045445222490061093,
      "loss": 4.7858,
      "mean_token_accuracy": 0.22934971749782562,
      "num_tokens": 51023004.0,
      "step": 22250
    },
    {
      "entropy": 5.030715656280518,
      "epoch": 2.1378482228626323,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004544315418650897,
      "loss": 4.7488,
      "mean_token_accuracy": 0.23142762035131453,
      "num_tokens": 51033993.0,
      "step": 22255
    },
    {
      "entropy": 5.062554502487183,
      "epoch": 2.138328530259366,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004544108546637008,
      "loss": 4.8336,
      "mean_token_accuracy": 0.232017982006073,
      "num_tokens": 51045269.0,
      "step": 22260
    },
    {
      "entropy": 5.158842706680298,
      "epoch": 2.1388088376561,
      "grad_norm": 1.125,
      "learning_rate": 0.0004543901632969247,
      "loss": 4.7627,
      "mean_token_accuracy": 0.231658872961998,
      "num_tokens": 51056044.0,
      "step": 22265
    },
    {
      "entropy": 5.176394414901734,
      "epoch": 2.139289145052834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045436946776524157,
      "loss": 4.8597,
      "mean_token_accuracy": 0.23165964782238008,
      "num_tokens": 51067096.0,
      "step": 22270
    },
    {
      "entropy": 5.1454658031463625,
      "epoch": 2.139769452449568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045434876806913204,
      "loss": 4.8645,
      "mean_token_accuracy": 0.22995989620685578,
      "num_tokens": 51080050.0,
      "step": 22275
    },
    {
      "entropy": 5.197056531906128,
      "epoch": 2.1402497598463017,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004543280642090767,
      "loss": 4.915,
      "mean_token_accuracy": 0.21641609370708464,
      "num_tokens": 51091333.0,
      "step": 22280
    },
    {
      "entropy": 5.2310422420501705,
      "epoch": 2.1407300672430356,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004543073561855562,
      "loss": 4.9362,
      "mean_token_accuracy": 0.2179610848426819,
      "num_tokens": 51102508.0,
      "step": 22285
    },
    {
      "entropy": 5.155963611602783,
      "epoch": 2.1412103746397695,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004542866439990513,
      "loss": 4.8546,
      "mean_token_accuracy": 0.2243368223309517,
      "num_tokens": 51114427.0,
      "step": 22290
    },
    {
      "entropy": 5.190688228607177,
      "epoch": 2.1416906820365034,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004542659276500429,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22406913042068483,
      "num_tokens": 51126486.0,
      "step": 22295
    },
    {
      "entropy": 5.134947919845581,
      "epoch": 2.1421709894332372,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045424520713901204,
      "loss": 4.8136,
      "mean_token_accuracy": 0.23166466355323792,
      "num_tokens": 51136813.0,
      "step": 22300
    },
    {
      "entropy": 5.087228059768677,
      "epoch": 2.142651296829971,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004542244824664396,
      "loss": 4.7382,
      "mean_token_accuracy": 0.23001331388950347,
      "num_tokens": 51148257.0,
      "step": 22305
    },
    {
      "entropy": 5.234271717071533,
      "epoch": 2.143131604226705,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045420375363280696,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2128526896238327,
      "num_tokens": 51159159.0,
      "step": 22310
    },
    {
      "entropy": 5.120167016983032,
      "epoch": 2.143611911623439,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045418302063859526,
      "loss": 4.8193,
      "mean_token_accuracy": 0.23187129199504852,
      "num_tokens": 51170900.0,
      "step": 22315
    },
    {
      "entropy": 5.270485067367554,
      "epoch": 2.1440922190201728,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045416228348428583,
      "loss": 4.9894,
      "mean_token_accuracy": 0.21440930664539337,
      "num_tokens": 51183707.0,
      "step": 22320
    },
    {
      "entropy": 5.127936124801636,
      "epoch": 2.1445725264169067,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045414154217036023,
      "loss": 4.7924,
      "mean_token_accuracy": 0.2344335749745369,
      "num_tokens": 51195114.0,
      "step": 22325
    },
    {
      "entropy": 5.103320598602295,
      "epoch": 2.1450528338136405,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045412079669730006,
      "loss": 4.7976,
      "mean_token_accuracy": 0.2332212060689926,
      "num_tokens": 51206210.0,
      "step": 22330
    },
    {
      "entropy": 5.127410936355591,
      "epoch": 2.1455331412103744,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004541000470655867,
      "loss": 4.8539,
      "mean_token_accuracy": 0.23173618018627168,
      "num_tokens": 51217657.0,
      "step": 22335
    },
    {
      "entropy": 5.233267498016358,
      "epoch": 2.1460134486071087,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045407929327570215,
      "loss": 4.9481,
      "mean_token_accuracy": 0.21768356561660768,
      "num_tokens": 51227836.0,
      "step": 22340
    },
    {
      "entropy": 5.15466628074646,
      "epoch": 2.1464937560038426,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004540585353281282,
      "loss": 4.8464,
      "mean_token_accuracy": 0.2337731808423996,
      "num_tokens": 51238936.0,
      "step": 22345
    },
    {
      "entropy": 5.208304595947266,
      "epoch": 2.1469740634005765,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004540377732233467,
      "loss": 4.9981,
      "mean_token_accuracy": 0.22375470995903016,
      "num_tokens": 51250696.0,
      "step": 22350
    },
    {
      "entropy": 5.265272331237793,
      "epoch": 2.1474543707973104,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004540170069618397,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22573766708374024,
      "num_tokens": 51262414.0,
      "step": 22355
    },
    {
      "entropy": 5.087259721755982,
      "epoch": 2.1479346781940443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045399623654408946,
      "loss": 4.8134,
      "mean_token_accuracy": 0.22306705564260482,
      "num_tokens": 51273267.0,
      "step": 22360
    },
    {
      "entropy": 5.094670104980469,
      "epoch": 2.148414985590778,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000453975461970578,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22763488292694092,
      "num_tokens": 51284783.0,
      "step": 22365
    },
    {
      "entropy": 5.17286319732666,
      "epoch": 2.148895292987512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004539546832417879,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22677937746047974,
      "num_tokens": 51296048.0,
      "step": 22370
    },
    {
      "entropy": 5.1727564334869385,
      "epoch": 2.149375600384246,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045393390035820136,
      "loss": 4.8573,
      "mean_token_accuracy": 0.22702993899583818,
      "num_tokens": 51306731.0,
      "step": 22375
    },
    {
      "entropy": 5.1857359409332275,
      "epoch": 2.14985590778098,
      "grad_norm": 1.09375,
      "learning_rate": 0.000453913113320301,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2177337408065796,
      "num_tokens": 51318431.0,
      "step": 22380
    },
    {
      "entropy": 5.224903154373169,
      "epoch": 2.1503362151777137,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004538923221285694,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22753535211086273,
      "num_tokens": 51330585.0,
      "step": 22385
    },
    {
      "entropy": 5.1841073513031,
      "epoch": 2.1508165225744476,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004538715267834893,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2226787105202675,
      "num_tokens": 51341936.0,
      "step": 22390
    },
    {
      "entropy": 5.170021390914917,
      "epoch": 2.1512968299711814,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004538507272855434,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22319029569625853,
      "num_tokens": 51353270.0,
      "step": 22395
    },
    {
      "entropy": 5.265185165405273,
      "epoch": 2.1517771373679153,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045382992363521486,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21633676886558534,
      "num_tokens": 51365470.0,
      "step": 22400
    },
    {
      "entropy": 5.183343601226807,
      "epoch": 2.152257444764649,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045380911583298633,
      "loss": 4.8935,
      "mean_token_accuracy": 0.224540314078331,
      "num_tokens": 51375911.0,
      "step": 22405
    },
    {
      "entropy": 5.17657699584961,
      "epoch": 2.152737752161383,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045378830387934123,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2181590884923935,
      "num_tokens": 51387082.0,
      "step": 22410
    },
    {
      "entropy": 5.179660844802856,
      "epoch": 2.1532180595581174,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004537674877747626,
      "loss": 4.8753,
      "mean_token_accuracy": 0.22653546929359436,
      "num_tokens": 51398826.0,
      "step": 22415
    },
    {
      "entropy": 5.216093635559082,
      "epoch": 2.1536983669548513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045374666751973365,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22463333308696748,
      "num_tokens": 51410848.0,
      "step": 22420
    },
    {
      "entropy": 5.176193571090698,
      "epoch": 2.154178674351585,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045372584311473784,
      "loss": 4.7816,
      "mean_token_accuracy": 0.23068543374538422,
      "num_tokens": 51422272.0,
      "step": 22425
    },
    {
      "entropy": 5.061978006362915,
      "epoch": 2.154658981748319,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004537050145602587,
      "loss": 4.8103,
      "mean_token_accuracy": 0.22533255070447922,
      "num_tokens": 51433139.0,
      "step": 22430
    },
    {
      "entropy": 5.165013408660888,
      "epoch": 2.155139289145053,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004536841818567798,
      "loss": 4.83,
      "mean_token_accuracy": 0.22087880671024324,
      "num_tokens": 51445582.0,
      "step": 22435
    },
    {
      "entropy": 5.150230550765992,
      "epoch": 2.155619596541787,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004536633450047847,
      "loss": 4.7572,
      "mean_token_accuracy": 0.23413965702056885,
      "num_tokens": 51456859.0,
      "step": 22440
    },
    {
      "entropy": 5.172560453414917,
      "epoch": 2.1560999039385207,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045364250400475734,
      "loss": 4.9348,
      "mean_token_accuracy": 0.2235237702727318,
      "num_tokens": 51468939.0,
      "step": 22445
    },
    {
      "entropy": 5.16064567565918,
      "epoch": 2.1565802113352546,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004536216588571814,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22203755229711533,
      "num_tokens": 51480390.0,
      "step": 22450
    },
    {
      "entropy": 5.1658436298370365,
      "epoch": 2.1570605187319885,
      "grad_norm": 1.046875,
      "learning_rate": 0.000453600809562541,
      "loss": 4.8668,
      "mean_token_accuracy": 0.22938430309295654,
      "num_tokens": 51493194.0,
      "step": 22455
    },
    {
      "entropy": 5.153738784790039,
      "epoch": 2.1575408261287223,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004535799561213202,
      "loss": 4.848,
      "mean_token_accuracy": 0.225484174489975,
      "num_tokens": 51505084.0,
      "step": 22460
    },
    {
      "entropy": 5.228030586242676,
      "epoch": 2.158021133525456,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000453559098534003,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22817500531673432,
      "num_tokens": 51516474.0,
      "step": 22465
    },
    {
      "entropy": 5.208297824859619,
      "epoch": 2.15850144092219,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004535382368010738,
      "loss": 4.889,
      "mean_token_accuracy": 0.2304280161857605,
      "num_tokens": 51526734.0,
      "step": 22470
    },
    {
      "entropy": 5.12373456954956,
      "epoch": 2.158981748318924,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045351737092301676,
      "loss": 4.8565,
      "mean_token_accuracy": 0.22616606056690217,
      "num_tokens": 51538230.0,
      "step": 22475
    },
    {
      "entropy": 5.2376728534698485,
      "epoch": 2.159462055715658,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004534965009003165,
      "loss": 4.8788,
      "mean_token_accuracy": 0.225794917345047,
      "num_tokens": 51551671.0,
      "step": 22480
    },
    {
      "entropy": 5.169781827926636,
      "epoch": 2.1599423631123917,
      "grad_norm": 1.25,
      "learning_rate": 0.0004534756267334576,
      "loss": 4.8525,
      "mean_token_accuracy": 0.22802109718322755,
      "num_tokens": 51563034.0,
      "step": 22485
    },
    {
      "entropy": 5.105741786956787,
      "epoch": 2.160422670509126,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045345474842292455,
      "loss": 4.7578,
      "mean_token_accuracy": 0.23153412491083145,
      "num_tokens": 51575210.0,
      "step": 22490
    },
    {
      "entropy": 5.18654465675354,
      "epoch": 2.16090297790586,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004534338659692022,
      "loss": 4.9345,
      "mean_token_accuracy": 0.222044537961483,
      "num_tokens": 51585586.0,
      "step": 22495
    },
    {
      "entropy": 5.202785539627075,
      "epoch": 2.161383285302594,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004534129793727753,
      "loss": 4.909,
      "mean_token_accuracy": 0.2195098415017128,
      "num_tokens": 51596519.0,
      "step": 22500
    },
    {
      "entropy": 5.19596495628357,
      "epoch": 2.1618635926993277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004533920886341288,
      "loss": 4.8526,
      "mean_token_accuracy": 0.2313298538327217,
      "num_tokens": 51608421.0,
      "step": 22505
    },
    {
      "entropy": 5.211212587356568,
      "epoch": 2.1623439000960616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004533711937537477,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2193269595503807,
      "num_tokens": 51620615.0,
      "step": 22510
    },
    {
      "entropy": 5.230288648605347,
      "epoch": 2.1628242074927955,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004533502947321171,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22001649737358092,
      "num_tokens": 51632797.0,
      "step": 22515
    },
    {
      "entropy": 5.163245725631714,
      "epoch": 2.1633045148895294,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004533293915697223,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2193766489624977,
      "num_tokens": 51644735.0,
      "step": 22520
    },
    {
      "entropy": 5.2032817840576175,
      "epoch": 2.1637848222862632,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045330848426704853,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2195388838648796,
      "num_tokens": 51656270.0,
      "step": 22525
    },
    {
      "entropy": 5.161894035339356,
      "epoch": 2.164265129682997,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004532875728245813,
      "loss": 4.829,
      "mean_token_accuracy": 0.2310422331094742,
      "num_tokens": 51668223.0,
      "step": 22530
    },
    {
      "entropy": 5.111179971694947,
      "epoch": 2.164745437079731,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045326665724280594,
      "loss": 4.7942,
      "mean_token_accuracy": 0.23143746703863144,
      "num_tokens": 51680298.0,
      "step": 22535
    },
    {
      "entropy": 5.235932493209839,
      "epoch": 2.165225744476465,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045324573752220814,
      "loss": 4.979,
      "mean_token_accuracy": 0.21768402606248854,
      "num_tokens": 51692803.0,
      "step": 22540
    },
    {
      "entropy": 5.180206632614135,
      "epoch": 2.1657060518731988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045322481366327365,
      "loss": 4.852,
      "mean_token_accuracy": 0.2232673928141594,
      "num_tokens": 51704478.0,
      "step": 22545
    },
    {
      "entropy": 5.17021803855896,
      "epoch": 2.1661863592699326,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004532038856664882,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22774208784103395,
      "num_tokens": 51714627.0,
      "step": 22550
    },
    {
      "entropy": 5.146081733703613,
      "epoch": 2.1666666666666665,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004531829535323376,
      "loss": 4.8197,
      "mean_token_accuracy": 0.2340054214000702,
      "num_tokens": 51726297.0,
      "step": 22555
    },
    {
      "entropy": 5.134461307525635,
      "epoch": 2.1671469740634004,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000453162017261308,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22708771973848343,
      "num_tokens": 51737676.0,
      "step": 22560
    },
    {
      "entropy": 5.1653602600097654,
      "epoch": 2.1676272814601343,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004531410768538854,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22862183153629304,
      "num_tokens": 51749532.0,
      "step": 22565
    },
    {
      "entropy": 5.162664747238159,
      "epoch": 2.168107588856868,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045312013231055596,
      "loss": 4.8997,
      "mean_token_accuracy": 0.22932156473398208,
      "num_tokens": 51761047.0,
      "step": 22570
    },
    {
      "entropy": 5.247588682174682,
      "epoch": 2.1685878962536025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045309918363180593,
      "loss": 4.8583,
      "mean_token_accuracy": 0.22851041406393052,
      "num_tokens": 51772134.0,
      "step": 22575
    },
    {
      "entropy": 5.288985538482666,
      "epoch": 2.1690682036503364,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045307823081812166,
      "loss": 4.9705,
      "mean_token_accuracy": 0.2183770701289177,
      "num_tokens": 51785775.0,
      "step": 22580
    },
    {
      "entropy": 5.126558780670166,
      "epoch": 2.1695485110470702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045305727386998977,
      "loss": 4.8252,
      "mean_token_accuracy": 0.22597796469926834,
      "num_tokens": 51797306.0,
      "step": 22585
    },
    {
      "entropy": 5.158930063247681,
      "epoch": 2.170028818443804,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004530363127878966,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21813196539878846,
      "num_tokens": 51809312.0,
      "step": 22590
    },
    {
      "entropy": 5.191654872894287,
      "epoch": 2.170509125840538,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045301534757232885,
      "loss": 4.8814,
      "mean_token_accuracy": 0.23201687037944793,
      "num_tokens": 51821646.0,
      "step": 22595
    },
    {
      "entropy": 5.145369672775269,
      "epoch": 2.170989433237272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004529943782237735,
      "loss": 4.7725,
      "mean_token_accuracy": 0.2311073660850525,
      "num_tokens": 51831546.0,
      "step": 22600
    },
    {
      "entropy": 5.163899755477905,
      "epoch": 2.1714697406340058,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045297340474271717,
      "loss": 4.8954,
      "mean_token_accuracy": 0.2291557103395462,
      "num_tokens": 51842273.0,
      "step": 22605
    },
    {
      "entropy": 5.212971305847168,
      "epoch": 2.1719500480307397,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004529524271296468,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22350149601697922,
      "num_tokens": 51854715.0,
      "step": 22610
    },
    {
      "entropy": 5.178461790084839,
      "epoch": 2.1724303554274735,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045293144538504943,
      "loss": 4.8701,
      "mean_token_accuracy": 0.22189754247665405,
      "num_tokens": 51866819.0,
      "step": 22615
    },
    {
      "entropy": 5.156960773468017,
      "epoch": 2.1729106628242074,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004529104595094124,
      "loss": 4.8378,
      "mean_token_accuracy": 0.23432862758636475,
      "num_tokens": 51878651.0,
      "step": 22620
    },
    {
      "entropy": 5.196591424942016,
      "epoch": 2.1733909702209413,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045288946950322264,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22412749677896499,
      "num_tokens": 51888640.0,
      "step": 22625
    },
    {
      "entropy": 5.111431217193603,
      "epoch": 2.173871277617675,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004528684753669677,
      "loss": 4.8109,
      "mean_token_accuracy": 0.23219498246908188,
      "num_tokens": 51899810.0,
      "step": 22630
    },
    {
      "entropy": 5.161106824874878,
      "epoch": 2.174351585014409,
      "grad_norm": 1.25,
      "learning_rate": 0.0004528474771011349,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2210506707429886,
      "num_tokens": 51910985.0,
      "step": 22635
    },
    {
      "entropy": 5.16982889175415,
      "epoch": 2.174831892411143,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045282647470621176,
      "loss": 4.8242,
      "mean_token_accuracy": 0.23005418330430985,
      "num_tokens": 51922550.0,
      "step": 22640
    },
    {
      "entropy": 5.094676971435547,
      "epoch": 2.175312199807877,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045280546818268595,
      "loss": 4.7333,
      "mean_token_accuracy": 0.2304110497236252,
      "num_tokens": 51934319.0,
      "step": 22645
    },
    {
      "entropy": 5.141055679321289,
      "epoch": 2.175792507204611,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004527844575310452,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21391458958387374,
      "num_tokens": 51945516.0,
      "step": 22650
    },
    {
      "entropy": 5.24350938796997,
      "epoch": 2.176272814601345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045276344275177715,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2190658837556839,
      "num_tokens": 51957033.0,
      "step": 22655
    },
    {
      "entropy": 5.181605195999145,
      "epoch": 2.176753121998079,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045274242384536984,
      "loss": 4.8652,
      "mean_token_accuracy": 0.22380332946777343,
      "num_tokens": 51968211.0,
      "step": 22660
    },
    {
      "entropy": 5.084350633621216,
      "epoch": 2.177233429394813,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004527214008123113,
      "loss": 4.7784,
      "mean_token_accuracy": 0.2312314361333847,
      "num_tokens": 51978709.0,
      "step": 22665
    },
    {
      "entropy": 5.144052934646607,
      "epoch": 2.1777137367915467,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004527003736530895,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22015300989151002,
      "num_tokens": 51989997.0,
      "step": 22670
    },
    {
      "entropy": 5.113739728927612,
      "epoch": 2.1781940441882806,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045267934236819265,
      "loss": 4.8288,
      "mean_token_accuracy": 0.22923711985349654,
      "num_tokens": 52001730.0,
      "step": 22675
    },
    {
      "entropy": 5.239670276641846,
      "epoch": 2.1786743515850144,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004526583069581091,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2210412949323654,
      "num_tokens": 52012267.0,
      "step": 22680
    },
    {
      "entropy": 5.242125225067139,
      "epoch": 2.1791546589817483,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004526372674233272,
      "loss": 4.943,
      "mean_token_accuracy": 0.21906615793704987,
      "num_tokens": 52024700.0,
      "step": 22685
    },
    {
      "entropy": 5.212616586685181,
      "epoch": 2.179634966378482,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045261622376433543,
      "loss": 4.8583,
      "mean_token_accuracy": 0.2245834067463875,
      "num_tokens": 52036685.0,
      "step": 22690
    },
    {
      "entropy": 5.170576333999634,
      "epoch": 2.180115273775216,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045259517598162237,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22653224170207978,
      "num_tokens": 52048911.0,
      "step": 22695
    },
    {
      "entropy": 5.204931974411011,
      "epoch": 2.18059558117195,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004525741240756766,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2263164758682251,
      "num_tokens": 52060039.0,
      "step": 22700
    },
    {
      "entropy": 5.142710256576538,
      "epoch": 2.181075888568684,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004525530680469871,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22858888655900955,
      "num_tokens": 52070549.0,
      "step": 22705
    },
    {
      "entropy": 5.224255847930908,
      "epoch": 2.1815561959654177,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045253200789604245,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21734580099582673,
      "num_tokens": 52081573.0,
      "step": 22710
    },
    {
      "entropy": 5.146978282928467,
      "epoch": 2.1820365033621516,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045251094362333186,
      "loss": 4.8624,
      "mean_token_accuracy": 0.2266298145055771,
      "num_tokens": 52093694.0,
      "step": 22715
    },
    {
      "entropy": 5.214025211334229,
      "epoch": 2.1825168107588855,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004524898752293441,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21957649886608124,
      "num_tokens": 52104784.0,
      "step": 22720
    },
    {
      "entropy": 5.231395483016968,
      "epoch": 2.18299711815562,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045246880271456857,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22006892412900925,
      "num_tokens": 52116444.0,
      "step": 22725
    },
    {
      "entropy": 5.087467050552368,
      "epoch": 2.1834774255523537,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004524477260794944,
      "loss": 4.8136,
      "mean_token_accuracy": 0.2303192973136902,
      "num_tokens": 52130264.0,
      "step": 22730
    },
    {
      "entropy": 5.149377059936524,
      "epoch": 2.1839577329490876,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045242664532461094,
      "loss": 4.8599,
      "mean_token_accuracy": 0.21612814515829087,
      "num_tokens": 52142688.0,
      "step": 22735
    },
    {
      "entropy": 5.264934635162353,
      "epoch": 2.1844380403458215,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045240556045040767,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21304726153612136,
      "num_tokens": 52154823.0,
      "step": 22740
    },
    {
      "entropy": 5.1733814716339115,
      "epoch": 2.1849183477425553,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045238447145737397,
      "loss": 4.8906,
      "mean_token_accuracy": 0.227194944024086,
      "num_tokens": 52165542.0,
      "step": 22745
    },
    {
      "entropy": 5.2543638229370115,
      "epoch": 2.185398655139289,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045236337834599966,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22979568243026732,
      "num_tokens": 52176475.0,
      "step": 22750
    },
    {
      "entropy": 5.109518194198609,
      "epoch": 2.185878962536023,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045234228111677434,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22141188830137254,
      "num_tokens": 52188467.0,
      "step": 22755
    },
    {
      "entropy": 5.170274448394776,
      "epoch": 2.186359269932757,
      "grad_norm": 1.125,
      "learning_rate": 0.0004523211797701878,
      "loss": 4.8056,
      "mean_token_accuracy": 0.22970423847436905,
      "num_tokens": 52199518.0,
      "step": 22760
    },
    {
      "entropy": 5.123329401016235,
      "epoch": 2.186839577329491,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045230007430673014,
      "loss": 4.8603,
      "mean_token_accuracy": 0.2266884848475456,
      "num_tokens": 52211429.0,
      "step": 22765
    },
    {
      "entropy": 5.2144537448883055,
      "epoch": 2.1873198847262247,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004522789647268911,
      "loss": 4.912,
      "mean_token_accuracy": 0.2235619455575943,
      "num_tokens": 52222258.0,
      "step": 22770
    },
    {
      "entropy": 5.1457315444946286,
      "epoch": 2.1878001921229586,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004522578510311609,
      "loss": 4.8124,
      "mean_token_accuracy": 0.22652983516454697,
      "num_tokens": 52233660.0,
      "step": 22775
    },
    {
      "entropy": 5.122683715820313,
      "epoch": 2.1882804995196925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045223673322002984,
      "loss": 4.834,
      "mean_token_accuracy": 0.2305731952190399,
      "num_tokens": 52244721.0,
      "step": 22780
    },
    {
      "entropy": 5.2564857482910154,
      "epoch": 2.1887608069164264,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045221561129398804,
      "loss": 4.9248,
      "mean_token_accuracy": 0.2205973222851753,
      "num_tokens": 52256099.0,
      "step": 22785
    },
    {
      "entropy": 5.159810590744018,
      "epoch": 2.1892411143131603,
      "grad_norm": 1.140625,
      "learning_rate": 0.000452194485253526,
      "loss": 4.872,
      "mean_token_accuracy": 0.22826613932847978,
      "num_tokens": 52266943.0,
      "step": 22790
    },
    {
      "entropy": 5.073787498474121,
      "epoch": 2.189721421709894,
      "grad_norm": 1.25,
      "learning_rate": 0.0004521733550991342,
      "loss": 4.861,
      "mean_token_accuracy": 0.2299852579832077,
      "num_tokens": 52279074.0,
      "step": 22795
    },
    {
      "entropy": 5.171516466140747,
      "epoch": 2.1902017291066285,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045215222083130316,
      "loss": 4.889,
      "mean_token_accuracy": 0.22333572506904603,
      "num_tokens": 52290613.0,
      "step": 22800
    },
    {
      "entropy": 5.215606546401977,
      "epoch": 2.1906820365033624,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004521310824505236,
      "loss": 4.8776,
      "mean_token_accuracy": 0.220848685503006,
      "num_tokens": 52301903.0,
      "step": 22805
    },
    {
      "entropy": 5.14372067451477,
      "epoch": 2.1911623439000962,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045210993995728623,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22416525483131408,
      "num_tokens": 52312738.0,
      "step": 22810
    },
    {
      "entropy": 5.128119802474975,
      "epoch": 2.19164265129683,
      "grad_norm": 1.015625,
      "learning_rate": 0.000452088793352082,
      "loss": 4.8282,
      "mean_token_accuracy": 0.2239854246377945,
      "num_tokens": 52325315.0,
      "step": 22815
    },
    {
      "entropy": 5.265017318725586,
      "epoch": 2.192122958693564,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004520676426354018,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2176401823759079,
      "num_tokens": 52338330.0,
      "step": 22820
    },
    {
      "entropy": 5.263670587539673,
      "epoch": 2.192603266090298,
      "grad_norm": 1.125,
      "learning_rate": 0.0004520464878077368,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21363643407821656,
      "num_tokens": 52349758.0,
      "step": 22825
    },
    {
      "entropy": 5.123260736465454,
      "epoch": 2.1930835734870318,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045202532886957805,
      "loss": 4.8247,
      "mean_token_accuracy": 0.22982172220945357,
      "num_tokens": 52361901.0,
      "step": 22830
    },
    {
      "entropy": 5.155756235122681,
      "epoch": 2.1935638808837656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045200416582141676,
      "loss": 4.8789,
      "mean_token_accuracy": 0.2261626586318016,
      "num_tokens": 52372912.0,
      "step": 22835
    },
    {
      "entropy": 5.115793037414551,
      "epoch": 2.1940441882804995,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004519829986637444,
      "loss": 4.8466,
      "mean_token_accuracy": 0.2281137526035309,
      "num_tokens": 52384270.0,
      "step": 22840
    },
    {
      "entropy": 5.213618230819702,
      "epoch": 2.1945244956772334,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004519618273970523,
      "loss": 4.9296,
      "mean_token_accuracy": 0.2197387382388115,
      "num_tokens": 52396555.0,
      "step": 22845
    },
    {
      "entropy": 5.1941437244415285,
      "epoch": 2.1950048030739673,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045194065202183205,
      "loss": 4.8502,
      "mean_token_accuracy": 0.22642536461353302,
      "num_tokens": 52408183.0,
      "step": 22850
    },
    {
      "entropy": 5.196371603012085,
      "epoch": 2.195485110470701,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004519194725385752,
      "loss": 5.0144,
      "mean_token_accuracy": 0.21943350732326508,
      "num_tokens": 52420285.0,
      "step": 22855
    },
    {
      "entropy": 5.208965158462524,
      "epoch": 2.195965417867435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045189828894777364,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22200609743595123,
      "num_tokens": 52431558.0,
      "step": 22860
    },
    {
      "entropy": 5.264218378067016,
      "epoch": 2.196445725264169,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045187710124991904,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21875936836004256,
      "num_tokens": 52443189.0,
      "step": 22865
    },
    {
      "entropy": 5.148245143890381,
      "epoch": 2.196926032660903,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004518559094455034,
      "loss": 4.8841,
      "mean_token_accuracy": 0.2238081783056259,
      "num_tokens": 52454781.0,
      "step": 22870
    },
    {
      "entropy": 5.098171377182007,
      "epoch": 2.1974063400576367,
      "grad_norm": 1.25,
      "learning_rate": 0.0004518347135350187,
      "loss": 4.8245,
      "mean_token_accuracy": 0.23320367187261581,
      "num_tokens": 52466424.0,
      "step": 22875
    },
    {
      "entropy": 5.169333171844483,
      "epoch": 2.1978866474543706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045181351351895703,
      "loss": 4.7857,
      "mean_token_accuracy": 0.23204689472913742,
      "num_tokens": 52477043.0,
      "step": 22880
    },
    {
      "entropy": 5.063090467453003,
      "epoch": 2.198366954851105,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004517923093978106,
      "loss": 4.7813,
      "mean_token_accuracy": 0.2327408656477928,
      "num_tokens": 52488321.0,
      "step": 22885
    },
    {
      "entropy": 5.149637460708618,
      "epoch": 2.1988472622478388,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004517711011720717,
      "loss": 4.8665,
      "mean_token_accuracy": 0.22444213628768922,
      "num_tokens": 52499071.0,
      "step": 22890
    },
    {
      "entropy": 5.174709701538086,
      "epoch": 2.1993275696445727,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004517498888422327,
      "loss": 4.8043,
      "mean_token_accuracy": 0.23680397719144822,
      "num_tokens": 52509846.0,
      "step": 22895
    },
    {
      "entropy": 5.252161502838135,
      "epoch": 2.1998078770413065,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004517286724087862,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2149966388940811,
      "num_tokens": 52521450.0,
      "step": 22900
    },
    {
      "entropy": 5.133204650878906,
      "epoch": 2.2002881844380404,
      "grad_norm": 1.125,
      "learning_rate": 0.0004517074518722247,
      "loss": 4.8731,
      "mean_token_accuracy": 0.23080717474222184,
      "num_tokens": 52533115.0,
      "step": 22905
    },
    {
      "entropy": 5.085250186920166,
      "epoch": 2.2007684918347743,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045168622723304084,
      "loss": 4.7806,
      "mean_token_accuracy": 0.23866580575704574,
      "num_tokens": 52543921.0,
      "step": 22910
    },
    {
      "entropy": 5.201054096221924,
      "epoch": 2.201248799231508,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004516649984917274,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23130674362182618,
      "num_tokens": 52554956.0,
      "step": 22915
    },
    {
      "entropy": 5.136899805068969,
      "epoch": 2.201729106628242,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045164376564877734,
      "loss": 4.8429,
      "mean_token_accuracy": 0.22546664476394654,
      "num_tokens": 52567386.0,
      "step": 22920
    },
    {
      "entropy": 5.184470558166504,
      "epoch": 2.202209414024976,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045162252870468354,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2255799502134323,
      "num_tokens": 52578533.0,
      "step": 22925
    },
    {
      "entropy": 5.26426248550415,
      "epoch": 2.20268972142171,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004516012876599391,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2140755146741867,
      "num_tokens": 52590283.0,
      "step": 22930
    },
    {
      "entropy": 5.126363229751587,
      "epoch": 2.2031700288184437,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045158004251503715,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23382412493228913,
      "num_tokens": 52602362.0,
      "step": 22935
    },
    {
      "entropy": 5.05814847946167,
      "epoch": 2.2036503362151776,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045155879327047087,
      "loss": 4.7797,
      "mean_token_accuracy": 0.2349344462156296,
      "num_tokens": 52614057.0,
      "step": 22940
    },
    {
      "entropy": 5.139747571945191,
      "epoch": 2.2041306436119115,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004515375399267338,
      "loss": 4.8582,
      "mean_token_accuracy": 0.2230058044195175,
      "num_tokens": 52624298.0,
      "step": 22945
    },
    {
      "entropy": 5.204195261001587,
      "epoch": 2.2046109510086453,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045151628248431925,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22475963681936265,
      "num_tokens": 52637043.0,
      "step": 22950
    },
    {
      "entropy": 5.199544286727905,
      "epoch": 2.2050912584053792,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045149502094372077,
      "loss": 4.9783,
      "mean_token_accuracy": 0.21587668657302855,
      "num_tokens": 52649326.0,
      "step": 22955
    },
    {
      "entropy": 5.206510877609253,
      "epoch": 2.2055715658021136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045147375530543195,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22508661150932313,
      "num_tokens": 52660973.0,
      "step": 22960
    },
    {
      "entropy": 5.1392961025238035,
      "epoch": 2.2060518731988474,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045145248556994653,
      "loss": 4.8058,
      "mean_token_accuracy": 0.23188504576683044,
      "num_tokens": 52671943.0,
      "step": 22965
    },
    {
      "entropy": 5.153354454040527,
      "epoch": 2.2065321805955813,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004514312117377584,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22007160037755966,
      "num_tokens": 52684634.0,
      "step": 22970
    },
    {
      "entropy": 5.151382303237915,
      "epoch": 2.207012487992315,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004514099338093613,
      "loss": 4.8718,
      "mean_token_accuracy": 0.23079041838645936,
      "num_tokens": 52695681.0,
      "step": 22975
    },
    {
      "entropy": 5.097003507614136,
      "epoch": 2.207492795389049,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004513886517852496,
      "loss": 4.7945,
      "mean_token_accuracy": 0.23297179043292998,
      "num_tokens": 52706967.0,
      "step": 22980
    },
    {
      "entropy": 5.177677297592163,
      "epoch": 2.207973102785783,
      "grad_norm": 1.015625,
      "learning_rate": 0.000451367365665917,
      "loss": 4.877,
      "mean_token_accuracy": 0.22195524573326111,
      "num_tokens": 52719226.0,
      "step": 22985
    },
    {
      "entropy": 5.045239877700806,
      "epoch": 2.208453410182517,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045134607545185785,
      "loss": 4.7282,
      "mean_token_accuracy": 0.2347163364291191,
      "num_tokens": 52730141.0,
      "step": 22990
    },
    {
      "entropy": 5.175697088241577,
      "epoch": 2.2089337175792507,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004513247811435666,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21491020768880845,
      "num_tokens": 52741059.0,
      "step": 22995
    },
    {
      "entropy": 5.247995281219483,
      "epoch": 2.2094140249759846,
      "grad_norm": 1.125,
      "learning_rate": 0.00045130348274153735,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2195878267288208,
      "num_tokens": 52752598.0,
      "step": 23000
    },
    {
      "entropy": 5.103432893753052,
      "epoch": 2.2098943323727185,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045128218024626486,
      "loss": 4.7953,
      "mean_token_accuracy": 0.22948150485754013,
      "num_tokens": 52764620.0,
      "step": 23005
    },
    {
      "entropy": 5.135836505889893,
      "epoch": 2.2103746397694524,
      "grad_norm": 1.125,
      "learning_rate": 0.0004512608736582436,
      "loss": 4.8076,
      "mean_token_accuracy": 0.22885997742414474,
      "num_tokens": 52774756.0,
      "step": 23010
    },
    {
      "entropy": 5.253208351135254,
      "epoch": 2.2108549471661862,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004512395629779682,
      "loss": 4.9594,
      "mean_token_accuracy": 0.2130097508430481,
      "num_tokens": 52786742.0,
      "step": 23015
    },
    {
      "entropy": 5.236729955673217,
      "epoch": 2.21133525456292,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004512182482059335,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22138423919677735,
      "num_tokens": 52798303.0,
      "step": 23020
    },
    {
      "entropy": 5.147168779373169,
      "epoch": 2.211815561959654,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004511969293426343,
      "loss": 4.8619,
      "mean_token_accuracy": 0.22781068533658982,
      "num_tokens": 52808930.0,
      "step": 23025
    },
    {
      "entropy": 5.117085647583008,
      "epoch": 2.212295869356388,
      "grad_norm": 1.6875,
      "learning_rate": 0.00045117560638856567,
      "loss": 4.8137,
      "mean_token_accuracy": 0.22577569484710694,
      "num_tokens": 52820066.0,
      "step": 23030
    },
    {
      "entropy": 5.167489290237427,
      "epoch": 2.212776176753122,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004511542793442225,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22667156159877777,
      "num_tokens": 52830839.0,
      "step": 23035
    },
    {
      "entropy": 5.187667655944824,
      "epoch": 2.213256484149856,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004511329482101001,
      "loss": 4.8735,
      "mean_token_accuracy": 0.21954104155302048,
      "num_tokens": 52843112.0,
      "step": 23040
    },
    {
      "entropy": 5.208319759368896,
      "epoch": 2.21373679154659,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004511116129866936,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22177437394857408,
      "num_tokens": 52854737.0,
      "step": 23045
    },
    {
      "entropy": 5.268733787536621,
      "epoch": 2.214217098943324,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045109027367449845,
      "loss": 4.9489,
      "mean_token_accuracy": 0.21879530251026152,
      "num_tokens": 52866876.0,
      "step": 23050
    },
    {
      "entropy": 5.235713100433349,
      "epoch": 2.2146974063400577,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045106893027400995,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22032684534788133,
      "num_tokens": 52878731.0,
      "step": 23055
    },
    {
      "entropy": 5.206970071792602,
      "epoch": 2.2151777137367916,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045104758278572375,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21747656613588334,
      "num_tokens": 52891011.0,
      "step": 23060
    },
    {
      "entropy": 5.09006896018982,
      "epoch": 2.2156580211335255,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004510262312101355,
      "loss": 4.8347,
      "mean_token_accuracy": 0.2280296951532364,
      "num_tokens": 52902334.0,
      "step": 23065
    },
    {
      "entropy": 5.273526763916015,
      "epoch": 2.2161383285302594,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004510048755477407,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2134373441338539,
      "num_tokens": 52913839.0,
      "step": 23070
    },
    {
      "entropy": 5.212164068222046,
      "epoch": 2.2166186359269933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004509835157990354,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22260777354240419,
      "num_tokens": 52925588.0,
      "step": 23075
    },
    {
      "entropy": 5.317955350875854,
      "epoch": 2.217098943323727,
      "grad_norm": 1.0,
      "learning_rate": 0.00045096215196451547,
      "loss": 5.0272,
      "mean_token_accuracy": 0.21796328872442244,
      "num_tokens": 52938474.0,
      "step": 23080
    },
    {
      "entropy": 5.110657501220703,
      "epoch": 2.217579250720461,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045094078404467683,
      "loss": 4.7958,
      "mean_token_accuracy": 0.22907410562038422,
      "num_tokens": 52950157.0,
      "step": 23085
    },
    {
      "entropy": 5.15514497756958,
      "epoch": 2.218059558117195,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045091941204001564,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22407884299755096,
      "num_tokens": 52960334.0,
      "step": 23090
    },
    {
      "entropy": 5.179567003250122,
      "epoch": 2.218539865513929,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000450898035951028,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22593673765659333,
      "num_tokens": 52972251.0,
      "step": 23095
    },
    {
      "entropy": 5.217549800872803,
      "epoch": 2.2190201729106627,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045087665577821034,
      "loss": 4.8505,
      "mean_token_accuracy": 0.22565948814153672,
      "num_tokens": 52983897.0,
      "step": 23100
    },
    {
      "entropy": 5.275551128387451,
      "epoch": 2.2195004803073966,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000450855271522059,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2170689269900322,
      "num_tokens": 52995821.0,
      "step": 23105
    },
    {
      "entropy": 5.180625295639038,
      "epoch": 2.219980787704131,
      "grad_norm": 1.0,
      "learning_rate": 0.00045083388318307044,
      "loss": 4.8853,
      "mean_token_accuracy": 0.2227293610572815,
      "num_tokens": 53007890.0,
      "step": 23110
    },
    {
      "entropy": 5.158633422851563,
      "epoch": 2.2204610951008648,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004508124907617411,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22459561377763748,
      "num_tokens": 53019390.0,
      "step": 23115
    },
    {
      "entropy": 5.148567485809326,
      "epoch": 2.2209414024975986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004507910942585679,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22907427698373795,
      "num_tokens": 53031318.0,
      "step": 23120
    },
    {
      "entropy": 5.155513525009155,
      "epoch": 2.2214217098943325,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004507696936740475,
      "loss": 4.7647,
      "mean_token_accuracy": 0.23568036705255507,
      "num_tokens": 53043121.0,
      "step": 23125
    },
    {
      "entropy": 5.137007236480713,
      "epoch": 2.2219020172910664,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004507482890086767,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22421053647994996,
      "num_tokens": 53055600.0,
      "step": 23130
    },
    {
      "entropy": 5.055635976791382,
      "epoch": 2.2223823246878003,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004507268802629525,
      "loss": 4.7657,
      "mean_token_accuracy": 0.23584290742874145,
      "num_tokens": 53067075.0,
      "step": 23135
    },
    {
      "entropy": 5.1109155178070065,
      "epoch": 2.222862632084534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004507054674373719,
      "loss": 4.8531,
      "mean_token_accuracy": 0.23000891506671906,
      "num_tokens": 53078374.0,
      "step": 23140
    },
    {
      "entropy": 5.1777702331542965,
      "epoch": 2.223342939481268,
      "grad_norm": 1.0,
      "learning_rate": 0.00045068405053243216,
      "loss": 4.8245,
      "mean_token_accuracy": 0.23206623941659926,
      "num_tokens": 53089177.0,
      "step": 23145
    },
    {
      "entropy": 5.075928735733032,
      "epoch": 2.223823246878002,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004506626295486304,
      "loss": 4.7156,
      "mean_token_accuracy": 0.23148219138383866,
      "num_tokens": 53099957.0,
      "step": 23150
    },
    {
      "entropy": 5.1710474491119385,
      "epoch": 2.224303554274736,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045064120448646405,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2264367312192917,
      "num_tokens": 53113369.0,
      "step": 23155
    },
    {
      "entropy": 5.152922677993774,
      "epoch": 2.2247838616714697,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004506197753464304,
      "loss": 4.8139,
      "mean_token_accuracy": 0.2275436833500862,
      "num_tokens": 53124647.0,
      "step": 23160
    },
    {
      "entropy": 5.19539475440979,
      "epoch": 2.2252641690682036,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045059834212902707,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22389136105775834,
      "num_tokens": 53135218.0,
      "step": 23165
    },
    {
      "entropy": 5.103829050064087,
      "epoch": 2.2257444764649374,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045057690483475167,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22636160552501677,
      "num_tokens": 53148077.0,
      "step": 23170
    },
    {
      "entropy": 5.181288766860962,
      "epoch": 2.2262247838616713,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004505554634641019,
      "loss": 4.901,
      "mean_token_accuracy": 0.22757795453071594,
      "num_tokens": 53159866.0,
      "step": 23175
    },
    {
      "entropy": 5.200931596755981,
      "epoch": 2.226705091258405,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045053401801757554,
      "loss": 4.8515,
      "mean_token_accuracy": 0.22717494815587996,
      "num_tokens": 53171151.0,
      "step": 23180
    },
    {
      "entropy": 5.19863452911377,
      "epoch": 2.227185398655139,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045051256849567054,
      "loss": 4.8767,
      "mean_token_accuracy": 0.22479392290115358,
      "num_tokens": 53182400.0,
      "step": 23185
    },
    {
      "entropy": 5.187387323379516,
      "epoch": 2.227665706051873,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045049111489888486,
      "loss": 5.0035,
      "mean_token_accuracy": 0.2139649733901024,
      "num_tokens": 53194248.0,
      "step": 23190
    },
    {
      "entropy": 5.2028357028961185,
      "epoch": 2.2281460134486073,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004504696572277165,
      "loss": 4.8774,
      "mean_token_accuracy": 0.2278245523571968,
      "num_tokens": 53205484.0,
      "step": 23195
    },
    {
      "entropy": 5.198713874816894,
      "epoch": 2.228626320845341,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045044819548266385,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22555887401103974,
      "num_tokens": 53217070.0,
      "step": 23200
    },
    {
      "entropy": 5.106890773773193,
      "epoch": 2.229106628242075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045042672966422506,
      "loss": 4.8228,
      "mean_token_accuracy": 0.22876110523939133,
      "num_tokens": 53228431.0,
      "step": 23205
    },
    {
      "entropy": 5.183371591567993,
      "epoch": 2.229586935638809,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045040525977289847,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21941139549016953,
      "num_tokens": 53240718.0,
      "step": 23210
    },
    {
      "entropy": 5.295283031463623,
      "epoch": 2.230067243035543,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004503837858091826,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22532128244638444,
      "num_tokens": 53251542.0,
      "step": 23215
    },
    {
      "entropy": 5.169456958770752,
      "epoch": 2.2305475504322767,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045036230777357604,
      "loss": 4.8308,
      "mean_token_accuracy": 0.2306185856461525,
      "num_tokens": 53261915.0,
      "step": 23220
    },
    {
      "entropy": 5.190240716934204,
      "epoch": 2.2310278578290106,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004503408256665774,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22584224492311478,
      "num_tokens": 53273287.0,
      "step": 23225
    },
    {
      "entropy": 5.189193248748779,
      "epoch": 2.2315081652257445,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045031933948868545,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22069835364818574,
      "num_tokens": 53284482.0,
      "step": 23230
    },
    {
      "entropy": 5.169965171813965,
      "epoch": 2.2319884726224783,
      "grad_norm": 1.0,
      "learning_rate": 0.00045029784924039903,
      "loss": 4.8127,
      "mean_token_accuracy": 0.2289542257785797,
      "num_tokens": 53296235.0,
      "step": 23235
    },
    {
      "entropy": 5.282018852233887,
      "epoch": 2.2324687800192122,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045027635492221716,
      "loss": 4.969,
      "mean_token_accuracy": 0.2114968866109848,
      "num_tokens": 53306784.0,
      "step": 23240
    },
    {
      "entropy": 5.181709146499633,
      "epoch": 2.232949087415946,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045025485653463866,
      "loss": 4.8705,
      "mean_token_accuracy": 0.2278911367058754,
      "num_tokens": 53318303.0,
      "step": 23245
    },
    {
      "entropy": 5.17683973312378,
      "epoch": 2.23342939481268,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004502333540781628,
      "loss": 4.8427,
      "mean_token_accuracy": 0.23332123309373856,
      "num_tokens": 53330196.0,
      "step": 23250
    },
    {
      "entropy": 5.1859557151794435,
      "epoch": 2.233909702209414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004502118475532888,
      "loss": 4.8561,
      "mean_token_accuracy": 0.22328293770551683,
      "num_tokens": 53342250.0,
      "step": 23255
    },
    {
      "entropy": 5.0922544479370115,
      "epoch": 2.2343900096061478,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000450190336960516,
      "loss": 4.8101,
      "mean_token_accuracy": 0.22355384826660157,
      "num_tokens": 53353050.0,
      "step": 23260
    },
    {
      "entropy": 5.129181385040283,
      "epoch": 2.2348703170028816,
      "grad_norm": 1.125,
      "learning_rate": 0.0004501688223003438,
      "loss": 4.8052,
      "mean_token_accuracy": 0.2344799295067787,
      "num_tokens": 53364702.0,
      "step": 23265
    },
    {
      "entropy": 5.16671233177185,
      "epoch": 2.235350624399616,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004501473035732717,
      "loss": 4.8226,
      "mean_token_accuracy": 0.22048887014389038,
      "num_tokens": 53375790.0,
      "step": 23270
    },
    {
      "entropy": 5.18066086769104,
      "epoch": 2.23583093179635,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004501257807797993,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22448172718286513,
      "num_tokens": 53387206.0,
      "step": 23275
    },
    {
      "entropy": 5.138333034515381,
      "epoch": 2.2363112391930837,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045010425392042624,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22484249323606492,
      "num_tokens": 53399519.0,
      "step": 23280
    },
    {
      "entropy": 5.160270261764526,
      "epoch": 2.2367915465898176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004500827229956524,
      "loss": 4.8094,
      "mean_token_accuracy": 0.227932171523571,
      "num_tokens": 53411187.0,
      "step": 23285
    },
    {
      "entropy": 5.226218271255493,
      "epoch": 2.2372718539865515,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045006118800597757,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2239303782582283,
      "num_tokens": 53421328.0,
      "step": 23290
    },
    {
      "entropy": 5.117600584030152,
      "epoch": 2.2377521613832854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045003964895190177,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22533937990665437,
      "num_tokens": 53433155.0,
      "step": 23295
    },
    {
      "entropy": 5.171433448791504,
      "epoch": 2.2382324687800192,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004500181058339251,
      "loss": 4.8676,
      "mean_token_accuracy": 0.2203913450241089,
      "num_tokens": 53445498.0,
      "step": 23300
    },
    {
      "entropy": 5.220692825317383,
      "epoch": 2.238712776176753,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004499965586525478,
      "loss": 4.8445,
      "mean_token_accuracy": 0.2300914093852043,
      "num_tokens": 53458005.0,
      "step": 23305
    },
    {
      "entropy": 5.176861572265625,
      "epoch": 2.239193083573487,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044997500740826993,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22914390563964843,
      "num_tokens": 53470052.0,
      "step": 23310
    },
    {
      "entropy": 5.1186638355255125,
      "epoch": 2.239673390970221,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000449953452101592,
      "loss": 4.8556,
      "mean_token_accuracy": 0.23240523487329484,
      "num_tokens": 53480665.0,
      "step": 23315
    },
    {
      "entropy": 5.167840909957886,
      "epoch": 2.2401536983669548,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044993189273301445,
      "loss": 4.8182,
      "mean_token_accuracy": 0.2299409955739975,
      "num_tokens": 53491903.0,
      "step": 23320
    },
    {
      "entropy": 5.154835367202759,
      "epoch": 2.2406340057636887,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004499103293030377,
      "loss": 4.8815,
      "mean_token_accuracy": 0.22766185998916627,
      "num_tokens": 53502806.0,
      "step": 23325
    },
    {
      "entropy": 5.153516340255737,
      "epoch": 2.2411143131604225,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004498887618121625,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23174736946821212,
      "num_tokens": 53513738.0,
      "step": 23330
    },
    {
      "entropy": 5.108708667755127,
      "epoch": 2.2415946205571564,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004498671902608897,
      "loss": 4.8458,
      "mean_token_accuracy": 0.22580721229314804,
      "num_tokens": 53525546.0,
      "step": 23335
    },
    {
      "entropy": 5.18180775642395,
      "epoch": 2.2420749279538903,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004498456146497198,
      "loss": 4.8299,
      "mean_token_accuracy": 0.22860410660505295,
      "num_tokens": 53537162.0,
      "step": 23340
    },
    {
      "entropy": 5.206007623672486,
      "epoch": 2.2425552353506246,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044982403497915405,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22737952172756196,
      "num_tokens": 53549165.0,
      "step": 23345
    },
    {
      "entropy": 5.211472415924073,
      "epoch": 2.2430355427473585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044980245124969333,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22777342647314072,
      "num_tokens": 53561641.0,
      "step": 23350
    },
    {
      "entropy": 5.1252336502075195,
      "epoch": 2.2435158501440924,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004497808634618386,
      "loss": 4.8507,
      "mean_token_accuracy": 0.23139424324035646,
      "num_tokens": 53572350.0,
      "step": 23355
    },
    {
      "entropy": 5.179977941513061,
      "epoch": 2.2439961575408263,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004497592716160913,
      "loss": 4.938,
      "mean_token_accuracy": 0.2143723949790001,
      "num_tokens": 53584191.0,
      "step": 23360
    },
    {
      "entropy": 5.289032745361328,
      "epoch": 2.24447646493756,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044973767571295273,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21851855218410493,
      "num_tokens": 53595021.0,
      "step": 23365
    },
    {
      "entropy": 5.268391561508179,
      "epoch": 2.244956772334294,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004497160757529241,
      "loss": 4.9178,
      "mean_token_accuracy": 0.22842417657375336,
      "num_tokens": 53606992.0,
      "step": 23370
    },
    {
      "entropy": 5.187631511688233,
      "epoch": 2.245437079731028,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044969447173650695,
      "loss": 4.8073,
      "mean_token_accuracy": 0.23904284238815307,
      "num_tokens": 53616548.0,
      "step": 23375
    },
    {
      "entropy": 5.167642211914062,
      "epoch": 2.245917387127762,
      "grad_norm": 1.125,
      "learning_rate": 0.000449672863664203,
      "loss": 4.8465,
      "mean_token_accuracy": 0.22922302335500716,
      "num_tokens": 53629078.0,
      "step": 23380
    },
    {
      "entropy": 5.1145933151245115,
      "epoch": 2.2463976945244957,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044965125153651375,
      "loss": 4.79,
      "mean_token_accuracy": 0.23830398172140121,
      "num_tokens": 53640059.0,
      "step": 23385
    },
    {
      "entropy": 5.205463790893555,
      "epoch": 2.2468780019212296,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004496296353539411,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21398036926984787,
      "num_tokens": 53651806.0,
      "step": 23390
    },
    {
      "entropy": 5.196995544433594,
      "epoch": 2.2473583093179634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004496080151169869,
      "loss": 4.9395,
      "mean_token_accuracy": 0.22113989144563675,
      "num_tokens": 53662180.0,
      "step": 23395
    },
    {
      "entropy": 5.159361553192139,
      "epoch": 2.2478386167146973,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044958639082615294,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22268953323364257,
      "num_tokens": 53674520.0,
      "step": 23400
    },
    {
      "entropy": 5.327778053283692,
      "epoch": 2.248318924111431,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004495647624819415,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2133573994040489,
      "num_tokens": 53686051.0,
      "step": 23405
    },
    {
      "entropy": 5.206648063659668,
      "epoch": 2.248799231508165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044954313008485457,
      "loss": 4.7385,
      "mean_token_accuracy": 0.24035776555538177,
      "num_tokens": 53697171.0,
      "step": 23410
    },
    {
      "entropy": 5.1649086475372314,
      "epoch": 2.249279538904899,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044952149363539453,
      "loss": 4.8835,
      "mean_token_accuracy": 0.22407249361276627,
      "num_tokens": 53707776.0,
      "step": 23415
    },
    {
      "entropy": 5.1878162860870365,
      "epoch": 2.249759846301633,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004494998531340635,
      "loss": 4.9207,
      "mean_token_accuracy": 0.2171652391552925,
      "num_tokens": 53719430.0,
      "step": 23420
    },
    {
      "entropy": 5.195713186264038,
      "epoch": 2.2502401536983667,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004494782085813641,
      "loss": 4.7724,
      "mean_token_accuracy": 0.23460965007543563,
      "num_tokens": 53730708.0,
      "step": 23425
    },
    {
      "entropy": 5.128457260131836,
      "epoch": 2.250720461095101,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004494565599777988,
      "loss": 4.8623,
      "mean_token_accuracy": 0.22770293802022934,
      "num_tokens": 53741623.0,
      "step": 23430
    },
    {
      "entropy": 5.115003252029419,
      "epoch": 2.251200768491835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044943490732387025,
      "loss": 4.8043,
      "mean_token_accuracy": 0.2282756209373474,
      "num_tokens": 53752348.0,
      "step": 23435
    },
    {
      "entropy": 5.180234718322754,
      "epoch": 2.251681075888569,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004494132506200811,
      "loss": 4.8527,
      "mean_token_accuracy": 0.22389007955789567,
      "num_tokens": 53763973.0,
      "step": 23440
    },
    {
      "entropy": 5.091251993179322,
      "epoch": 2.2521613832853027,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004493915898669341,
      "loss": 4.7159,
      "mean_token_accuracy": 0.23478296250104905,
      "num_tokens": 53775139.0,
      "step": 23445
    },
    {
      "entropy": 5.156144094467163,
      "epoch": 2.2526416906820366,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004493699250649323,
      "loss": 4.9151,
      "mean_token_accuracy": 0.23127202093601226,
      "num_tokens": 53788324.0,
      "step": 23450
    },
    {
      "entropy": 5.205100154876709,
      "epoch": 2.2531219980787704,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004493482562145786,
      "loss": 4.8078,
      "mean_token_accuracy": 0.229248109459877,
      "num_tokens": 53800019.0,
      "step": 23455
    },
    {
      "entropy": 5.219074010848999,
      "epoch": 2.2536023054755043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044932658331637605,
      "loss": 4.9381,
      "mean_token_accuracy": 0.21974605023860933,
      "num_tokens": 53812155.0,
      "step": 23460
    },
    {
      "entropy": 5.123564863204956,
      "epoch": 2.254082612872238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004493049063708279,
      "loss": 4.8417,
      "mean_token_accuracy": 0.23188960254192353,
      "num_tokens": 53823770.0,
      "step": 23465
    },
    {
      "entropy": 5.29152398109436,
      "epoch": 2.254562920268972,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044928322537843746,
      "loss": 5.0221,
      "mean_token_accuracy": 0.21831232458353042,
      "num_tokens": 53836037.0,
      "step": 23470
    },
    {
      "entropy": 5.1482994079589846,
      "epoch": 2.255043227665706,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044926154033970793,
      "loss": 4.7897,
      "mean_token_accuracy": 0.2300397664308548,
      "num_tokens": 53848517.0,
      "step": 23475
    },
    {
      "entropy": 5.138360452651978,
      "epoch": 2.25552353506244,
      "grad_norm": 1.34375,
      "learning_rate": 0.000449239851255143,
      "loss": 4.8203,
      "mean_token_accuracy": 0.23666814863681793,
      "num_tokens": 53860143.0,
      "step": 23480
    },
    {
      "entropy": 5.221700954437256,
      "epoch": 2.2560038424591737,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044921815812524606,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22522461265325547,
      "num_tokens": 53871173.0,
      "step": 23485
    },
    {
      "entropy": 5.154909896850586,
      "epoch": 2.2564841498559076,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044919646095052077,
      "loss": 4.824,
      "mean_token_accuracy": 0.23161608278751372,
      "num_tokens": 53882791.0,
      "step": 23490
    },
    {
      "entropy": 5.151592874526978,
      "epoch": 2.2569644572526415,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000449174759731471,
      "loss": 4.8909,
      "mean_token_accuracy": 0.2270645633339882,
      "num_tokens": 53894785.0,
      "step": 23495
    },
    {
      "entropy": 5.205832052230835,
      "epoch": 2.2574447646493754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044915305446860046,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22653023898601532,
      "num_tokens": 53905989.0,
      "step": 23500
    },
    {
      "entropy": 5.162853479385376,
      "epoch": 2.2579250720461097,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044913134516241305,
      "loss": 4.7842,
      "mean_token_accuracy": 0.2272715985774994,
      "num_tokens": 53918233.0,
      "step": 23505
    },
    {
      "entropy": 5.113063001632691,
      "epoch": 2.2584053794428436,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004491096318134129,
      "loss": 4.8657,
      "mean_token_accuracy": 0.22407424151897432,
      "num_tokens": 53929886.0,
      "step": 23510
    },
    {
      "entropy": 5.271719121932984,
      "epoch": 2.2588856868395775,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004490879144221042,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20740476995706558,
      "num_tokens": 53941788.0,
      "step": 23515
    },
    {
      "entropy": 5.208265972137451,
      "epoch": 2.2593659942363113,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044906619298899097,
      "loss": 4.8264,
      "mean_token_accuracy": 0.23068186193704604,
      "num_tokens": 53952753.0,
      "step": 23520
    },
    {
      "entropy": 5.132901048660278,
      "epoch": 2.2598463016330452,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004490444675145777,
      "loss": 4.8551,
      "mean_token_accuracy": 0.2307449012994766,
      "num_tokens": 53964435.0,
      "step": 23525
    },
    {
      "entropy": 5.199666738510132,
      "epoch": 2.260326609029779,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004490227379993686,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22592173218727113,
      "num_tokens": 53976906.0,
      "step": 23530
    },
    {
      "entropy": 5.206274080276489,
      "epoch": 2.260806916426513,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004490010044438683,
      "loss": 4.8254,
      "mean_token_accuracy": 0.23390084654092788,
      "num_tokens": 53988076.0,
      "step": 23535
    },
    {
      "entropy": 5.048392391204834,
      "epoch": 2.261287223823247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044897926684858133,
      "loss": 4.7465,
      "mean_token_accuracy": 0.234316685795784,
      "num_tokens": 53998896.0,
      "step": 23540
    },
    {
      "entropy": 5.0900774002075195,
      "epoch": 2.2617675312199808,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044895752521401246,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23143697232007981,
      "num_tokens": 54010855.0,
      "step": 23545
    },
    {
      "entropy": 5.210872030258178,
      "epoch": 2.2622478386167146,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004489357795406663,
      "loss": 4.8686,
      "mean_token_accuracy": 0.23042750656604766,
      "num_tokens": 54021630.0,
      "step": 23550
    },
    {
      "entropy": 5.1503386974334715,
      "epoch": 2.2627281460134485,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004489140298290479,
      "loss": 4.8006,
      "mean_token_accuracy": 0.22928445786237717,
      "num_tokens": 54033715.0,
      "step": 23555
    },
    {
      "entropy": 5.132248258590698,
      "epoch": 2.2632084534101824,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044889227607966217,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22405778765678405,
      "num_tokens": 54046633.0,
      "step": 23560
    },
    {
      "entropy": 5.187770175933838,
      "epoch": 2.2636887608069163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044887051829301406,
      "loss": 4.8637,
      "mean_token_accuracy": 0.22688241600990294,
      "num_tokens": 54058389.0,
      "step": 23565
    },
    {
      "entropy": 5.215956401824951,
      "epoch": 2.26416906820365,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044884875646960886,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22362007200717926,
      "num_tokens": 54070338.0,
      "step": 23570
    },
    {
      "entropy": 5.136712503433228,
      "epoch": 2.264649375600384,
      "grad_norm": 1.125,
      "learning_rate": 0.00044882699060995175,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22938944399356842,
      "num_tokens": 54082511.0,
      "step": 23575
    },
    {
      "entropy": 5.150363254547119,
      "epoch": 2.2651296829971184,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004488052207145481,
      "loss": 4.8299,
      "mean_token_accuracy": 0.2277448445558548,
      "num_tokens": 54094141.0,
      "step": 23580
    },
    {
      "entropy": 5.157951021194458,
      "epoch": 2.2656099903938522,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044878344678390324,
      "loss": 4.9379,
      "mean_token_accuracy": 0.22273894250392914,
      "num_tokens": 54104837.0,
      "step": 23585
    },
    {
      "entropy": 5.251724481582642,
      "epoch": 2.266090297790586,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044876166881852286,
      "loss": 4.9094,
      "mean_token_accuracy": 0.21741730123758315,
      "num_tokens": 54116924.0,
      "step": 23590
    },
    {
      "entropy": 5.19210467338562,
      "epoch": 2.26657060518732,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004487398868189125,
      "loss": 4.8464,
      "mean_token_accuracy": 0.2306140884757042,
      "num_tokens": 54127187.0,
      "step": 23595
    },
    {
      "entropy": 5.147652912139892,
      "epoch": 2.267050912584054,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044871810078557777,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23084039539098739,
      "num_tokens": 54139024.0,
      "step": 23600
    },
    {
      "entropy": 5.176128244400024,
      "epoch": 2.2675312199807878,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004486963107190247,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22294869720935823,
      "num_tokens": 54150643.0,
      "step": 23605
    },
    {
      "entropy": 5.232387590408325,
      "epoch": 2.2680115273775217,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044867451661975894,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22561298310756683,
      "num_tokens": 54162171.0,
      "step": 23610
    },
    {
      "entropy": 5.169717931747437,
      "epoch": 2.2684918347742555,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044865271848828673,
      "loss": 4.864,
      "mean_token_accuracy": 0.2257572591304779,
      "num_tokens": 54173104.0,
      "step": 23615
    },
    {
      "entropy": 5.146307802200317,
      "epoch": 2.2689721421709894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000448630916325114,
      "loss": 4.8528,
      "mean_token_accuracy": 0.22701227515935898,
      "num_tokens": 54183956.0,
      "step": 23620
    },
    {
      "entropy": 5.213671398162842,
      "epoch": 2.2694524495677233,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000448609110130747,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22616536170244217,
      "num_tokens": 54195087.0,
      "step": 23625
    },
    {
      "entropy": 5.136537647247314,
      "epoch": 2.269932756964457,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044858729990569193,
      "loss": 4.8008,
      "mean_token_accuracy": 0.23043912202119826,
      "num_tokens": 54206684.0,
      "step": 23630
    },
    {
      "entropy": 5.115725135803222,
      "epoch": 2.270413064361191,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044856548565045523,
      "loss": 4.8287,
      "mean_token_accuracy": 0.2271506279706955,
      "num_tokens": 54218155.0,
      "step": 23635
    },
    {
      "entropy": 5.203822469711303,
      "epoch": 2.270893371757925,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044854366736554323,
      "loss": 4.8986,
      "mean_token_accuracy": 0.22781532853841782,
      "num_tokens": 54229124.0,
      "step": 23640
    },
    {
      "entropy": 5.143389892578125,
      "epoch": 2.271373679154659,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044852184505146274,
      "loss": 4.8367,
      "mean_token_accuracy": 0.23365228474140168,
      "num_tokens": 54240889.0,
      "step": 23645
    },
    {
      "entropy": 5.128871202468872,
      "epoch": 2.2718539865513927,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004485000187087202,
      "loss": 4.8031,
      "mean_token_accuracy": 0.2366361916065216,
      "num_tokens": 54252388.0,
      "step": 23650
    },
    {
      "entropy": 5.173939514160156,
      "epoch": 2.272334293948127,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004484781883378224,
      "loss": 4.8502,
      "mean_token_accuracy": 0.22676964700222016,
      "num_tokens": 54264702.0,
      "step": 23655
    },
    {
      "entropy": 5.207448148727417,
      "epoch": 2.2728146013448605,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044845635393927623,
      "loss": 4.9297,
      "mean_token_accuracy": 0.22353605180978775,
      "num_tokens": 54277064.0,
      "step": 23660
    },
    {
      "entropy": 5.122329378128052,
      "epoch": 2.273294908741595,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004484345155135886,
      "loss": 4.8053,
      "mean_token_accuracy": 0.23280777931213378,
      "num_tokens": 54287958.0,
      "step": 23665
    },
    {
      "entropy": 5.113739442825318,
      "epoch": 2.2737752161383287,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004484126730612664,
      "loss": 4.8391,
      "mean_token_accuracy": 0.22920786291360856,
      "num_tokens": 54299428.0,
      "step": 23670
    },
    {
      "entropy": 5.205555438995361,
      "epoch": 2.2742555235350626,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000448390826582817,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22190964370965957,
      "num_tokens": 54312249.0,
      "step": 23675
    },
    {
      "entropy": 5.2560567378997805,
      "epoch": 2.2747358309317964,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044836897607874744,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21853582113981246,
      "num_tokens": 54324408.0,
      "step": 23680
    },
    {
      "entropy": 5.192885494232177,
      "epoch": 2.2752161383285303,
      "grad_norm": 1.03125,
      "learning_rate": 0.000448347121549565,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22656177431344987,
      "num_tokens": 54336092.0,
      "step": 23685
    },
    {
      "entropy": 5.107464361190796,
      "epoch": 2.275696445725264,
      "grad_norm": 1.0,
      "learning_rate": 0.0004483252629957771,
      "loss": 4.771,
      "mean_token_accuracy": 0.23931982964277268,
      "num_tokens": 54347231.0,
      "step": 23690
    },
    {
      "entropy": 5.161984920501709,
      "epoch": 2.276176753121998,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044830340041789133,
      "loss": 4.8356,
      "mean_token_accuracy": 0.2328764796257019,
      "num_tokens": 54359096.0,
      "step": 23695
    },
    {
      "entropy": 5.123661088943481,
      "epoch": 2.276657060518732,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004482815338164152,
      "loss": 4.7519,
      "mean_token_accuracy": 0.2374079316854477,
      "num_tokens": 54370433.0,
      "step": 23700
    },
    {
      "entropy": 5.171488428115845,
      "epoch": 2.277137367915466,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004482596631918564,
      "loss": 4.8732,
      "mean_token_accuracy": 0.21768954545259475,
      "num_tokens": 54381430.0,
      "step": 23705
    },
    {
      "entropy": 5.151374626159668,
      "epoch": 2.2776176753121997,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044823778854472267,
      "loss": 4.8449,
      "mean_token_accuracy": 0.22217780500650405,
      "num_tokens": 54392840.0,
      "step": 23710
    },
    {
      "entropy": 5.269924354553223,
      "epoch": 2.2780979827089336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004482159098755219,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2190593495965004,
      "num_tokens": 54405776.0,
      "step": 23715
    },
    {
      "entropy": 5.126699924468994,
      "epoch": 2.2785782901056675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004481940271847621,
      "loss": 4.8014,
      "mean_token_accuracy": 0.23142738044261932,
      "num_tokens": 54417392.0,
      "step": 23720
    },
    {
      "entropy": 5.185228538513184,
      "epoch": 2.2790585975024014,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004481721404729513,
      "loss": 4.9718,
      "mean_token_accuracy": 0.21291757375001907,
      "num_tokens": 54428407.0,
      "step": 23725
    },
    {
      "entropy": 5.245624637603759,
      "epoch": 2.2795389048991357,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004481502497405975,
      "loss": 4.8792,
      "mean_token_accuracy": 0.2271825611591339,
      "num_tokens": 54439667.0,
      "step": 23730
    },
    {
      "entropy": 5.126108312606812,
      "epoch": 2.280019212295869,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004481283549882091,
      "loss": 4.8146,
      "mean_token_accuracy": 0.22970257550477982,
      "num_tokens": 54451966.0,
      "step": 23735
    },
    {
      "entropy": 5.247702884674072,
      "epoch": 2.2804995196926034,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044810645621629443,
      "loss": 4.9974,
      "mean_token_accuracy": 0.2148939073085785,
      "num_tokens": 54462896.0,
      "step": 23740
    },
    {
      "entropy": 5.188837242126465,
      "epoch": 2.2809798270893373,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044808455342536176,
      "loss": 4.9043,
      "mean_token_accuracy": 0.23161743581295013,
      "num_tokens": 54474056.0,
      "step": 23745
    },
    {
      "entropy": 5.18371353149414,
      "epoch": 2.281460134486071,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044806264661591976,
      "loss": 4.8742,
      "mean_token_accuracy": 0.2211504802107811,
      "num_tokens": 54485846.0,
      "step": 23750
    },
    {
      "entropy": 5.156873559951782,
      "epoch": 2.281940441882805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004480407357884771,
      "loss": 4.8117,
      "mean_token_accuracy": 0.22742047756910325,
      "num_tokens": 54496020.0,
      "step": 23755
    },
    {
      "entropy": 5.23224024772644,
      "epoch": 2.282420749279539,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044801882094354226,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2141735315322876,
      "num_tokens": 54507388.0,
      "step": 23760
    },
    {
      "entropy": 5.211707067489624,
      "epoch": 2.282901056676273,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004479969020816242,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2313588410615921,
      "num_tokens": 54519159.0,
      "step": 23765
    },
    {
      "entropy": 5.217219686508178,
      "epoch": 2.2833813640730067,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044797497920323175,
      "loss": 5.024,
      "mean_token_accuracy": 0.22224070876836777,
      "num_tokens": 54530258.0,
      "step": 23770
    },
    {
      "entropy": 5.2389076232910154,
      "epoch": 2.2838616714697406,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004479530523088739,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2157914638519287,
      "num_tokens": 54542626.0,
      "step": 23775
    },
    {
      "entropy": 5.193422126770019,
      "epoch": 2.2843419788664745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004479311213990598,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22455762922763825,
      "num_tokens": 54553762.0,
      "step": 23780
    },
    {
      "entropy": 5.175043106079102,
      "epoch": 2.2848222862632084,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044790918647429854,
      "loss": 4.8049,
      "mean_token_accuracy": 0.22902195006608964,
      "num_tokens": 54565484.0,
      "step": 23785
    },
    {
      "entropy": 5.129402112960816,
      "epoch": 2.2853025936599423,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044788724753509935,
      "loss": 4.8376,
      "mean_token_accuracy": 0.23449129313230516,
      "num_tokens": 54577321.0,
      "step": 23790
    },
    {
      "entropy": 5.111122274398804,
      "epoch": 2.285782901056676,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004478653045819717,
      "loss": 4.8198,
      "mean_token_accuracy": 0.2362240105867386,
      "num_tokens": 54587796.0,
      "step": 23795
    },
    {
      "entropy": 5.223103713989258,
      "epoch": 2.28626320845341,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004478433576154249,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2198309525847435,
      "num_tokens": 54600637.0,
      "step": 23800
    },
    {
      "entropy": 5.251637744903564,
      "epoch": 2.286743515850144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004478214066359687,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22575539201498032,
      "num_tokens": 54610690.0,
      "step": 23805
    },
    {
      "entropy": 5.214708232879639,
      "epoch": 2.287223823246878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044779945164411254,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22056688815355302,
      "num_tokens": 54622417.0,
      "step": 23810
    },
    {
      "entropy": 5.1703653812408445,
      "epoch": 2.287704130643612,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004477774926403662,
      "loss": 4.8162,
      "mean_token_accuracy": 0.2343436896800995,
      "num_tokens": 54633190.0,
      "step": 23815
    },
    {
      "entropy": 5.19298620223999,
      "epoch": 2.288184438040346,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004477555296252396,
      "loss": 4.84,
      "mean_token_accuracy": 0.22868522852659226,
      "num_tokens": 54643507.0,
      "step": 23820
    },
    {
      "entropy": 5.0273322582244875,
      "epoch": 2.28866474543708,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044773356259924255,
      "loss": 4.7554,
      "mean_token_accuracy": 0.23708846420049667,
      "num_tokens": 54653835.0,
      "step": 23825
    },
    {
      "entropy": 5.206317329406739,
      "epoch": 2.2891450528338138,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044771159156288505,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22870426774024963,
      "num_tokens": 54664908.0,
      "step": 23830
    },
    {
      "entropy": 5.156438493728638,
      "epoch": 2.2896253602305476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004476896165166772,
      "loss": 4.8548,
      "mean_token_accuracy": 0.2282841056585312,
      "num_tokens": 54677714.0,
      "step": 23835
    },
    {
      "entropy": 5.167556810379028,
      "epoch": 2.2901056676272815,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044766763746112936,
      "loss": 4.8696,
      "mean_token_accuracy": 0.2245771586894989,
      "num_tokens": 54688919.0,
      "step": 23840
    },
    {
      "entropy": 5.158967638015747,
      "epoch": 2.2905859750240154,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004476456543967517,
      "loss": 4.877,
      "mean_token_accuracy": 0.22782885879278184,
      "num_tokens": 54700089.0,
      "step": 23845
    },
    {
      "entropy": 5.191954946517944,
      "epoch": 2.2910662824207493,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044762366732405454,
      "loss": 4.8159,
      "mean_token_accuracy": 0.23828311860561371,
      "num_tokens": 54711233.0,
      "step": 23850
    },
    {
      "entropy": 5.194373226165771,
      "epoch": 2.291546589817483,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004476016762435484,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2213611587882042,
      "num_tokens": 54723140.0,
      "step": 23855
    },
    {
      "entropy": 5.202805376052856,
      "epoch": 2.292026897214217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004475796811557439,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22901684939861297,
      "num_tokens": 54734196.0,
      "step": 23860
    },
    {
      "entropy": 5.230012273788452,
      "epoch": 2.292507204610951,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044755768206115155,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21177158653736114,
      "num_tokens": 54747395.0,
      "step": 23865
    },
    {
      "entropy": 5.214230585098266,
      "epoch": 2.292987512007685,
      "grad_norm": 1.125,
      "learning_rate": 0.0004475356789602823,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22546781450510026,
      "num_tokens": 54761137.0,
      "step": 23870
    },
    {
      "entropy": 5.161330318450927,
      "epoch": 2.2934678194044187,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044751367185364696,
      "loss": 4.7983,
      "mean_token_accuracy": 0.22843362838029863,
      "num_tokens": 54772181.0,
      "step": 23875
    },
    {
      "entropy": 5.1077433109283445,
      "epoch": 2.2939481268011526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044749166074175634,
      "loss": 4.8434,
      "mean_token_accuracy": 0.23038306534290315,
      "num_tokens": 54783759.0,
      "step": 23880
    },
    {
      "entropy": 5.146162796020508,
      "epoch": 2.2944284341978864,
      "grad_norm": 1.25,
      "learning_rate": 0.00044746964562512154,
      "loss": 4.7717,
      "mean_token_accuracy": 0.22758061736822127,
      "num_tokens": 54796265.0,
      "step": 23885
    },
    {
      "entropy": 5.110705614089966,
      "epoch": 2.2949087415946208,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044744762650425376,
      "loss": 4.8649,
      "mean_token_accuracy": 0.227969129383564,
      "num_tokens": 54808159.0,
      "step": 23890
    },
    {
      "entropy": 5.244549131393432,
      "epoch": 2.2953890489913547,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044742560337966415,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22455647736787795,
      "num_tokens": 54819070.0,
      "step": 23895
    },
    {
      "entropy": 5.1161716938018795,
      "epoch": 2.2958693563880885,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000447403576251864,
      "loss": 4.8464,
      "mean_token_accuracy": 0.22402366250753403,
      "num_tokens": 54829210.0,
      "step": 23900
    },
    {
      "entropy": 5.128628540039062,
      "epoch": 2.2963496637848224,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004473815451213648,
      "loss": 4.8192,
      "mean_token_accuracy": 0.23280172795057297,
      "num_tokens": 54841238.0,
      "step": 23905
    },
    {
      "entropy": 5.214447689056397,
      "epoch": 2.2968299711815563,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004473595099886779,
      "loss": 4.8821,
      "mean_token_accuracy": 0.23338940888643264,
      "num_tokens": 54852531.0,
      "step": 23910
    },
    {
      "entropy": 5.17368369102478,
      "epoch": 2.29731027857829,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004473374708543151,
      "loss": 4.851,
      "mean_token_accuracy": 0.22883552461862564,
      "num_tokens": 54864089.0,
      "step": 23915
    },
    {
      "entropy": 5.137069988250732,
      "epoch": 2.297790585975024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004473154277187878,
      "loss": 4.8222,
      "mean_token_accuracy": 0.22690456062555314,
      "num_tokens": 54874956.0,
      "step": 23920
    },
    {
      "entropy": 5.215553617477417,
      "epoch": 2.298270893371758,
      "grad_norm": 1.125,
      "learning_rate": 0.00044729338058260805,
      "loss": 4.8632,
      "mean_token_accuracy": 0.2255019575357437,
      "num_tokens": 54885608.0,
      "step": 23925
    },
    {
      "entropy": 5.1210661888122555,
      "epoch": 2.298751200768492,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004472713294462876,
      "loss": 4.8157,
      "mean_token_accuracy": 0.23032934963703156,
      "num_tokens": 54896782.0,
      "step": 23930
    },
    {
      "entropy": 5.1588414192199705,
      "epoch": 2.2992315081652257,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044724927431033843,
      "loss": 4.8863,
      "mean_token_accuracy": 0.22166333794593812,
      "num_tokens": 54908179.0,
      "step": 23935
    },
    {
      "entropy": 5.196999502182007,
      "epoch": 2.2997118155619596,
      "grad_norm": 1.375,
      "learning_rate": 0.0004472272151752725,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22370546460151672,
      "num_tokens": 54921114.0,
      "step": 23940
    },
    {
      "entropy": 5.218417835235596,
      "epoch": 2.3001921229586935,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004472051520416022,
      "loss": 4.8393,
      "mean_token_accuracy": 0.22720366418361665,
      "num_tokens": 54931398.0,
      "step": 23945
    },
    {
      "entropy": 5.254206562042237,
      "epoch": 2.3006724303554273,
      "grad_norm": 1.125,
      "learning_rate": 0.0004471830849098395,
      "loss": 4.8995,
      "mean_token_accuracy": 0.224046228826046,
      "num_tokens": 54942354.0,
      "step": 23950
    },
    {
      "entropy": 5.127012538909912,
      "epoch": 2.3011527377521612,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044716101378049683,
      "loss": 4.8283,
      "mean_token_accuracy": 0.23748761862516404,
      "num_tokens": 54953700.0,
      "step": 23955
    },
    {
      "entropy": 5.010592699050903,
      "epoch": 2.301633045148895,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044713893865408667,
      "loss": 4.7841,
      "mean_token_accuracy": 0.23399144113063813,
      "num_tokens": 54965544.0,
      "step": 23960
    },
    {
      "entropy": 5.143846750259399,
      "epoch": 2.3021133525456294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004471168595311215,
      "loss": 4.8016,
      "mean_token_accuracy": 0.23451133370399474,
      "num_tokens": 54976811.0,
      "step": 23965
    },
    {
      "entropy": 5.16963267326355,
      "epoch": 2.302593659942363,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044709477641211395,
      "loss": 4.8142,
      "mean_token_accuracy": 0.23631610870361328,
      "num_tokens": 54987508.0,
      "step": 23970
    },
    {
      "entropy": 5.228953075408936,
      "epoch": 2.303073967339097,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004470726892975767,
      "loss": 5.0213,
      "mean_token_accuracy": 0.2115531012415886,
      "num_tokens": 55000276.0,
      "step": 23975
    },
    {
      "entropy": 5.194169807434082,
      "epoch": 2.303554274735831,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044705059818802255,
      "loss": 4.8566,
      "mean_token_accuracy": 0.23450987040996552,
      "num_tokens": 55012397.0,
      "step": 23980
    },
    {
      "entropy": 5.158816814422607,
      "epoch": 2.304034582132565,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004470285030839644,
      "loss": 4.7986,
      "mean_token_accuracy": 0.22937736362218858,
      "num_tokens": 55024358.0,
      "step": 23985
    },
    {
      "entropy": 5.09406270980835,
      "epoch": 2.304514889529299,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044700640398591526,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23102694898843765,
      "num_tokens": 55035887.0,
      "step": 23990
    },
    {
      "entropy": 5.204601049423218,
      "epoch": 2.3049951969260327,
      "grad_norm": 1.125,
      "learning_rate": 0.0004469843008943881,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21698899269104005,
      "num_tokens": 55048775.0,
      "step": 23995
    },
    {
      "entropy": 5.247527313232422,
      "epoch": 2.3054755043227666,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004469621938098962,
      "loss": 4.8819,
      "mean_token_accuracy": 0.23007541596889497,
      "num_tokens": 55059994.0,
      "step": 24000
    },
    {
      "epoch": 2.3054755043227666,
      "eval_entropy": 5.0264807440256964,
      "eval_loss": 4.978430271148682,
      "eval_mean_token_accuracy": 0.22936128535789724,
      "eval_num_tokens": 55059994.0,
      "eval_runtime": 26.4647,
      "eval_samples_per_second": 1239.952,
      "eval_steps_per_second": 154.999,
      "step": 24000
    },
    {
      "entropy": 5.208881282806397,
      "epoch": 2.3059558117195005,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004469400827329528,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22136543542146683,
      "num_tokens": 55071950.0,
      "step": 24005
    },
    {
      "entropy": 5.054619169235229,
      "epoch": 2.3064361191162344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004469179676640712,
      "loss": 4.7534,
      "mean_token_accuracy": 0.2388426184654236,
      "num_tokens": 55085260.0,
      "step": 24010
    },
    {
      "entropy": 5.18996033668518,
      "epoch": 2.3069164265129682,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004468958486037649,
      "loss": 4.7787,
      "mean_token_accuracy": 0.22916346341371535,
      "num_tokens": 55095794.0,
      "step": 24015
    },
    {
      "entropy": 5.147738456726074,
      "epoch": 2.307396733909702,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004468737255525474,
      "loss": 4.8548,
      "mean_token_accuracy": 0.2300407648086548,
      "num_tokens": 55106476.0,
      "step": 24020
    },
    {
      "entropy": 5.156685018539429,
      "epoch": 2.307877041306436,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004468515985109324,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22908006310462953,
      "num_tokens": 55118277.0,
      "step": 24025
    },
    {
      "entropy": 5.178712320327759,
      "epoch": 2.30835734870317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004468294674794335,
      "loss": 4.8684,
      "mean_token_accuracy": 0.2241617575287819,
      "num_tokens": 55129902.0,
      "step": 24030
    },
    {
      "entropy": 5.092375707626343,
      "epoch": 2.3088376560999038,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004468073324585647,
      "loss": 4.8259,
      "mean_token_accuracy": 0.2244735151529312,
      "num_tokens": 55143342.0,
      "step": 24035
    },
    {
      "entropy": 5.2473749160766605,
      "epoch": 2.309317963496638,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004467851934488398,
      "loss": 4.9085,
      "mean_token_accuracy": 0.23109859377145767,
      "num_tokens": 55154116.0,
      "step": 24040
    },
    {
      "entropy": 5.187033224105835,
      "epoch": 2.3097982708933715,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004467630504507727,
      "loss": 4.8523,
      "mean_token_accuracy": 0.2286726102232933,
      "num_tokens": 55165548.0,
      "step": 24045
    },
    {
      "entropy": 5.193554878234863,
      "epoch": 2.310278578290106,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004467409034648776,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22234465628862382,
      "num_tokens": 55176047.0,
      "step": 24050
    },
    {
      "entropy": 5.143213367462158,
      "epoch": 2.3107588856868397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004467187524916688,
      "loss": 4.8761,
      "mean_token_accuracy": 0.22765274047851564,
      "num_tokens": 55188715.0,
      "step": 24055
    },
    {
      "entropy": 5.049026823043823,
      "epoch": 2.3112391930835736,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004466965975316604,
      "loss": 4.7304,
      "mean_token_accuracy": 0.24033090323209763,
      "num_tokens": 55198988.0,
      "step": 24060
    },
    {
      "entropy": 5.088257265090943,
      "epoch": 2.3117195004803075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044667443858536685,
      "loss": 4.7143,
      "mean_token_accuracy": 0.24029948860406875,
      "num_tokens": 55210000.0,
      "step": 24065
    },
    {
      "entropy": 5.180831813812256,
      "epoch": 2.3121998078770414,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004466522756533026,
      "loss": 4.864,
      "mean_token_accuracy": 0.22100847959518433,
      "num_tokens": 55221794.0,
      "step": 24070
    },
    {
      "entropy": 5.228139877319336,
      "epoch": 2.3126801152737753,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004466301087359822,
      "loss": 4.9146,
      "mean_token_accuracy": 0.21756412088871002,
      "num_tokens": 55233166.0,
      "step": 24075
    },
    {
      "entropy": 5.112611722946167,
      "epoch": 2.313160422670509,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044660793783392035,
      "loss": 4.7964,
      "mean_token_accuracy": 0.23226374089717866,
      "num_tokens": 55244179.0,
      "step": 24080
    },
    {
      "entropy": 5.135675144195557,
      "epoch": 2.313640730067243,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004465857629476317,
      "loss": 4.8239,
      "mean_token_accuracy": 0.22816379815340043,
      "num_tokens": 55255556.0,
      "step": 24085
    },
    {
      "entropy": 5.286248350143433,
      "epoch": 2.314121037463977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004465635840776312,
      "loss": 4.984,
      "mean_token_accuracy": 0.22022192031145096,
      "num_tokens": 55267001.0,
      "step": 24090
    },
    {
      "entropy": 5.239016008377075,
      "epoch": 2.314601344860711,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044654140122443373,
      "loss": 4.8843,
      "mean_token_accuracy": 0.22621733248233794,
      "num_tokens": 55278253.0,
      "step": 24095
    },
    {
      "entropy": 5.161130666732788,
      "epoch": 2.3150816522574447,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004465192143885542,
      "loss": 4.8441,
      "mean_token_accuracy": 0.22874267846345903,
      "num_tokens": 55289601.0,
      "step": 24100
    },
    {
      "entropy": 5.132434272766114,
      "epoch": 2.3155619596541785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044649702357050787,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22044461965560913,
      "num_tokens": 55300661.0,
      "step": 24105
    },
    {
      "entropy": 5.142382192611694,
      "epoch": 2.3160422670509124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004464748287708099,
      "loss": 4.8403,
      "mean_token_accuracy": 0.23081352263689042,
      "num_tokens": 55311890.0,
      "step": 24110
    },
    {
      "entropy": 5.094909286499023,
      "epoch": 2.3165225744476463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044645262998997557,
      "loss": 4.7862,
      "mean_token_accuracy": 0.23823101371526717,
      "num_tokens": 55322241.0,
      "step": 24115
    },
    {
      "entropy": 5.160297441482544,
      "epoch": 2.31700288184438,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044643042722852024,
      "loss": 4.7703,
      "mean_token_accuracy": 0.23754269182682036,
      "num_tokens": 55334064.0,
      "step": 24120
    },
    {
      "entropy": 5.190683364868164,
      "epoch": 2.3174831892411145,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004464082204869595,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22489411979913712,
      "num_tokens": 55346266.0,
      "step": 24125
    },
    {
      "entropy": 5.115088272094726,
      "epoch": 2.3179634966378484,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004463860097658088,
      "loss": 4.8055,
      "mean_token_accuracy": 0.23235896974802017,
      "num_tokens": 55357395.0,
      "step": 24130
    },
    {
      "entropy": 5.133263492584229,
      "epoch": 2.3184438040345823,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004463637950655839,
      "loss": 4.8123,
      "mean_token_accuracy": 0.22790854424238205,
      "num_tokens": 55367982.0,
      "step": 24135
    },
    {
      "entropy": 5.18132586479187,
      "epoch": 2.318924111431316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044634157638680054,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22918518781661987,
      "num_tokens": 55379822.0,
      "step": 24140
    },
    {
      "entropy": 5.185386800765992,
      "epoch": 2.31940441882805,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044631935372997455,
      "loss": 4.8517,
      "mean_token_accuracy": 0.22267881482839585,
      "num_tokens": 55391271.0,
      "step": 24145
    },
    {
      "entropy": 5.1493151664733885,
      "epoch": 2.319884726224784,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004462971270956219,
      "loss": 4.7713,
      "mean_token_accuracy": 0.2302739664912224,
      "num_tokens": 55401514.0,
      "step": 24150
    },
    {
      "entropy": 5.1712487697601315,
      "epoch": 2.320365033621518,
      "grad_norm": 1.125,
      "learning_rate": 0.0004462748964842586,
      "loss": 4.8936,
      "mean_token_accuracy": 0.2291257843375206,
      "num_tokens": 55412979.0,
      "step": 24155
    },
    {
      "entropy": 5.122518253326416,
      "epoch": 2.3208453410182517,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004462526618964008,
      "loss": 4.8145,
      "mean_token_accuracy": 0.22606973946094513,
      "num_tokens": 55423985.0,
      "step": 24160
    },
    {
      "entropy": 5.13782320022583,
      "epoch": 2.3213256484149856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004462304233325647,
      "loss": 4.8454,
      "mean_token_accuracy": 0.22875836491584778,
      "num_tokens": 55434638.0,
      "step": 24165
    },
    {
      "entropy": 5.1494279384613035,
      "epoch": 2.3218059558117194,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004462081807932666,
      "loss": 4.8978,
      "mean_token_accuracy": 0.21485513150691987,
      "num_tokens": 55445402.0,
      "step": 24170
    },
    {
      "entropy": 5.203578186035156,
      "epoch": 2.3222862632084533,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004461859342790229,
      "loss": 4.8196,
      "mean_token_accuracy": 0.2319582536816597,
      "num_tokens": 55456707.0,
      "step": 24175
    },
    {
      "entropy": 5.1227963924407955,
      "epoch": 2.322766570605187,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004461636837903501,
      "loss": 4.8006,
      "mean_token_accuracy": 0.22644427567720413,
      "num_tokens": 55468473.0,
      "step": 24180
    },
    {
      "entropy": 5.057804775238037,
      "epoch": 2.323246878001921,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004461414293277649,
      "loss": 4.7109,
      "mean_token_accuracy": 0.23536432087421416,
      "num_tokens": 55479623.0,
      "step": 24185
    },
    {
      "entropy": 5.199443054199219,
      "epoch": 2.323727185398655,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004461191708917839,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2325580045580864,
      "num_tokens": 55490036.0,
      "step": 24190
    },
    {
      "entropy": 5.137261819839478,
      "epoch": 2.324207492795389,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044609690848292376,
      "loss": 4.877,
      "mean_token_accuracy": 0.2290624350309372,
      "num_tokens": 55501605.0,
      "step": 24195
    },
    {
      "entropy": 5.1834478855133055,
      "epoch": 2.324687800192123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004460746421017016,
      "loss": 4.7771,
      "mean_token_accuracy": 0.23958743661642073,
      "num_tokens": 55512029.0,
      "step": 24200
    },
    {
      "entropy": 5.208399391174316,
      "epoch": 2.325168107588857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044605237174863405,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22820637822151185,
      "num_tokens": 55524260.0,
      "step": 24205
    },
    {
      "entropy": 5.0786158561706545,
      "epoch": 2.325648414985591,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004460300974242386,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22656827569007873,
      "num_tokens": 55536520.0,
      "step": 24210
    },
    {
      "entropy": 5.194007730484008,
      "epoch": 2.326128722382325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004460078191290319,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21970550566911698,
      "num_tokens": 55547896.0,
      "step": 24215
    },
    {
      "entropy": 5.236271905899048,
      "epoch": 2.3266090297790587,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044598553686353153,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22853832989931105,
      "num_tokens": 55559229.0,
      "step": 24220
    },
    {
      "entropy": 5.180901908874512,
      "epoch": 2.3270893371757926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044596325062825476,
      "loss": 4.8481,
      "mean_token_accuracy": 0.2260160267353058,
      "num_tokens": 55569531.0,
      "step": 24225
    },
    {
      "entropy": 5.1274223804473875,
      "epoch": 2.3275696445725265,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004459409604237189,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2251845583319664,
      "num_tokens": 55580454.0,
      "step": 24230
    },
    {
      "entropy": 5.117392110824585,
      "epoch": 2.3280499519692603,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044591866625044154,
      "loss": 4.8208,
      "mean_token_accuracy": 0.23696815818548203,
      "num_tokens": 55591336.0,
      "step": 24235
    },
    {
      "entropy": 5.2030986785888675,
      "epoch": 2.3285302593659942,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004458963681089403,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22540994882583618,
      "num_tokens": 55601955.0,
      "step": 24240
    },
    {
      "entropy": 5.227811241149903,
      "epoch": 2.329010566762728,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004458740659997328,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21928848922252656,
      "num_tokens": 55614130.0,
      "step": 24245
    },
    {
      "entropy": 5.171048212051391,
      "epoch": 2.329490874159462,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004458517599233369,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22540259957313538,
      "num_tokens": 55625447.0,
      "step": 24250
    },
    {
      "entropy": 5.163996982574463,
      "epoch": 2.329971181556196,
      "grad_norm": 1.0,
      "learning_rate": 0.0004458294498802706,
      "loss": 4.8304,
      "mean_token_accuracy": 0.23129211962223054,
      "num_tokens": 55636426.0,
      "step": 24255
    },
    {
      "entropy": 5.074560260772705,
      "epoch": 2.3304514889529298,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004458071358710516,
      "loss": 4.799,
      "mean_token_accuracy": 0.22179099917411804,
      "num_tokens": 55647095.0,
      "step": 24260
    },
    {
      "entropy": 5.102037572860718,
      "epoch": 2.3309317963496636,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004457848178961981,
      "loss": 4.7793,
      "mean_token_accuracy": 0.22693680822849274,
      "num_tokens": 55660362.0,
      "step": 24265
    },
    {
      "entropy": 5.17506628036499,
      "epoch": 2.3314121037463975,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044576249595622833,
      "loss": 4.8059,
      "mean_token_accuracy": 0.23493959307670592,
      "num_tokens": 55671760.0,
      "step": 24270
    },
    {
      "entropy": 5.104570722579956,
      "epoch": 2.331892411143132,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004457401700516603,
      "loss": 4.8529,
      "mean_token_accuracy": 0.22955234348773956,
      "num_tokens": 55683826.0,
      "step": 24275
    },
    {
      "entropy": 5.203718900680542,
      "epoch": 2.3323727185398653,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044571784018301267,
      "loss": 4.8466,
      "mean_token_accuracy": 0.23341756463050842,
      "num_tokens": 55694939.0,
      "step": 24280
    },
    {
      "entropy": 5.150739908218384,
      "epoch": 2.3328530259365996,
      "grad_norm": 1.0,
      "learning_rate": 0.00044569550635080365,
      "loss": 4.8681,
      "mean_token_accuracy": 0.22752155661582946,
      "num_tokens": 55707101.0,
      "step": 24285
    },
    {
      "entropy": 5.120617198944092,
      "epoch": 2.3333333333333335,
      "grad_norm": 1.0,
      "learning_rate": 0.00044567316855555184,
      "loss": 4.7651,
      "mean_token_accuracy": 0.23352627605199813,
      "num_tokens": 55718418.0,
      "step": 24290
    },
    {
      "entropy": 5.235289144515991,
      "epoch": 2.3338136407300674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004456508267977759,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2175884649157524,
      "num_tokens": 55729229.0,
      "step": 24295
    },
    {
      "entropy": 5.223034906387329,
      "epoch": 2.3342939481268012,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044562848107799444,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22382025569677352,
      "num_tokens": 55740719.0,
      "step": 24300
    },
    {
      "entropy": 5.124778175354004,
      "epoch": 2.334774255523535,
      "grad_norm": 1.59375,
      "learning_rate": 0.00044560613139672627,
      "loss": 4.7577,
      "mean_token_accuracy": 0.2346653163433075,
      "num_tokens": 55752363.0,
      "step": 24305
    },
    {
      "entropy": 5.072336053848266,
      "epoch": 2.335254562920269,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044558377775449036,
      "loss": 4.7783,
      "mean_token_accuracy": 0.23282758742570878,
      "num_tokens": 55763851.0,
      "step": 24310
    },
    {
      "entropy": 5.143740272521972,
      "epoch": 2.335734870317003,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044556142015180573,
      "loss": 4.8278,
      "mean_token_accuracy": 0.23359145671129228,
      "num_tokens": 55776118.0,
      "step": 24315
    },
    {
      "entropy": 5.065422868728637,
      "epoch": 2.3362151777137368,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044553905858919134,
      "loss": 4.7486,
      "mean_token_accuracy": 0.23730223774909973,
      "num_tokens": 55788106.0,
      "step": 24320
    },
    {
      "entropy": 5.1730293273925785,
      "epoch": 2.3366954851104706,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004455166930671664,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22975881695747374,
      "num_tokens": 55798608.0,
      "step": 24325
    },
    {
      "entropy": 5.241461515426636,
      "epoch": 2.3371757925072045,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044549432358625014,
      "loss": 4.8883,
      "mean_token_accuracy": 0.22333077490329742,
      "num_tokens": 55810963.0,
      "step": 24330
    },
    {
      "entropy": 5.115719938278199,
      "epoch": 2.3376560999039384,
      "grad_norm": 0.96875,
      "learning_rate": 0.000445471950146962,
      "loss": 4.7529,
      "mean_token_accuracy": 0.23350639790296554,
      "num_tokens": 55821593.0,
      "step": 24335
    },
    {
      "entropy": 5.199576663970947,
      "epoch": 2.3381364073006723,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004454495727498214,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22622345834970475,
      "num_tokens": 55833602.0,
      "step": 24340
    },
    {
      "entropy": 5.259151649475098,
      "epoch": 2.338616714697406,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004454271913953478,
      "loss": 4.966,
      "mean_token_accuracy": 0.21381545960903167,
      "num_tokens": 55844745.0,
      "step": 24345
    },
    {
      "entropy": 5.291757202148437,
      "epoch": 2.3390970220941405,
      "grad_norm": 1.0,
      "learning_rate": 0.00044540480608406093,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2201313316822052,
      "num_tokens": 55857193.0,
      "step": 24350
    },
    {
      "entropy": 5.156404447555542,
      "epoch": 2.339577329490874,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004453824168164804,
      "loss": 4.8344,
      "mean_token_accuracy": 0.2253891035914421,
      "num_tokens": 55868954.0,
      "step": 24355
    },
    {
      "entropy": 5.09840407371521,
      "epoch": 2.3400576368876083,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004453600235931261,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2253431960940361,
      "num_tokens": 55880485.0,
      "step": 24360
    },
    {
      "entropy": 5.221623086929322,
      "epoch": 2.340537944284342,
      "grad_norm": 1.046875,
      "learning_rate": 0.000445337626414518,
      "loss": 4.8934,
      "mean_token_accuracy": 0.22641885429620742,
      "num_tokens": 55893043.0,
      "step": 24365
    },
    {
      "entropy": 5.11188588142395,
      "epoch": 2.341018251681076,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044531522528117593,
      "loss": 4.7726,
      "mean_token_accuracy": 0.2291277229785919,
      "num_tokens": 55905776.0,
      "step": 24370
    },
    {
      "entropy": 5.129027318954468,
      "epoch": 2.34149855907781,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044529282019362007,
      "loss": 4.8396,
      "mean_token_accuracy": 0.2337314084172249,
      "num_tokens": 55916424.0,
      "step": 24375
    },
    {
      "entropy": 5.240195846557617,
      "epoch": 2.341978866474544,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044527041115237056,
      "loss": 4.9954,
      "mean_token_accuracy": 0.21366714239120482,
      "num_tokens": 55928510.0,
      "step": 24380
    },
    {
      "entropy": 5.259598684310913,
      "epoch": 2.3424591738712777,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004452479981579477,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22125631719827651,
      "num_tokens": 55940536.0,
      "step": 24385
    },
    {
      "entropy": 5.204502916336059,
      "epoch": 2.3429394812680115,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004452255812108719,
      "loss": 4.8788,
      "mean_token_accuracy": 0.2271536558866501,
      "num_tokens": 55951837.0,
      "step": 24390
    },
    {
      "entropy": 5.002779388427735,
      "epoch": 2.3434197886647454,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004452031603116636,
      "loss": 4.712,
      "mean_token_accuracy": 0.2432721048593521,
      "num_tokens": 55963641.0,
      "step": 24395
    },
    {
      "entropy": 5.177810001373291,
      "epoch": 2.3439000960614793,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044518073546084326,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2269780457019806,
      "num_tokens": 55975205.0,
      "step": 24400
    },
    {
      "entropy": 5.213040113449097,
      "epoch": 2.344380403458213,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004451583066589316,
      "loss": 4.9078,
      "mean_token_accuracy": 0.2243112325668335,
      "num_tokens": 55985456.0,
      "step": 24405
    },
    {
      "entropy": 5.1889537334442135,
      "epoch": 2.344860710854947,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044513587390644925,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22910890132188796,
      "num_tokens": 55997946.0,
      "step": 24410
    },
    {
      "entropy": 5.219141054153442,
      "epoch": 2.345341018251681,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044511343720391724,
      "loss": 4.8791,
      "mean_token_accuracy": 0.2220301628112793,
      "num_tokens": 56008907.0,
      "step": 24415
    },
    {
      "entropy": 5.142144298553466,
      "epoch": 2.345821325648415,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004450909965518563,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22765176892280578,
      "num_tokens": 56019153.0,
      "step": 24420
    },
    {
      "entropy": 5.028193616867066,
      "epoch": 2.3463016330451487,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044506855195078755,
      "loss": 4.7081,
      "mean_token_accuracy": 0.2380005568265915,
      "num_tokens": 56029944.0,
      "step": 24425
    },
    {
      "entropy": 5.194375038146973,
      "epoch": 2.3467819404418826,
      "grad_norm": 1.0,
      "learning_rate": 0.00044504610340123185,
      "loss": 4.8299,
      "mean_token_accuracy": 0.2232501283288002,
      "num_tokens": 56041782.0,
      "step": 24430
    },
    {
      "entropy": 5.260311889648437,
      "epoch": 2.347262247838617,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00044502365090371066,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22843010872602462,
      "num_tokens": 56052803.0,
      "step": 24435
    },
    {
      "entropy": 5.182580280303955,
      "epoch": 2.347742555235351,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004450011944587452,
      "loss": 4.8574,
      "mean_token_accuracy": 0.2287664830684662,
      "num_tokens": 56062503.0,
      "step": 24440
    },
    {
      "entropy": 5.106720924377441,
      "epoch": 2.3482228626320847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044497873406685673,
      "loss": 4.8283,
      "mean_token_accuracy": 0.22876937538385392,
      "num_tokens": 56073147.0,
      "step": 24445
    },
    {
      "entropy": 5.192853116989136,
      "epoch": 2.3487031700288186,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004449562697285669,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22287747114896775,
      "num_tokens": 56084450.0,
      "step": 24450
    },
    {
      "entropy": 5.102748012542724,
      "epoch": 2.3491834774255524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044493380144439707,
      "loss": 4.7384,
      "mean_token_accuracy": 0.23315025717020035,
      "num_tokens": 56094869.0,
      "step": 24455
    },
    {
      "entropy": 5.112974691390991,
      "epoch": 2.3496637848222863,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000444911329214869,
      "loss": 4.7932,
      "mean_token_accuracy": 0.23882840424776078,
      "num_tokens": 56105640.0,
      "step": 24460
    },
    {
      "entropy": 5.21599063873291,
      "epoch": 2.35014409221902,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044488885304050434,
      "loss": 4.9228,
      "mean_token_accuracy": 0.22374927252531052,
      "num_tokens": 56116756.0,
      "step": 24465
    },
    {
      "entropy": 5.181053829193115,
      "epoch": 2.350624399615754,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000444866372921825,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22566210329532624,
      "num_tokens": 56129674.0,
      "step": 24470
    },
    {
      "entropy": 5.16304726600647,
      "epoch": 2.351104707012488,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044484388885935287,
      "loss": 4.9086,
      "mean_token_accuracy": 0.225508214533329,
      "num_tokens": 56141360.0,
      "step": 24475
    },
    {
      "entropy": 5.120934534072876,
      "epoch": 2.351585014409222,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044482140085361005,
      "loss": 4.7367,
      "mean_token_accuracy": 0.24205872863531114,
      "num_tokens": 56152995.0,
      "step": 24480
    },
    {
      "entropy": 5.2052747249603275,
      "epoch": 2.3520653218059557,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044479890890511853,
      "loss": 4.9239,
      "mean_token_accuracy": 0.22731008231639863,
      "num_tokens": 56164515.0,
      "step": 24485
    },
    {
      "entropy": 5.156913614273071,
      "epoch": 2.3525456292026896,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044477641301440054,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2303958684206009,
      "num_tokens": 56175983.0,
      "step": 24490
    },
    {
      "entropy": 5.097546768188477,
      "epoch": 2.3530259365994235,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004447539131819784,
      "loss": 4.7673,
      "mean_token_accuracy": 0.23214149326086045,
      "num_tokens": 56187176.0,
      "step": 24495
    },
    {
      "entropy": 5.15512900352478,
      "epoch": 2.3535062439961574,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044473140940837436,
      "loss": 4.8339,
      "mean_token_accuracy": 0.22434639781713486,
      "num_tokens": 56198713.0,
      "step": 24500
    },
    {
      "entropy": 5.0931768894195555,
      "epoch": 2.3539865513928913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044470890169411107,
      "loss": 4.8138,
      "mean_token_accuracy": 0.22905541211366653,
      "num_tokens": 56210473.0,
      "step": 24505
    },
    {
      "entropy": 5.155054330825806,
      "epoch": 2.3544668587896256,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000444686390039711,
      "loss": 4.9204,
      "mean_token_accuracy": 0.22321080267429352,
      "num_tokens": 56221850.0,
      "step": 24510
    },
    {
      "entropy": 5.158477640151977,
      "epoch": 2.3549471661863595,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004446638744456968,
      "loss": 4.8245,
      "mean_token_accuracy": 0.22970156818628312,
      "num_tokens": 56232359.0,
      "step": 24515
    },
    {
      "entropy": 5.1029047012329105,
      "epoch": 2.3554274735830933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044464135491259135,
      "loss": 4.8004,
      "mean_token_accuracy": 0.23129784166812897,
      "num_tokens": 56243385.0,
      "step": 24520
    },
    {
      "entropy": 5.145754337310791,
      "epoch": 2.3559077809798272,
      "grad_norm": 1.0,
      "learning_rate": 0.0004446188314409172,
      "loss": 4.8016,
      "mean_token_accuracy": 0.23743956238031388,
      "num_tokens": 56254713.0,
      "step": 24525
    },
    {
      "entropy": 5.122840309143067,
      "epoch": 2.356388088376561,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004445963040311975,
      "loss": 4.768,
      "mean_token_accuracy": 0.239417727291584,
      "num_tokens": 56266227.0,
      "step": 24530
    },
    {
      "entropy": 5.239283418655395,
      "epoch": 2.356868395773295,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044457377268395526,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21653226464986802,
      "num_tokens": 56276258.0,
      "step": 24535
    },
    {
      "entropy": 5.220578384399414,
      "epoch": 2.357348703170029,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044455123739971355,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22171889692544938,
      "num_tokens": 56287909.0,
      "step": 24540
    },
    {
      "entropy": 5.20037989616394,
      "epoch": 2.3578290105667628,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044452869817899554,
      "loss": 4.8948,
      "mean_token_accuracy": 0.23091649264097214,
      "num_tokens": 56299161.0,
      "step": 24545
    },
    {
      "entropy": 5.17031774520874,
      "epoch": 2.3583093179634966,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004445061550223246,
      "loss": 4.875,
      "mean_token_accuracy": 0.22813785970211028,
      "num_tokens": 56311333.0,
      "step": 24550
    },
    {
      "entropy": 5.145089340209961,
      "epoch": 2.3587896253602305,
      "grad_norm": 1.0,
      "learning_rate": 0.00044448360793022403,
      "loss": 4.7536,
      "mean_token_accuracy": 0.2421611174941063,
      "num_tokens": 56322349.0,
      "step": 24555
    },
    {
      "entropy": 5.16576452255249,
      "epoch": 2.3592699327569644,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004444610569032174,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22823168188333512,
      "num_tokens": 56333169.0,
      "step": 24560
    },
    {
      "entropy": 5.0576637268066404,
      "epoch": 2.3597502401536983,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004444385019418281,
      "loss": 4.7607,
      "mean_token_accuracy": 0.23781062066555023,
      "num_tokens": 56345419.0,
      "step": 24565
    },
    {
      "entropy": 5.173731470108033,
      "epoch": 2.360230547550432,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044441594304658004,
      "loss": 4.7781,
      "mean_token_accuracy": 0.23144095242023469,
      "num_tokens": 56355848.0,
      "step": 24570
    },
    {
      "entropy": 5.23099455833435,
      "epoch": 2.360710854947166,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004443933802179968,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2173900455236435,
      "num_tokens": 56368369.0,
      "step": 24575
    },
    {
      "entropy": 5.10700216293335,
      "epoch": 2.3611911623439,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044437081345660224,
      "loss": 4.7886,
      "mean_token_accuracy": 0.23209902942180632,
      "num_tokens": 56379069.0,
      "step": 24580
    },
    {
      "entropy": 5.152787733078003,
      "epoch": 2.3616714697406342,
      "grad_norm": 1.25,
      "learning_rate": 0.0004443482427629204,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22924163043498993,
      "num_tokens": 56390325.0,
      "step": 24585
    },
    {
      "entropy": 5.212109470367432,
      "epoch": 2.3621517771373677,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004443256681374751,
      "loss": 4.9016,
      "mean_token_accuracy": 0.23101179003715516,
      "num_tokens": 56401336.0,
      "step": 24590
    },
    {
      "entropy": 5.179910707473755,
      "epoch": 2.362632084534102,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004443030895807907,
      "loss": 4.9315,
      "mean_token_accuracy": 0.22279146015644075,
      "num_tokens": 56412547.0,
      "step": 24595
    },
    {
      "entropy": 5.303513240814209,
      "epoch": 2.363112391930836,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044428050709339117,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22898103147745133,
      "num_tokens": 56423757.0,
      "step": 24600
    },
    {
      "entropy": 5.114374589920044,
      "epoch": 2.3635926993275698,
      "grad_norm": 1.125,
      "learning_rate": 0.000444257920675801,
      "loss": 4.8591,
      "mean_token_accuracy": 0.2314160943031311,
      "num_tokens": 56433936.0,
      "step": 24605
    },
    {
      "entropy": 5.090903043746948,
      "epoch": 2.3640730067243036,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044423533032854454,
      "loss": 4.7984,
      "mean_token_accuracy": 0.23257745802402496,
      "num_tokens": 56446504.0,
      "step": 24610
    },
    {
      "entropy": 5.218806028366089,
      "epoch": 2.3645533141210375,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004442127360521462,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22271284461021423,
      "num_tokens": 56459019.0,
      "step": 24615
    },
    {
      "entropy": 5.171198320388794,
      "epoch": 2.3650336215177714,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004441901378471306,
      "loss": 4.8564,
      "mean_token_accuracy": 0.2300712615251541,
      "num_tokens": 56470208.0,
      "step": 24620
    },
    {
      "entropy": 5.156589984893799,
      "epoch": 2.3655139289145053,
      "grad_norm": 1.125,
      "learning_rate": 0.00044416753571402233,
      "loss": 4.8649,
      "mean_token_accuracy": 0.2206113114953041,
      "num_tokens": 56482188.0,
      "step": 24625
    },
    {
      "entropy": 5.13980073928833,
      "epoch": 2.365994236311239,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004441449296533462,
      "loss": 4.7921,
      "mean_token_accuracy": 0.22808422446250914,
      "num_tokens": 56493599.0,
      "step": 24630
    },
    {
      "entropy": 5.226800060272216,
      "epoch": 2.366474543707973,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044412231966562717,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21908214092254638,
      "num_tokens": 56504448.0,
      "step": 24635
    },
    {
      "entropy": 5.1718910217285154,
      "epoch": 2.366954851104707,
      "grad_norm": 1.0,
      "learning_rate": 0.00044409970575139,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22469279021024705,
      "num_tokens": 56517361.0,
      "step": 24640
    },
    {
      "entropy": 5.1234580993652346,
      "epoch": 2.367435158501441,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004440770879111598,
      "loss": 4.7339,
      "mean_token_accuracy": 0.23614125847816467,
      "num_tokens": 56530152.0,
      "step": 24645
    },
    {
      "entropy": 5.259749269485473,
      "epoch": 2.3679154658981747,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044405446614546163,
      "loss": 4.9298,
      "mean_token_accuracy": 0.226273213326931,
      "num_tokens": 56541121.0,
      "step": 24650
    },
    {
      "entropy": 5.108071613311767,
      "epoch": 2.3683957732949086,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004440318404548208,
      "loss": 4.7603,
      "mean_token_accuracy": 0.23324883878231048,
      "num_tokens": 56551697.0,
      "step": 24655
    },
    {
      "entropy": 5.277592372894287,
      "epoch": 2.3688760806916425,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044400921083976246,
      "loss": 5.0573,
      "mean_token_accuracy": 0.2089755967259407,
      "num_tokens": 56564250.0,
      "step": 24660
    },
    {
      "entropy": 5.231600952148438,
      "epoch": 2.3693563880883763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004439865773008122,
      "loss": 4.8607,
      "mean_token_accuracy": 0.2207558423280716,
      "num_tokens": 56574963.0,
      "step": 24665
    },
    {
      "entropy": 5.114921569824219,
      "epoch": 2.3698366954851107,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004439639398384953,
      "loss": 4.8206,
      "mean_token_accuracy": 0.2344073697924614,
      "num_tokens": 56585516.0,
      "step": 24670
    },
    {
      "entropy": 5.0840599060058596,
      "epoch": 2.3703170028818445,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044394129845333756,
      "loss": 4.7795,
      "mean_token_accuracy": 0.23086483925580978,
      "num_tokens": 56597459.0,
      "step": 24675
    },
    {
      "entropy": 5.172188711166382,
      "epoch": 2.3707973102785784,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004439186531458645,
      "loss": 4.8562,
      "mean_token_accuracy": 0.2296239972114563,
      "num_tokens": 56608942.0,
      "step": 24680
    },
    {
      "entropy": 5.227690601348877,
      "epoch": 2.3712776176753123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044389600391660185,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22034805864095688,
      "num_tokens": 56621444.0,
      "step": 24685
    },
    {
      "entropy": 5.203283643722534,
      "epoch": 2.371757925072046,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044387335076607554,
      "loss": 4.7592,
      "mean_token_accuracy": 0.23789092451333999,
      "num_tokens": 56631734.0,
      "step": 24690
    },
    {
      "entropy": 5.094226741790772,
      "epoch": 2.37223823246878,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004438506936948115,
      "loss": 4.829,
      "mean_token_accuracy": 0.23035890311002732,
      "num_tokens": 56644140.0,
      "step": 24695
    },
    {
      "entropy": 5.072202968597412,
      "epoch": 2.372718539865514,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044382803270333565,
      "loss": 4.7728,
      "mean_token_accuracy": 0.2377172440290451,
      "num_tokens": 56655071.0,
      "step": 24700
    },
    {
      "entropy": 5.167391157150268,
      "epoch": 2.373198847262248,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004438053677921743,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23240213096141815,
      "num_tokens": 56667434.0,
      "step": 24705
    },
    {
      "entropy": 5.222027015686035,
      "epoch": 2.3736791546589817,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044378269896185344,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22397251725196837,
      "num_tokens": 56677675.0,
      "step": 24710
    },
    {
      "entropy": 5.150626707077026,
      "epoch": 2.3741594620557156,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004437600262128996,
      "loss": 4.8146,
      "mean_token_accuracy": 0.22969537526369094,
      "num_tokens": 56687849.0,
      "step": 24715
    },
    {
      "entropy": 5.175713777542114,
      "epoch": 2.3746397694524495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000443737349545839,
      "loss": 4.838,
      "mean_token_accuracy": 0.22802821397781373,
      "num_tokens": 56699637.0,
      "step": 24720
    },
    {
      "entropy": 5.238686990737915,
      "epoch": 2.3751200768491834,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044371466896119823,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22427304834127426,
      "num_tokens": 56711412.0,
      "step": 24725
    },
    {
      "entropy": 5.297528982162476,
      "epoch": 2.3756003842459172,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044369198445950384,
      "loss": 4.9543,
      "mean_token_accuracy": 0.2290105253458023,
      "num_tokens": 56723379.0,
      "step": 24730
    },
    {
      "entropy": 5.183445835113526,
      "epoch": 2.376080691642651,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004436692960412824,
      "loss": 4.949,
      "mean_token_accuracy": 0.22714217007160187,
      "num_tokens": 56734936.0,
      "step": 24735
    },
    {
      "entropy": 5.123217821121216,
      "epoch": 2.376560999039385,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004436466037070608,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23787820488214492,
      "num_tokens": 56746925.0,
      "step": 24740
    },
    {
      "entropy": 5.241061019897461,
      "epoch": 2.3770413064361193,
      "grad_norm": 1.125,
      "learning_rate": 0.00044362390745736585,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2284475862979889,
      "num_tokens": 56759034.0,
      "step": 24745
    },
    {
      "entropy": 5.18985767364502,
      "epoch": 2.377521613832853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004436012072927245,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22885988503694535,
      "num_tokens": 56770882.0,
      "step": 24750
    },
    {
      "entropy": 5.059929895401001,
      "epoch": 2.378001921229587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044357850321366375,
      "loss": 4.7897,
      "mean_token_accuracy": 0.2338315039873123,
      "num_tokens": 56782110.0,
      "step": 24755
    },
    {
      "entropy": 5.148529434204102,
      "epoch": 2.378482228626321,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004435557952207107,
      "loss": 4.8156,
      "mean_token_accuracy": 0.22787342220544815,
      "num_tokens": 56793689.0,
      "step": 24760
    },
    {
      "entropy": 5.119774675369262,
      "epoch": 2.378962536023055,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044353308331439257,
      "loss": 4.8363,
      "mean_token_accuracy": 0.23164703398942948,
      "num_tokens": 56805315.0,
      "step": 24765
    },
    {
      "entropy": 5.187684059143066,
      "epoch": 2.3794428434197887,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004435103674952367,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22694899141788483,
      "num_tokens": 56816738.0,
      "step": 24770
    },
    {
      "entropy": 5.276415061950684,
      "epoch": 2.3799231508165226,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044348764776377047,
      "loss": 5.0569,
      "mean_token_accuracy": 0.2162120133638382,
      "num_tokens": 56828613.0,
      "step": 24775
    },
    {
      "entropy": 5.257706880569458,
      "epoch": 2.3804034582132565,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004434649241205214,
      "loss": 4.8853,
      "mean_token_accuracy": 0.2237246036529541,
      "num_tokens": 56840373.0,
      "step": 24780
    },
    {
      "entropy": 5.172229671478272,
      "epoch": 2.3808837656099904,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044344219656601704,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2187038093805313,
      "num_tokens": 56850581.0,
      "step": 24785
    },
    {
      "entropy": 5.182748079299927,
      "epoch": 2.3813640730067243,
      "grad_norm": 1.203125,
      "learning_rate": 0.000443419465100785,
      "loss": 4.8339,
      "mean_token_accuracy": 0.2342318519949913,
      "num_tokens": 56861576.0,
      "step": 24790
    },
    {
      "entropy": 5.238910293579101,
      "epoch": 2.381844380403458,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004433967297253531,
      "loss": 4.9466,
      "mean_token_accuracy": 0.22809169590473174,
      "num_tokens": 56872833.0,
      "step": 24795
    },
    {
      "entropy": 5.199377870559692,
      "epoch": 2.382324687800192,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044337399044024924,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2292731985449791,
      "num_tokens": 56884367.0,
      "step": 24800
    },
    {
      "entropy": 5.04766583442688,
      "epoch": 2.382804995196926,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004433512472460012,
      "loss": 4.7416,
      "mean_token_accuracy": 0.2403650090098381,
      "num_tokens": 56895233.0,
      "step": 24805
    },
    {
      "entropy": 5.097816085815429,
      "epoch": 2.38328530259366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044332850014313713,
      "loss": 4.8636,
      "mean_token_accuracy": 0.2280310794711113,
      "num_tokens": 56907133.0,
      "step": 24810
    },
    {
      "entropy": 5.2622997760772705,
      "epoch": 2.3837656099903937,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004433057491321851,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22770422250032424,
      "num_tokens": 56919372.0,
      "step": 24815
    },
    {
      "entropy": 5.196826982498169,
      "epoch": 2.384245917387128,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044328299421367333,
      "loss": 4.8971,
      "mean_token_accuracy": 0.22053462713956834,
      "num_tokens": 56930523.0,
      "step": 24820
    },
    {
      "entropy": 5.06194281578064,
      "epoch": 2.3847262247838614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004432602353881302,
      "loss": 4.7358,
      "mean_token_accuracy": 0.23345678448677062,
      "num_tokens": 56941324.0,
      "step": 24825
    },
    {
      "entropy": 5.024616670608521,
      "epoch": 2.3852065321805958,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044323747265608395,
      "loss": 4.731,
      "mean_token_accuracy": 0.2327443167567253,
      "num_tokens": 56952359.0,
      "step": 24830
    },
    {
      "entropy": 5.159119653701782,
      "epoch": 2.3856868395773296,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004432147060180632,
      "loss": 4.8139,
      "mean_token_accuracy": 0.23433706909418106,
      "num_tokens": 56963456.0,
      "step": 24835
    },
    {
      "entropy": 5.086129665374756,
      "epoch": 2.3861671469740635,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044319193547459645,
      "loss": 4.7366,
      "mean_token_accuracy": 0.2367536962032318,
      "num_tokens": 56975029.0,
      "step": 24840
    },
    {
      "entropy": 5.104020977020264,
      "epoch": 2.3866474543707974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004431691610262124,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22686857134103774,
      "num_tokens": 56987159.0,
      "step": 24845
    },
    {
      "entropy": 5.182588529586792,
      "epoch": 2.3871277617675313,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044314638267343976,
      "loss": 4.9071,
      "mean_token_accuracy": 0.2303967759013176,
      "num_tokens": 56998359.0,
      "step": 24850
    },
    {
      "entropy": 5.186446475982666,
      "epoch": 2.387608069164265,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004431236004168075,
      "loss": 4.8842,
      "mean_token_accuracy": 0.2184463247656822,
      "num_tokens": 57009980.0,
      "step": 24855
    },
    {
      "entropy": 5.181201457977295,
      "epoch": 2.388088376560999,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004431008142568444,
      "loss": 4.861,
      "mean_token_accuracy": 0.2269979938864708,
      "num_tokens": 57022531.0,
      "step": 24860
    },
    {
      "entropy": 5.246126413345337,
      "epoch": 2.388568683957733,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044307802419407954,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21920094043016433,
      "num_tokens": 57034746.0,
      "step": 24865
    },
    {
      "entropy": 5.164705991744995,
      "epoch": 2.389048991354467,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004430552302290421,
      "loss": 4.8763,
      "mean_token_accuracy": 0.226024629175663,
      "num_tokens": 57046496.0,
      "step": 24870
    },
    {
      "entropy": 5.114444160461426,
      "epoch": 2.3895292987512007,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004430324323622611,
      "loss": 4.8192,
      "mean_token_accuracy": 0.22778339684009552,
      "num_tokens": 57059083.0,
      "step": 24875
    },
    {
      "entropy": 5.194456338882446,
      "epoch": 2.3900096061479346,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044300963059426605,
      "loss": 4.8261,
      "mean_token_accuracy": 0.23668499290943146,
      "num_tokens": 57070077.0,
      "step": 24880
    },
    {
      "entropy": 5.194018030166626,
      "epoch": 2.3904899135446684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044298682492558637,
      "loss": 4.7959,
      "mean_token_accuracy": 0.22870599031448363,
      "num_tokens": 57081397.0,
      "step": 24885
    },
    {
      "entropy": 5.16359395980835,
      "epoch": 2.3909702209414023,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044296401535675136,
      "loss": 4.8814,
      "mean_token_accuracy": 0.22575587183237075,
      "num_tokens": 57093882.0,
      "step": 24890
    },
    {
      "entropy": 5.241537046432495,
      "epoch": 2.3914505283381366,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044294120188829056,
      "loss": 4.9557,
      "mean_token_accuracy": 0.22046367526054383,
      "num_tokens": 57104603.0,
      "step": 24895
    },
    {
      "entropy": 5.261300706863404,
      "epoch": 2.39193083573487,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004429183845207339,
      "loss": 4.9161,
      "mean_token_accuracy": 0.22526892721652986,
      "num_tokens": 57115512.0,
      "step": 24900
    },
    {
      "entropy": 5.1826756477355955,
      "epoch": 2.3924111431316044,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004428955632546108,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22585040628910064,
      "num_tokens": 57125934.0,
      "step": 24905
    },
    {
      "entropy": 5.104539775848389,
      "epoch": 2.3928914505283383,
      "grad_norm": 1.0,
      "learning_rate": 0.0004428727380904514,
      "loss": 4.7612,
      "mean_token_accuracy": 0.24016287922859192,
      "num_tokens": 57136437.0,
      "step": 24910
    },
    {
      "entropy": 5.141263246536255,
      "epoch": 2.393371757925072,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044284990902878545,
      "loss": 4.8432,
      "mean_token_accuracy": 0.21777609288692473,
      "num_tokens": 57147542.0,
      "step": 24915
    },
    {
      "entropy": 5.1258820533752445,
      "epoch": 2.393852065321806,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044282707607014304,
      "loss": 4.8269,
      "mean_token_accuracy": 0.2250780791044235,
      "num_tokens": 57159248.0,
      "step": 24920
    },
    {
      "entropy": 5.15540246963501,
      "epoch": 2.39433237271854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044280423921505427,
      "loss": 4.8432,
      "mean_token_accuracy": 0.22625250667333602,
      "num_tokens": 57171627.0,
      "step": 24925
    },
    {
      "entropy": 5.1938642978668215,
      "epoch": 2.394812680115274,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004427813984640493,
      "loss": 4.8377,
      "mean_token_accuracy": 0.23405671864748,
      "num_tokens": 57181604.0,
      "step": 24930
    },
    {
      "entropy": 5.081823825836182,
      "epoch": 2.3952929875120077,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004427585538176585,
      "loss": 4.7484,
      "mean_token_accuracy": 0.2361682504415512,
      "num_tokens": 57191409.0,
      "step": 24935
    },
    {
      "entropy": 5.098561191558838,
      "epoch": 2.3957732949087416,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044273570527641223,
      "loss": 4.8062,
      "mean_token_accuracy": 0.2310518130660057,
      "num_tokens": 57204606.0,
      "step": 24940
    },
    {
      "entropy": 5.209570360183716,
      "epoch": 2.3962536023054755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044271285284084097,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22264860570430756,
      "num_tokens": 57215641.0,
      "step": 24945
    },
    {
      "entropy": 5.181931209564209,
      "epoch": 2.3967339097022093,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004426899965114752,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21922363340854645,
      "num_tokens": 57226520.0,
      "step": 24950
    },
    {
      "entropy": 5.219824361801147,
      "epoch": 2.3972142170989432,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044266713628884566,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2174603059887886,
      "num_tokens": 57237859.0,
      "step": 24955
    },
    {
      "entropy": 5.2045619010925295,
      "epoch": 2.397694524495677,
      "grad_norm": 1.125,
      "learning_rate": 0.00044264427217348315,
      "loss": 4.7727,
      "mean_token_accuracy": 0.2394646465778351,
      "num_tokens": 57248727.0,
      "step": 24960
    },
    {
      "entropy": 5.120057821273804,
      "epoch": 2.398174831892411,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004426214041659184,
      "loss": 4.8245,
      "mean_token_accuracy": 0.23315869867801667,
      "num_tokens": 57259644.0,
      "step": 24965
    },
    {
      "entropy": 5.198820495605469,
      "epoch": 2.398655139289145,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004425985322666824,
      "loss": 4.8824,
      "mean_token_accuracy": 0.22823734283447267,
      "num_tokens": 57271140.0,
      "step": 24970
    },
    {
      "entropy": 5.177461576461792,
      "epoch": 2.3991354466858787,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004425756564763061,
      "loss": 4.8475,
      "mean_token_accuracy": 0.23296259641647338,
      "num_tokens": 57281205.0,
      "step": 24975
    },
    {
      "entropy": 5.14650149345398,
      "epoch": 2.399615754082613,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044255277679532075,
      "loss": 4.8187,
      "mean_token_accuracy": 0.2288091778755188,
      "num_tokens": 57292056.0,
      "step": 24980
    },
    {
      "entropy": 5.159703016281128,
      "epoch": 2.400096061479347,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044252989322425735,
      "loss": 4.8279,
      "mean_token_accuracy": 0.232200588285923,
      "num_tokens": 57303444.0,
      "step": 24985
    },
    {
      "entropy": 5.161099433898926,
      "epoch": 2.400576368876081,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044250700576364734,
      "loss": 4.8326,
      "mean_token_accuracy": 0.2258690759539604,
      "num_tokens": 57314784.0,
      "step": 24990
    },
    {
      "entropy": 5.06792631149292,
      "epoch": 2.4010566762728147,
      "grad_norm": 1.0,
      "learning_rate": 0.000442484114414022,
      "loss": 4.7549,
      "mean_token_accuracy": 0.23772234171628953,
      "num_tokens": 57325896.0,
      "step": 24995
    },
    {
      "entropy": 5.122263956069946,
      "epoch": 2.4015369836695486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004424612191759129,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22501546740531922,
      "num_tokens": 57338338.0,
      "step": 25000
    },
    {
      "entropy": 5.197610139846802,
      "epoch": 2.4020172910662825,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004424383200498515,
      "loss": 4.8896,
      "mean_token_accuracy": 0.2271687537431717,
      "num_tokens": 57350038.0,
      "step": 25005
    },
    {
      "entropy": 5.226184940338134,
      "epoch": 2.4024975984630164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004424154170363696,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2262828081846237,
      "num_tokens": 57362371.0,
      "step": 25010
    },
    {
      "entropy": 5.1837273120880125,
      "epoch": 2.4029779058597502,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004423925101359987,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22076825946569442,
      "num_tokens": 57374032.0,
      "step": 25015
    },
    {
      "entropy": 5.132214689254761,
      "epoch": 2.403458213256484,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004423695993492709,
      "loss": 4.8211,
      "mean_token_accuracy": 0.2303972586989403,
      "num_tokens": 57385705.0,
      "step": 25020
    },
    {
      "entropy": 5.118337678909302,
      "epoch": 2.403938520653218,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004423466846767179,
      "loss": 4.7849,
      "mean_token_accuracy": 0.23668774366378784,
      "num_tokens": 57396863.0,
      "step": 25025
    },
    {
      "entropy": 5.331267976760865,
      "epoch": 2.404418828049952,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044232376611887185,
      "loss": 5.0471,
      "mean_token_accuracy": 0.2171625390648842,
      "num_tokens": 57408486.0,
      "step": 25030
    },
    {
      "entropy": 5.202099990844727,
      "epoch": 2.4048991354466858,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044230084367626477,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22726424038410187,
      "num_tokens": 57419899.0,
      "step": 25035
    },
    {
      "entropy": 5.1954042434692385,
      "epoch": 2.4053794428434196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004422779173494288,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22586333155632018,
      "num_tokens": 57431620.0,
      "step": 25040
    },
    {
      "entropy": 5.134331226348877,
      "epoch": 2.4058597502401535,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004422549871388965,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22100483626127243,
      "num_tokens": 57443985.0,
      "step": 25045
    },
    {
      "entropy": 5.2015503406524655,
      "epoch": 2.4063400576368874,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044223205304519994,
      "loss": 4.8473,
      "mean_token_accuracy": 0.22980797439813613,
      "num_tokens": 57455160.0,
      "step": 25050
    },
    {
      "entropy": 5.1681419849395756,
      "epoch": 2.4068203650336217,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004422091150688717,
      "loss": 4.8771,
      "mean_token_accuracy": 0.2347584456205368,
      "num_tokens": 57467557.0,
      "step": 25055
    },
    {
      "entropy": 5.159786653518677,
      "epoch": 2.4073006724303556,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004421861732104443,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22550536394119264,
      "num_tokens": 57478749.0,
      "step": 25060
    },
    {
      "entropy": 5.14363317489624,
      "epoch": 2.4077809798270895,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004421632274704504,
      "loss": 4.8168,
      "mean_token_accuracy": 0.2298109635710716,
      "num_tokens": 57489899.0,
      "step": 25065
    },
    {
      "entropy": 5.269582223892212,
      "epoch": 2.4082612872238234,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004421402778494227,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22471741139888762,
      "num_tokens": 57502086.0,
      "step": 25070
    },
    {
      "entropy": 5.089459753036499,
      "epoch": 2.4087415946205573,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004421173243478941,
      "loss": 4.7717,
      "mean_token_accuracy": 0.23339829444885254,
      "num_tokens": 57513555.0,
      "step": 25075
    },
    {
      "entropy": 5.220227384567261,
      "epoch": 2.409221902017291,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044209436696639745,
      "loss": 4.8562,
      "mean_token_accuracy": 0.2232603460550308,
      "num_tokens": 57525100.0,
      "step": 25080
    },
    {
      "entropy": 5.1753706455230715,
      "epoch": 2.409702209414025,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044207140570546574,
      "loss": 4.8312,
      "mean_token_accuracy": 0.22801396399736404,
      "num_tokens": 57536355.0,
      "step": 25085
    },
    {
      "entropy": 5.059002685546875,
      "epoch": 2.410182516810759,
      "grad_norm": 1.0,
      "learning_rate": 0.00044204844056563216,
      "loss": 4.7891,
      "mean_token_accuracy": 0.2342930853366852,
      "num_tokens": 57547260.0,
      "step": 25090
    },
    {
      "entropy": 5.150734615325928,
      "epoch": 2.410662824207493,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004420254715474297,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22953960448503494,
      "num_tokens": 57558299.0,
      "step": 25095
    },
    {
      "entropy": 5.065666675567627,
      "epoch": 2.4111431316042267,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044200249865139187,
      "loss": 4.8236,
      "mean_token_accuracy": 0.23243228495121002,
      "num_tokens": 57569393.0,
      "step": 25100
    },
    {
      "entropy": 5.163590812683106,
      "epoch": 2.4116234390009605,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044197952187805185,
      "loss": 4.8643,
      "mean_token_accuracy": 0.22479417622089387,
      "num_tokens": 57580585.0,
      "step": 25105
    },
    {
      "entropy": 5.233012676239014,
      "epoch": 2.4121037463976944,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044195654122794324,
      "loss": 4.926,
      "mean_token_accuracy": 0.22503596693277358,
      "num_tokens": 57592887.0,
      "step": 25110
    },
    {
      "entropy": 5.245716905593872,
      "epoch": 2.4125840537944283,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004419335567015994,
      "loss": 4.9253,
      "mean_token_accuracy": 0.21861074715852738,
      "num_tokens": 57605129.0,
      "step": 25115
    },
    {
      "entropy": 5.067347764968872,
      "epoch": 2.413064361191162,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004419105682995542,
      "loss": 4.747,
      "mean_token_accuracy": 0.23804236799478531,
      "num_tokens": 57616852.0,
      "step": 25120
    },
    {
      "entropy": 5.143368911743164,
      "epoch": 2.413544668587896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004418875760223411,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2219969928264618,
      "num_tokens": 57629808.0,
      "step": 25125
    },
    {
      "entropy": 5.282468461990357,
      "epoch": 2.4140249759846304,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044186457987049405,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22217965126037598,
      "num_tokens": 57641025.0,
      "step": 25130
    },
    {
      "entropy": 5.097671222686768,
      "epoch": 2.414505283381364,
      "grad_norm": 1.0625,
      "learning_rate": 0.000441841579844547,
      "loss": 4.8178,
      "mean_token_accuracy": 0.23018050342798232,
      "num_tokens": 57652240.0,
      "step": 25135
    },
    {
      "entropy": 5.06086106300354,
      "epoch": 2.414985590778098,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004418185759450338,
      "loss": 4.7628,
      "mean_token_accuracy": 0.2420410543680191,
      "num_tokens": 57663842.0,
      "step": 25140
    },
    {
      "entropy": 5.150963640213012,
      "epoch": 2.415465898174832,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004417955681724887,
      "loss": 4.7864,
      "mean_token_accuracy": 0.23592489957809448,
      "num_tokens": 57674299.0,
      "step": 25145
    },
    {
      "entropy": 5.073912525177002,
      "epoch": 2.415946205571566,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044177255652744576,
      "loss": 4.7938,
      "mean_token_accuracy": 0.23812906295061112,
      "num_tokens": 57684858.0,
      "step": 25150
    },
    {
      "entropy": 5.211230945587158,
      "epoch": 2.4164265129683,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044174954101043926,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22357097417116165,
      "num_tokens": 57696542.0,
      "step": 25155
    },
    {
      "entropy": 5.1052830696105955,
      "epoch": 2.4169068203650337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044172652162200354,
      "loss": 4.7671,
      "mean_token_accuracy": 0.23986016660928727,
      "num_tokens": 57708603.0,
      "step": 25160
    },
    {
      "entropy": 5.111344671249389,
      "epoch": 2.4173871277617676,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004417034983626731,
      "loss": 4.8572,
      "mean_token_accuracy": 0.22818624824285508,
      "num_tokens": 57720273.0,
      "step": 25165
    },
    {
      "entropy": 5.152489137649536,
      "epoch": 2.4178674351585014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004416804712329825,
      "loss": 4.7796,
      "mean_token_accuracy": 0.22938569635152817,
      "num_tokens": 57730915.0,
      "step": 25170
    },
    {
      "entropy": 5.071908760070801,
      "epoch": 2.4183477425552353,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044165744023346614,
      "loss": 4.7984,
      "mean_token_accuracy": 0.23985119462013244,
      "num_tokens": 57743379.0,
      "step": 25175
    },
    {
      "entropy": 5.118296575546265,
      "epoch": 2.418828049951969,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044163440536465904,
      "loss": 4.8412,
      "mean_token_accuracy": 0.229281684756279,
      "num_tokens": 57754441.0,
      "step": 25180
    },
    {
      "entropy": 5.227838563919067,
      "epoch": 2.419308357348703,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044161136662709577,
      "loss": 4.8472,
      "mean_token_accuracy": 0.2235688880085945,
      "num_tokens": 57765701.0,
      "step": 25185
    },
    {
      "entropy": 5.17133903503418,
      "epoch": 2.419788664745437,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044158832402131133,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2249674916267395,
      "num_tokens": 57776774.0,
      "step": 25190
    },
    {
      "entropy": 5.195941925048828,
      "epoch": 2.420268972142171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044156527754784066,
      "loss": 4.8969,
      "mean_token_accuracy": 0.2269780606031418,
      "num_tokens": 57787618.0,
      "step": 25195
    },
    {
      "entropy": 5.273792362213134,
      "epoch": 2.4207492795389047,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044154222720721887,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21989891678094864,
      "num_tokens": 57800133.0,
      "step": 25200
    },
    {
      "entropy": 5.181040859222412,
      "epoch": 2.421229586935639,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000441519172999981,
      "loss": 4.8116,
      "mean_token_accuracy": 0.23111788928508759,
      "num_tokens": 57812736.0,
      "step": 25205
    },
    {
      "entropy": 5.130559825897217,
      "epoch": 2.4217098943323725,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004414961149266625,
      "loss": 4.8137,
      "mean_token_accuracy": 0.232513864338398,
      "num_tokens": 57824459.0,
      "step": 25210
    },
    {
      "entropy": 5.170581102371216,
      "epoch": 2.422190201729107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044147305298779856,
      "loss": 4.885,
      "mean_token_accuracy": 0.22741056680679322,
      "num_tokens": 57836292.0,
      "step": 25215
    },
    {
      "entropy": 5.12136287689209,
      "epoch": 2.4226705091258407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004414499871839247,
      "loss": 4.8306,
      "mean_token_accuracy": 0.2290224775671959,
      "num_tokens": 57848197.0,
      "step": 25220
    },
    {
      "entropy": 5.162555885314942,
      "epoch": 2.4231508165225746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004414269175155763,
      "loss": 4.8266,
      "mean_token_accuracy": 0.2278594732284546,
      "num_tokens": 57859491.0,
      "step": 25225
    },
    {
      "entropy": 5.0802319049835205,
      "epoch": 2.4236311239193085,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004414038439832891,
      "loss": 4.742,
      "mean_token_accuracy": 0.23962815701961518,
      "num_tokens": 57870199.0,
      "step": 25230
    },
    {
      "entropy": 5.174031829833984,
      "epoch": 2.4241114313160423,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004413807665875988,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22426361739635467,
      "num_tokens": 57882059.0,
      "step": 25235
    },
    {
      "entropy": 5.138275384902954,
      "epoch": 2.4245917387127762,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044135768532904104,
      "loss": 4.8403,
      "mean_token_accuracy": 0.234387569129467,
      "num_tokens": 57892316.0,
      "step": 25240
    },
    {
      "entropy": 5.08466420173645,
      "epoch": 2.42507204610951,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000441334600208152,
      "loss": 4.7999,
      "mean_token_accuracy": 0.2360578238964081,
      "num_tokens": 57903273.0,
      "step": 25245
    },
    {
      "entropy": 5.175006008148193,
      "epoch": 2.425552353506244,
      "grad_norm": 1.0,
      "learning_rate": 0.00044131151122546724,
      "loss": 4.923,
      "mean_token_accuracy": 0.22188410609960557,
      "num_tokens": 57914484.0,
      "step": 25250
    },
    {
      "entropy": 5.155128240585327,
      "epoch": 2.426032660902978,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044128841838152313,
      "loss": 4.7498,
      "mean_token_accuracy": 0.2311859905719757,
      "num_tokens": 57925213.0,
      "step": 25255
    },
    {
      "entropy": 5.135756874084473,
      "epoch": 2.4265129682997117,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004412653216768558,
      "loss": 4.8267,
      "mean_token_accuracy": 0.22824662178754807,
      "num_tokens": 57937422.0,
      "step": 25260
    },
    {
      "entropy": 5.119731760025024,
      "epoch": 2.4269932756964456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004412422211120013,
      "loss": 4.8173,
      "mean_token_accuracy": 0.23299674689769745,
      "num_tokens": 57947576.0,
      "step": 25265
    },
    {
      "entropy": 5.126709127426148,
      "epoch": 2.4274735830931795,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004412191166874961,
      "loss": 4.8037,
      "mean_token_accuracy": 0.234458327293396,
      "num_tokens": 57958998.0,
      "step": 25270
    },
    {
      "entropy": 5.189592361450195,
      "epoch": 2.4279538904899134,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004411960084038766,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22754490822553636,
      "num_tokens": 57970126.0,
      "step": 25275
    },
    {
      "entropy": 5.136882686614991,
      "epoch": 2.4284341978866473,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044117289626167917,
      "loss": 4.8276,
      "mean_token_accuracy": 0.2325281247496605,
      "num_tokens": 57981523.0,
      "step": 25280
    },
    {
      "entropy": 5.1236231327056885,
      "epoch": 2.428914505283381,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004411497802614406,
      "loss": 4.7728,
      "mean_token_accuracy": 0.23790016770362854,
      "num_tokens": 57992113.0,
      "step": 25285
    },
    {
      "entropy": 5.2023553371429445,
      "epoch": 2.4293948126801155,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004411266604036975,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22422919273376465,
      "num_tokens": 58005163.0,
      "step": 25290
    },
    {
      "entropy": 5.1836565971374515,
      "epoch": 2.4298751200768494,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044110353668898674,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22558384388685226,
      "num_tokens": 58015095.0,
      "step": 25295
    },
    {
      "entropy": 5.116381359100342,
      "epoch": 2.4303554274735832,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004410804091178449,
      "loss": 4.7714,
      "mean_token_accuracy": 0.2314148411154747,
      "num_tokens": 58025522.0,
      "step": 25300
    },
    {
      "entropy": 5.131482028961182,
      "epoch": 2.430835734870317,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004410572776908092,
      "loss": 4.835,
      "mean_token_accuracy": 0.22797947525978088,
      "num_tokens": 58036567.0,
      "step": 25305
    },
    {
      "entropy": 5.16339955329895,
      "epoch": 2.431316042267051,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044103414240841664,
      "loss": 4.7953,
      "mean_token_accuracy": 0.23575274646282196,
      "num_tokens": 58048321.0,
      "step": 25310
    },
    {
      "entropy": 5.159911251068115,
      "epoch": 2.431796349663785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004410110032712043,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2228917136788368,
      "num_tokens": 58060505.0,
      "step": 25315
    },
    {
      "entropy": 5.127581024169922,
      "epoch": 2.4322766570605188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004409878602797094,
      "loss": 4.8132,
      "mean_token_accuracy": 0.23894816786050796,
      "num_tokens": 58071926.0,
      "step": 25320
    },
    {
      "entropy": 5.174221658706665,
      "epoch": 2.4327569644572526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044096471343446923,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23890531808137894,
      "num_tokens": 58082963.0,
      "step": 25325
    },
    {
      "entropy": 5.156064558029175,
      "epoch": 2.4332372718539865,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004409415627360213,
      "loss": 4.8488,
      "mean_token_accuracy": 0.23025713115930557,
      "num_tokens": 58095343.0,
      "step": 25330
    },
    {
      "entropy": 5.0660255432128904,
      "epoch": 2.4337175792507204,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044091840818490303,
      "loss": 4.7611,
      "mean_token_accuracy": 0.23841883093118668,
      "num_tokens": 58106763.0,
      "step": 25335
    },
    {
      "entropy": 5.144422006607056,
      "epoch": 2.4341978866474543,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044089524978165197,
      "loss": 4.767,
      "mean_token_accuracy": 0.23548780530691146,
      "num_tokens": 58116942.0,
      "step": 25340
    },
    {
      "entropy": 5.0666886329650875,
      "epoch": 2.434678194044188,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044087208752680577,
      "loss": 4.7437,
      "mean_token_accuracy": 0.23562292754650116,
      "num_tokens": 58127365.0,
      "step": 25345
    },
    {
      "entropy": 5.134285926818848,
      "epoch": 2.435158501440922,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004408489214209023,
      "loss": 4.8753,
      "mean_token_accuracy": 0.22586984634399415,
      "num_tokens": 58137572.0,
      "step": 25350
    },
    {
      "entropy": 5.231930828094482,
      "epoch": 2.435638808837656,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004408257514644793,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22125782668590546,
      "num_tokens": 58150012.0,
      "step": 25355
    },
    {
      "entropy": 5.078060007095337,
      "epoch": 2.43611911623439,
      "grad_norm": 1.0,
      "learning_rate": 0.00044080257765807476,
      "loss": 4.6749,
      "mean_token_accuracy": 0.2422065705060959,
      "num_tokens": 58160611.0,
      "step": 25360
    },
    {
      "entropy": 5.218013620376587,
      "epoch": 2.436599423631124,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004407794000022267,
      "loss": 4.959,
      "mean_token_accuracy": 0.2184649258852005,
      "num_tokens": 58172113.0,
      "step": 25365
    },
    {
      "entropy": 5.187936305999756,
      "epoch": 2.437079731027858,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004407562184974732,
      "loss": 4.8477,
      "mean_token_accuracy": 0.22827996611595153,
      "num_tokens": 58183118.0,
      "step": 25370
    },
    {
      "entropy": 5.112111043930054,
      "epoch": 2.437560038424592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004407330331443526,
      "loss": 4.8427,
      "mean_token_accuracy": 0.22580569833517075,
      "num_tokens": 58194534.0,
      "step": 25375
    },
    {
      "entropy": 5.1823060512542725,
      "epoch": 2.438040345821326,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000440709843943403,
      "loss": 4.883,
      "mean_token_accuracy": 0.22593903094530104,
      "num_tokens": 58205952.0,
      "step": 25380
    },
    {
      "entropy": 5.172022724151612,
      "epoch": 2.4385206532180597,
      "grad_norm": 1.03125,
      "learning_rate": 0.000440686650895163,
      "loss": 4.8145,
      "mean_token_accuracy": 0.2260005757212639,
      "num_tokens": 58217811.0,
      "step": 25385
    },
    {
      "entropy": 5.230978536605835,
      "epoch": 2.4390009606147935,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044066345400017084,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22314143925905228,
      "num_tokens": 58228658.0,
      "step": 25390
    },
    {
      "entropy": 5.1500050067901615,
      "epoch": 2.4394812680115274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044064025325896524,
      "loss": 4.8972,
      "mean_token_accuracy": 0.2250830873847008,
      "num_tokens": 58242130.0,
      "step": 25395
    },
    {
      "entropy": 5.166266012191772,
      "epoch": 2.4399615754082613,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044061704867208484,
      "loss": 4.8297,
      "mean_token_accuracy": 0.22895791083574296,
      "num_tokens": 58253533.0,
      "step": 25400
    },
    {
      "entropy": 5.174663877487182,
      "epoch": 2.440441882804995,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044059384024006825,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22701561450958252,
      "num_tokens": 58264381.0,
      "step": 25405
    },
    {
      "entropy": 5.145359659194947,
      "epoch": 2.440922190201729,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004405706279634545,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22682986557483673,
      "num_tokens": 58275114.0,
      "step": 25410
    },
    {
      "entropy": 5.103869819641114,
      "epoch": 2.441402497598463,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044054741184278243,
      "loss": 4.7759,
      "mean_token_accuracy": 0.23141957968473434,
      "num_tokens": 58287349.0,
      "step": 25415
    },
    {
      "entropy": 5.121177244186401,
      "epoch": 2.441882804995197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044052419187859095,
      "loss": 4.7969,
      "mean_token_accuracy": 0.23908806145191192,
      "num_tokens": 58298787.0,
      "step": 25420
    },
    {
      "entropy": 5.130848932266235,
      "epoch": 2.4423631123919307,
      "grad_norm": 1.125,
      "learning_rate": 0.0004405009680714193,
      "loss": 4.9036,
      "mean_token_accuracy": 0.222383052110672,
      "num_tokens": 58311019.0,
      "step": 25425
    },
    {
      "entropy": 5.103860855102539,
      "epoch": 2.4428434197886646,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004404777404218065,
      "loss": 4.816,
      "mean_token_accuracy": 0.23279712349176407,
      "num_tokens": 58322692.0,
      "step": 25430
    },
    {
      "entropy": 5.169567394256592,
      "epoch": 2.4433237271853985,
      "grad_norm": 1.03125,
      "learning_rate": 0.000440454508930292,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22787571996450423,
      "num_tokens": 58333638.0,
      "step": 25435
    },
    {
      "entropy": 5.141612815856933,
      "epoch": 2.443804034582133,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004404312735974152,
      "loss": 4.8478,
      "mean_token_accuracy": 0.2263951763510704,
      "num_tokens": 58345291.0,
      "step": 25440
    },
    {
      "entropy": 5.205932092666626,
      "epoch": 2.4442843419788662,
      "grad_norm": 1.0,
      "learning_rate": 0.00044040803442371533,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22903018146753312,
      "num_tokens": 58357303.0,
      "step": 25445
    },
    {
      "entropy": 5.1495006561279295,
      "epoch": 2.4447646493756006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004403847914097321,
      "loss": 4.8307,
      "mean_token_accuracy": 0.2296408921480179,
      "num_tokens": 58369804.0,
      "step": 25450
    },
    {
      "entropy": 5.129746341705323,
      "epoch": 2.4452449567723344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044036154455600517,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22930045127868653,
      "num_tokens": 58381027.0,
      "step": 25455
    },
    {
      "entropy": 5.130937814712524,
      "epoch": 2.4457252641690683,
      "grad_norm": 1.0,
      "learning_rate": 0.0004403382938630741,
      "loss": 4.8063,
      "mean_token_accuracy": 0.2266253799200058,
      "num_tokens": 58392165.0,
      "step": 25460
    },
    {
      "entropy": 5.162167644500732,
      "epoch": 2.446205571565802,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044031503933147887,
      "loss": 4.8382,
      "mean_token_accuracy": 0.22998449951410294,
      "num_tokens": 58402494.0,
      "step": 25465
    },
    {
      "entropy": 5.087490844726562,
      "epoch": 2.446685878962536,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044029178096175934,
      "loss": 4.7754,
      "mean_token_accuracy": 0.23356178104877473,
      "num_tokens": 58413735.0,
      "step": 25470
    },
    {
      "entropy": 5.171435070037842,
      "epoch": 2.44716618635927,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004402685187544554,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22690863609313966,
      "num_tokens": 58425977.0,
      "step": 25475
    },
    {
      "entropy": 5.161852264404297,
      "epoch": 2.447646493756004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004402452527101072,
      "loss": 4.8258,
      "mean_token_accuracy": 0.23154951483011246,
      "num_tokens": 58436845.0,
      "step": 25480
    },
    {
      "entropy": 5.148421716690064,
      "epoch": 2.4481268011527377,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004402219828292549,
      "loss": 4.7813,
      "mean_token_accuracy": 0.2364364430308342,
      "num_tokens": 58448113.0,
      "step": 25485
    },
    {
      "entropy": 5.211902713775634,
      "epoch": 2.4486071085494716,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004401987091124388,
      "loss": 4.8469,
      "mean_token_accuracy": 0.2206164538860321,
      "num_tokens": 58459226.0,
      "step": 25490
    },
    {
      "entropy": 5.209065198898315,
      "epoch": 2.4490874159462055,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004401754315601992,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22686037570238113,
      "num_tokens": 58471354.0,
      "step": 25495
    },
    {
      "entropy": 5.202803039550782,
      "epoch": 2.4495677233429394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004401521501730765,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23199114948511124,
      "num_tokens": 58482859.0,
      "step": 25500
    },
    {
      "entropy": 5.219443702697754,
      "epoch": 2.4500480307396733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044012886495161144,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21931190192699432,
      "num_tokens": 58496351.0,
      "step": 25505
    },
    {
      "entropy": 5.201704359054565,
      "epoch": 2.450528338136407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004401055758963443,
      "loss": 4.8748,
      "mean_token_accuracy": 0.23221035152673722,
      "num_tokens": 58508055.0,
      "step": 25510
    },
    {
      "entropy": 5.091471529006958,
      "epoch": 2.451008645533141,
      "grad_norm": 1.0,
      "learning_rate": 0.000440082283007816,
      "loss": 4.8232,
      "mean_token_accuracy": 0.23124151974916457,
      "num_tokens": 58519344.0,
      "step": 25515
    },
    {
      "entropy": 5.243907499313354,
      "epoch": 2.451488952929875,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044005898628656734,
      "loss": 4.922,
      "mean_token_accuracy": 0.22056668251752853,
      "num_tokens": 58530141.0,
      "step": 25520
    },
    {
      "entropy": 5.225683784484863,
      "epoch": 2.4519692603266092,
      "grad_norm": 1.078125,
      "learning_rate": 0.000440035685733139,
      "loss": 4.8287,
      "mean_token_accuracy": 0.23096541464328765,
      "num_tokens": 58541527.0,
      "step": 25525
    },
    {
      "entropy": 5.245523977279663,
      "epoch": 2.452449567723343,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004400123813480722,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2287449687719345,
      "num_tokens": 58552205.0,
      "step": 25530
    },
    {
      "entropy": 5.1628148555755615,
      "epoch": 2.452929875120077,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043998907313190787,
      "loss": 4.7843,
      "mean_token_accuracy": 0.23279560804367067,
      "num_tokens": 58563563.0,
      "step": 25535
    },
    {
      "entropy": 5.134886455535889,
      "epoch": 2.453410182516811,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004399657610851873,
      "loss": 4.8418,
      "mean_token_accuracy": 0.22083672732114792,
      "num_tokens": 58575605.0,
      "step": 25540
    },
    {
      "entropy": 5.219349431991577,
      "epoch": 2.4538904899135447,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043994244520845146,
      "loss": 4.9572,
      "mean_token_accuracy": 0.22009041905403137,
      "num_tokens": 58587761.0,
      "step": 25545
    },
    {
      "entropy": 5.277775669097901,
      "epoch": 2.4543707973102786,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004399191255022418,
      "loss": 4.7977,
      "mean_token_accuracy": 0.2291880190372467,
      "num_tokens": 58598541.0,
      "step": 25550
    },
    {
      "entropy": 5.125376462936401,
      "epoch": 2.4548511047070125,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004398958019670998,
      "loss": 4.8361,
      "mean_token_accuracy": 0.23202893435955046,
      "num_tokens": 58610833.0,
      "step": 25555
    },
    {
      "entropy": 5.171410274505615,
      "epoch": 2.4553314121037464,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043987247460356696,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22461301684379578,
      "num_tokens": 58620857.0,
      "step": 25560
    },
    {
      "entropy": 5.124122762680054,
      "epoch": 2.4558117195004803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004398491434121847,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23651879131793976,
      "num_tokens": 58631245.0,
      "step": 25565
    },
    {
      "entropy": 5.146027088165283,
      "epoch": 2.456292026897214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000439825808393495,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23377207964658736,
      "num_tokens": 58642995.0,
      "step": 25570
    },
    {
      "entropy": 5.125596046447754,
      "epoch": 2.456772334293948,
      "grad_norm": 1.125,
      "learning_rate": 0.0004398024695480394,
      "loss": 4.7786,
      "mean_token_accuracy": 0.23809215277433396,
      "num_tokens": 58654817.0,
      "step": 25575
    },
    {
      "entropy": 5.17879548072815,
      "epoch": 2.457252641690682,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004397791268763598,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23346019685268402,
      "num_tokens": 58665912.0,
      "step": 25580
    },
    {
      "entropy": 5.149789094924927,
      "epoch": 2.457732949087416,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043975578037899814,
      "loss": 4.8371,
      "mean_token_accuracy": 0.22949791103601455,
      "num_tokens": 58678080.0,
      "step": 25585
    },
    {
      "entropy": 5.184147644042969,
      "epoch": 2.4582132564841497,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004397324300564966,
      "loss": 4.7931,
      "mean_token_accuracy": 0.23076532930135726,
      "num_tokens": 58689902.0,
      "step": 25590
    },
    {
      "entropy": 5.1674305438995365,
      "epoch": 2.4586935638808836,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004397090759093971,
      "loss": 4.799,
      "mean_token_accuracy": 0.23421066105365754,
      "num_tokens": 58701857.0,
      "step": 25595
    },
    {
      "entropy": 5.143260431289673,
      "epoch": 2.459173871277618,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043968571793824194,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22512925416231155,
      "num_tokens": 58712775.0,
      "step": 25600
    },
    {
      "entropy": 5.156573820114136,
      "epoch": 2.4596541786743518,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004396623561435734,
      "loss": 4.7876,
      "mean_token_accuracy": 0.23024412393569946,
      "num_tokens": 58723583.0,
      "step": 25605
    },
    {
      "entropy": 5.190837430953979,
      "epoch": 2.4601344860710856,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004396389905259339,
      "loss": 4.8987,
      "mean_token_accuracy": 0.2300587311387062,
      "num_tokens": 58734187.0,
      "step": 25610
    },
    {
      "entropy": 5.1675457000732425,
      "epoch": 2.4606147934678195,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00043961562108586603,
      "loss": 4.8377,
      "mean_token_accuracy": 0.22763992547988893,
      "num_tokens": 58746138.0,
      "step": 25615
    },
    {
      "entropy": 5.212983465194702,
      "epoch": 2.4610951008645534,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043959224782391215,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22069852203130721,
      "num_tokens": 58757765.0,
      "step": 25620
    },
    {
      "entropy": 5.324716567993164,
      "epoch": 2.4615754082612873,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000439568870740615,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21120384186506272,
      "num_tokens": 58769957.0,
      "step": 25625
    },
    {
      "entropy": 5.316012287139893,
      "epoch": 2.462055715658021,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004395454898365174,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22655819058418275,
      "num_tokens": 58781019.0,
      "step": 25630
    },
    {
      "entropy": 5.153783178329467,
      "epoch": 2.462536023054755,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043952210511216205,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22483428865671157,
      "num_tokens": 58793427.0,
      "step": 25635
    },
    {
      "entropy": 5.202746820449829,
      "epoch": 2.463016330451489,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043949871656809205,
      "loss": 4.923,
      "mean_token_accuracy": 0.22232692539691926,
      "num_tokens": 58804406.0,
      "step": 25640
    },
    {
      "entropy": 5.217516994476318,
      "epoch": 2.463496637848223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043947532420485024,
      "loss": 4.8373,
      "mean_token_accuracy": 0.23095616400241853,
      "num_tokens": 58815783.0,
      "step": 25645
    },
    {
      "entropy": 5.185921669006348,
      "epoch": 2.4639769452449567,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004394519280229798,
      "loss": 4.9534,
      "mean_token_accuracy": 0.22906674742698668,
      "num_tokens": 58826968.0,
      "step": 25650
    },
    {
      "entropy": 5.177308130264282,
      "epoch": 2.4644572526416906,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043942852802302397,
      "loss": 4.8151,
      "mean_token_accuracy": 0.2277105540037155,
      "num_tokens": 58839090.0,
      "step": 25655
    },
    {
      "entropy": 5.15227370262146,
      "epoch": 2.4649375600384245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004394051242055259,
      "loss": 4.7919,
      "mean_token_accuracy": 0.22970083504915237,
      "num_tokens": 58851636.0,
      "step": 25660
    },
    {
      "entropy": 5.20171217918396,
      "epoch": 2.4654178674351583,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004393817165710291,
      "loss": 4.8905,
      "mean_token_accuracy": 0.23236202597618102,
      "num_tokens": 58863777.0,
      "step": 25665
    },
    {
      "entropy": 5.17203369140625,
      "epoch": 2.465898174831892,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043935830512007687,
      "loss": 4.8579,
      "mean_token_accuracy": 0.22548525184392929,
      "num_tokens": 58875454.0,
      "step": 25670
    },
    {
      "entropy": 5.1638326168060305,
      "epoch": 2.4663784822286265,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00043933488985321286,
      "loss": 4.8143,
      "mean_token_accuracy": 0.2344038352370262,
      "num_tokens": 58887932.0,
      "step": 25675
    },
    {
      "entropy": 5.148482131958008,
      "epoch": 2.46685878962536,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004393114707709807,
      "loss": 4.8453,
      "mean_token_accuracy": 0.2225457414984703,
      "num_tokens": 58900167.0,
      "step": 25680
    },
    {
      "entropy": 5.108645629882813,
      "epoch": 2.4673390970220943,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004392880478739241,
      "loss": 4.7933,
      "mean_token_accuracy": 0.23522593975067138,
      "num_tokens": 58911769.0,
      "step": 25685
    },
    {
      "entropy": 5.1371378898620605,
      "epoch": 2.467819404418828,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004392646211625869,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23506252467632294,
      "num_tokens": 58923747.0,
      "step": 25690
    },
    {
      "entropy": 5.218261575698852,
      "epoch": 2.468299711815562,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004392411906375129,
      "loss": 4.8783,
      "mean_token_accuracy": 0.22780343741178513,
      "num_tokens": 58934801.0,
      "step": 25695
    },
    {
      "entropy": 5.1515491008758545,
      "epoch": 2.468780019212296,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043921775629924615,
      "loss": 4.862,
      "mean_token_accuracy": 0.23388772308826447,
      "num_tokens": 58946566.0,
      "step": 25700
    },
    {
      "entropy": 5.124656629562378,
      "epoch": 2.46926032660903,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043919431814833077,
      "loss": 4.7878,
      "mean_token_accuracy": 0.2369292214512825,
      "num_tokens": 58957066.0,
      "step": 25705
    },
    {
      "entropy": 5.118686532974243,
      "epoch": 2.4697406340057637,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043917087618531084,
      "loss": 4.7407,
      "mean_token_accuracy": 0.2336021363735199,
      "num_tokens": 58967553.0,
      "step": 25710
    },
    {
      "entropy": 5.15559344291687,
      "epoch": 2.4702209414024976,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004391474304107307,
      "loss": 4.8565,
      "mean_token_accuracy": 0.2323279619216919,
      "num_tokens": 58979038.0,
      "step": 25715
    },
    {
      "entropy": 5.1457830429077145,
      "epoch": 2.4707012487992315,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043912398082513463,
      "loss": 4.8041,
      "mean_token_accuracy": 0.22675961256027222,
      "num_tokens": 58989564.0,
      "step": 25720
    },
    {
      "entropy": 5.124704027175904,
      "epoch": 2.4711815561959654,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000439100527429067,
      "loss": 4.8534,
      "mean_token_accuracy": 0.23475688844919204,
      "num_tokens": 59001789.0,
      "step": 25725
    },
    {
      "entropy": 5.1620948791503904,
      "epoch": 2.4716618635926992,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043907707022307243,
      "loss": 4.8703,
      "mean_token_accuracy": 0.23116497248411177,
      "num_tokens": 59013518.0,
      "step": 25730
    },
    {
      "entropy": 5.10627965927124,
      "epoch": 2.472142170989433,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043905360920769553,
      "loss": 4.7787,
      "mean_token_accuracy": 0.2364992991089821,
      "num_tokens": 59024739.0,
      "step": 25735
    },
    {
      "entropy": 5.148125171661377,
      "epoch": 2.472622478386167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000439030144383481,
      "loss": 4.7971,
      "mean_token_accuracy": 0.23215530812740326,
      "num_tokens": 59036142.0,
      "step": 25740
    },
    {
      "entropy": 5.198448133468628,
      "epoch": 2.473102785782901,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043900667575097355,
      "loss": 4.8801,
      "mean_token_accuracy": 0.23001099079847337,
      "num_tokens": 59046675.0,
      "step": 25745
    },
    {
      "entropy": 5.157300281524658,
      "epoch": 2.473583093179635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004389832033107181,
      "loss": 4.8016,
      "mean_token_accuracy": 0.23102212697267532,
      "num_tokens": 59057193.0,
      "step": 25750
    },
    {
      "entropy": 5.1085583686828615,
      "epoch": 2.4740634005763686,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043895972706325953,
      "loss": 4.8182,
      "mean_token_accuracy": 0.22804644107818603,
      "num_tokens": 59069501.0,
      "step": 25755
    },
    {
      "entropy": 5.226031541824341,
      "epoch": 2.474543707973103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000438936247009143,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21831804215908052,
      "num_tokens": 59080973.0,
      "step": 25760
    },
    {
      "entropy": 5.1503297805786135,
      "epoch": 2.475024015369837,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043891276314891365,
      "loss": 4.8237,
      "mean_token_accuracy": 0.22746793627738954,
      "num_tokens": 59092541.0,
      "step": 25765
    },
    {
      "entropy": 5.103377914428711,
      "epoch": 2.4755043227665707,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004388892754831166,
      "loss": 4.8082,
      "mean_token_accuracy": 0.22803625017404555,
      "num_tokens": 59103936.0,
      "step": 25770
    },
    {
      "entropy": 5.106954956054688,
      "epoch": 2.4759846301633046,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004388657840122973,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23232102543115615,
      "num_tokens": 59114714.0,
      "step": 25775
    },
    {
      "entropy": 5.201672458648682,
      "epoch": 2.4764649375600385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000438842288737001,
      "loss": 4.8498,
      "mean_token_accuracy": 0.22411305010318755,
      "num_tokens": 59126678.0,
      "step": 25780
    },
    {
      "entropy": 5.171486186981201,
      "epoch": 2.4769452449567724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043881878965777325,
      "loss": 4.8509,
      "mean_token_accuracy": 0.23180041760206221,
      "num_tokens": 59137732.0,
      "step": 25785
    },
    {
      "entropy": 5.112175512313843,
      "epoch": 2.4774255523535063,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043879528677515973,
      "loss": 4.8159,
      "mean_token_accuracy": 0.22645105570554733,
      "num_tokens": 59149231.0,
      "step": 25790
    },
    {
      "entropy": 5.256683015823365,
      "epoch": 2.47790585975024,
      "grad_norm": 1.0,
      "learning_rate": 0.00043877178008970596,
      "loss": 4.9475,
      "mean_token_accuracy": 0.21218567937612534,
      "num_tokens": 59161033.0,
      "step": 25795
    },
    {
      "entropy": 5.122001266479492,
      "epoch": 2.478386167146974,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004387482696019578,
      "loss": 4.7434,
      "mean_token_accuracy": 0.24140208065509797,
      "num_tokens": 59171906.0,
      "step": 25800
    },
    {
      "entropy": 5.184522199630737,
      "epoch": 2.478866474543708,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043872475531246105,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22715393006801604,
      "num_tokens": 59185036.0,
      "step": 25805
    },
    {
      "entropy": 5.186851072311401,
      "epoch": 2.479346781940442,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043870123722176166,
      "loss": 4.911,
      "mean_token_accuracy": 0.22600208818912507,
      "num_tokens": 59196397.0,
      "step": 25810
    },
    {
      "entropy": 5.203660774230957,
      "epoch": 2.4798270893371757,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004386777153304056,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2253268450498581,
      "num_tokens": 59208180.0,
      "step": 25815
    },
    {
      "entropy": 5.305365991592407,
      "epoch": 2.4803073967339095,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043865418963893896,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21980006843805314,
      "num_tokens": 59219651.0,
      "step": 25820
    },
    {
      "entropy": 5.073846244812012,
      "epoch": 2.4807877041306434,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004386306601479081,
      "loss": 4.7377,
      "mean_token_accuracy": 0.23738843351602554,
      "num_tokens": 59231327.0,
      "step": 25825
    },
    {
      "entropy": 5.104054880142212,
      "epoch": 2.4812680115273773,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004386071268578591,
      "loss": 4.8485,
      "mean_token_accuracy": 0.22758934944868087,
      "num_tokens": 59243606.0,
      "step": 25830
    },
    {
      "entropy": 5.216682767868042,
      "epoch": 2.4817483189241116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043858358976933844,
      "loss": 4.8783,
      "mean_token_accuracy": 0.2246907651424408,
      "num_tokens": 59255824.0,
      "step": 25835
    },
    {
      "entropy": 5.152724075317383,
      "epoch": 2.4822286263208455,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043856004888289264,
      "loss": 4.7723,
      "mean_token_accuracy": 0.23192077428102492,
      "num_tokens": 59267405.0,
      "step": 25840
    },
    {
      "entropy": 5.158254861831665,
      "epoch": 2.4827089337175794,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004385365041990681,
      "loss": 4.8596,
      "mean_token_accuracy": 0.22911245077848436,
      "num_tokens": 59279063.0,
      "step": 25845
    },
    {
      "entropy": 4.982639169692993,
      "epoch": 2.4831892411143133,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004385129557184116,
      "loss": 4.622,
      "mean_token_accuracy": 0.2503858581185341,
      "num_tokens": 59289838.0,
      "step": 25850
    },
    {
      "entropy": 5.092452812194824,
      "epoch": 2.483669548511047,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043848940344146976,
      "loss": 4.7555,
      "mean_token_accuracy": 0.23404320776462556,
      "num_tokens": 59301505.0,
      "step": 25855
    },
    {
      "entropy": 5.049518394470215,
      "epoch": 2.484149855907781,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004384658473687894,
      "loss": 4.7322,
      "mean_token_accuracy": 0.23690066188573838,
      "num_tokens": 59312934.0,
      "step": 25860
    },
    {
      "entropy": 5.129816722869873,
      "epoch": 2.484630163304515,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004384422875009176,
      "loss": 4.8156,
      "mean_token_accuracy": 0.23358032405376433,
      "num_tokens": 59324104.0,
      "step": 25865
    },
    {
      "entropy": 5.177162981033325,
      "epoch": 2.485110470701249,
      "grad_norm": 1.0,
      "learning_rate": 0.0004384187238384011,
      "loss": 4.8515,
      "mean_token_accuracy": 0.22914069443941115,
      "num_tokens": 59334978.0,
      "step": 25870
    },
    {
      "entropy": 5.069106483459473,
      "epoch": 2.4855907780979827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004383951563817871,
      "loss": 4.7832,
      "mean_token_accuracy": 0.23123769462108612,
      "num_tokens": 59345779.0,
      "step": 25875
    },
    {
      "entropy": 5.216180467605591,
      "epoch": 2.4860710854947166,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004383715851316227,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22725493162870408,
      "num_tokens": 59358870.0,
      "step": 25880
    },
    {
      "entropy": 5.284058570861816,
      "epoch": 2.4865513928914504,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043834801008845527,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22689099311828614,
      "num_tokens": 59369831.0,
      "step": 25885
    },
    {
      "entropy": 5.116775274276733,
      "epoch": 2.4870317002881843,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004383244312528321,
      "loss": 4.7426,
      "mean_token_accuracy": 0.23881611973047256,
      "num_tokens": 59381105.0,
      "step": 25890
    },
    {
      "entropy": 5.140939092636108,
      "epoch": 2.487512007684918,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004383008486253006,
      "loss": 4.8337,
      "mean_token_accuracy": 0.22626545578241347,
      "num_tokens": 59391923.0,
      "step": 25895
    },
    {
      "entropy": 5.166287517547607,
      "epoch": 2.487992315081652,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043827726220640827,
      "loss": 4.8176,
      "mean_token_accuracy": 0.23532552123069764,
      "num_tokens": 59402986.0,
      "step": 25900
    },
    {
      "entropy": 5.129156112670898,
      "epoch": 2.488472622478386,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043825367199670274,
      "loss": 4.8271,
      "mean_token_accuracy": 0.23425014466047286,
      "num_tokens": 59413656.0,
      "step": 25905
    },
    {
      "entropy": 5.076101493835449,
      "epoch": 2.4889529298751203,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004382300779967318,
      "loss": 4.7652,
      "mean_token_accuracy": 0.23708924800157546,
      "num_tokens": 59425396.0,
      "step": 25910
    },
    {
      "entropy": 5.207878446578979,
      "epoch": 2.489433237271854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043820648020704303,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2305879443883896,
      "num_tokens": 59436859.0,
      "step": 25915
    },
    {
      "entropy": 5.205628061294556,
      "epoch": 2.489913544668588,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043818287862818444,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22559942603111266,
      "num_tokens": 59448415.0,
      "step": 25920
    },
    {
      "entropy": 5.101398992538452,
      "epoch": 2.490393852065322,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000438159273260704,
      "loss": 4.7346,
      "mean_token_accuracy": 0.23690586537122726,
      "num_tokens": 59459477.0,
      "step": 25925
    },
    {
      "entropy": 5.092023038864136,
      "epoch": 2.490874159462056,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004381356641051497,
      "loss": 4.7685,
      "mean_token_accuracy": 0.24000615179538726,
      "num_tokens": 59471184.0,
      "step": 25930
    },
    {
      "entropy": 5.115890550613403,
      "epoch": 2.4913544668587897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004381120511620697,
      "loss": 4.7854,
      "mean_token_accuracy": 0.23824024051427842,
      "num_tokens": 59482706.0,
      "step": 25935
    },
    {
      "entropy": 5.1309174537658695,
      "epoch": 2.4918347742555236,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043808843443201217,
      "loss": 4.8511,
      "mean_token_accuracy": 0.2269692301750183,
      "num_tokens": 59493822.0,
      "step": 25940
    },
    {
      "entropy": 5.153273439407348,
      "epoch": 2.4923150816522575,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004380648139155255,
      "loss": 4.767,
      "mean_token_accuracy": 0.22716633677482606,
      "num_tokens": 59504577.0,
      "step": 25945
    },
    {
      "entropy": 5.134180784225464,
      "epoch": 2.4927953890489913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004380411896131581,
      "loss": 4.8418,
      "mean_token_accuracy": 0.22312534600496292,
      "num_tokens": 59517395.0,
      "step": 25950
    },
    {
      "entropy": 5.170784282684326,
      "epoch": 2.493275696445725,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043801756152545836,
      "loss": 4.8356,
      "mean_token_accuracy": 0.23404995501041412,
      "num_tokens": 59529270.0,
      "step": 25955
    },
    {
      "entropy": 5.20110559463501,
      "epoch": 2.493756003842459,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043799392965297496,
      "loss": 4.842,
      "mean_token_accuracy": 0.2237558513879776,
      "num_tokens": 59540866.0,
      "step": 25960
    },
    {
      "entropy": 5.2281773567199705,
      "epoch": 2.494236311239193,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004379702939962564,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22138237059116364,
      "num_tokens": 59553201.0,
      "step": 25965
    },
    {
      "entropy": 5.094724225997925,
      "epoch": 2.494716618635927,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004379466545558516,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23468267321586608,
      "num_tokens": 59563054.0,
      "step": 25970
    },
    {
      "entropy": 5.069637870788574,
      "epoch": 2.4951969260326607,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043792301133230933,
      "loss": 4.7596,
      "mean_token_accuracy": 0.23659998774528504,
      "num_tokens": 59574849.0,
      "step": 25975
    },
    {
      "entropy": 5.0768898010253904,
      "epoch": 2.4956772334293946,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004378993643261785,
      "loss": 4.7124,
      "mean_token_accuracy": 0.23979566097259522,
      "num_tokens": 59585567.0,
      "step": 25980
    },
    {
      "entropy": 5.125411128997802,
      "epoch": 2.496157540826129,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043787571353800814,
      "loss": 4.822,
      "mean_token_accuracy": 0.23586249500513076,
      "num_tokens": 59596810.0,
      "step": 25985
    },
    {
      "entropy": 5.100441980361938,
      "epoch": 2.4966378482228624,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004378520589683475,
      "loss": 4.8155,
      "mean_token_accuracy": 0.23190304040908813,
      "num_tokens": 59608704.0,
      "step": 25990
    },
    {
      "entropy": 5.080446672439575,
      "epoch": 2.4971181556195967,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043782840061774544,
      "loss": 4.7398,
      "mean_token_accuracy": 0.2458975687623024,
      "num_tokens": 59619325.0,
      "step": 25995
    },
    {
      "entropy": 5.183133268356324,
      "epoch": 2.4975984630163306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043780473848675143,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22434959560632706,
      "num_tokens": 59630181.0,
      "step": 26000
    },
    {
      "entropy": 5.141420793533325,
      "epoch": 2.4980787704130645,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004377810725759149,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22240401953458785,
      "num_tokens": 59642722.0,
      "step": 26005
    },
    {
      "entropy": 5.19269118309021,
      "epoch": 2.4985590778097984,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043775740288578516,
      "loss": 4.8442,
      "mean_token_accuracy": 0.22953615337610245,
      "num_tokens": 59653675.0,
      "step": 26010
    },
    {
      "entropy": 5.253700065612793,
      "epoch": 2.4990393852065322,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004377337294169118,
      "loss": 4.933,
      "mean_token_accuracy": 0.2214443415403366,
      "num_tokens": 59664897.0,
      "step": 26015
    },
    {
      "entropy": 5.1191002368927006,
      "epoch": 2.499519692603266,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043771005216984457,
      "loss": 4.8047,
      "mean_token_accuracy": 0.2336744859814644,
      "num_tokens": 59676554.0,
      "step": 26020
    },
    {
      "entropy": 5.187154102325439,
      "epoch": 2.5,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000437686371145133,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22486110627651215,
      "num_tokens": 59688030.0,
      "step": 26025
    },
    {
      "entropy": 5.12536997795105,
      "epoch": 2.500480307396734,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000437662686343327,
      "loss": 4.8134,
      "mean_token_accuracy": 0.2282300591468811,
      "num_tokens": 59699721.0,
      "step": 26030
    },
    {
      "entropy": 5.109609842300415,
      "epoch": 2.5009606147934678,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004376389977649764,
      "loss": 4.789,
      "mean_token_accuracy": 0.2274756610393524,
      "num_tokens": 59710426.0,
      "step": 26035
    },
    {
      "entropy": 5.069755363464355,
      "epoch": 2.5014409221902016,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004376153054106313,
      "loss": 4.7663,
      "mean_token_accuracy": 0.23743103444576263,
      "num_tokens": 59720837.0,
      "step": 26040
    },
    {
      "entropy": 5.149607992172241,
      "epoch": 2.5019212295869355,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004375916092808416,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22865049690008163,
      "num_tokens": 59732045.0,
      "step": 26045
    },
    {
      "entropy": 5.166457223892212,
      "epoch": 2.5024015369836694,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004375679093761575,
      "loss": 4.8547,
      "mean_token_accuracy": 0.22209447473287583,
      "num_tokens": 59744735.0,
      "step": 26050
    },
    {
      "entropy": 5.154904985427857,
      "epoch": 2.5028818443804033,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043754420569712925,
      "loss": 4.7656,
      "mean_token_accuracy": 0.23750779330730437,
      "num_tokens": 59755604.0,
      "step": 26055
    },
    {
      "entropy": 5.078918743133545,
      "epoch": 2.5033621517771376,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043752049824430736,
      "loss": 4.8031,
      "mean_token_accuracy": 0.23734070658683776,
      "num_tokens": 59765892.0,
      "step": 26060
    },
    {
      "entropy": 5.113362884521484,
      "epoch": 2.503842459173871,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043749678701824197,
      "loss": 4.8307,
      "mean_token_accuracy": 0.225986510515213,
      "num_tokens": 59779068.0,
      "step": 26065
    },
    {
      "entropy": 5.218837308883667,
      "epoch": 2.5043227665706054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004374730720194837,
      "loss": 4.8795,
      "mean_token_accuracy": 0.23196474611759185,
      "num_tokens": 59789894.0,
      "step": 26070
    },
    {
      "entropy": 5.070636653900147,
      "epoch": 2.5048030739673393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004374493532485832,
      "loss": 4.7847,
      "mean_token_accuracy": 0.24057650119066237,
      "num_tokens": 59801384.0,
      "step": 26075
    },
    {
      "entropy": 5.118451070785523,
      "epoch": 2.505283381364073,
      "grad_norm": 1.015625,
      "learning_rate": 0.000437425630706091,
      "loss": 4.8119,
      "mean_token_accuracy": 0.2307392790913582,
      "num_tokens": 59813471.0,
      "step": 26080
    },
    {
      "entropy": 5.241332864761352,
      "epoch": 2.505763688760807,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000437401904392558,
      "loss": 4.901,
      "mean_token_accuracy": 0.22997792959213256,
      "num_tokens": 59824495.0,
      "step": 26085
    },
    {
      "entropy": 5.238813018798828,
      "epoch": 2.506243996157541,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043737817430853504,
      "loss": 4.9097,
      "mean_token_accuracy": 0.2185683086514473,
      "num_tokens": 59835588.0,
      "step": 26090
    },
    {
      "entropy": 5.178942394256592,
      "epoch": 2.506724303554275,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043735444045457303,
      "loss": 4.87,
      "mean_token_accuracy": 0.22750694155693055,
      "num_tokens": 59847093.0,
      "step": 26095
    },
    {
      "entropy": 5.107070636749268,
      "epoch": 2.5072046109510087,
      "grad_norm": 1.125,
      "learning_rate": 0.00043733070283122306,
      "loss": 4.7475,
      "mean_token_accuracy": 0.23755243718624114,
      "num_tokens": 59858109.0,
      "step": 26100
    },
    {
      "entropy": 5.102501153945923,
      "epoch": 2.5076849183477425,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043730696143903607,
      "loss": 4.7811,
      "mean_token_accuracy": 0.22890698164701462,
      "num_tokens": 59870161.0,
      "step": 26105
    },
    {
      "entropy": 5.1666699886322025,
      "epoch": 2.5081652257444764,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004372832162785635,
      "loss": 4.8353,
      "mean_token_accuracy": 0.2319941535592079,
      "num_tokens": 59882068.0,
      "step": 26110
    },
    {
      "entropy": 5.181774711608886,
      "epoch": 2.5086455331412103,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004372594673503565,
      "loss": 4.7883,
      "mean_token_accuracy": 0.23984845876693725,
      "num_tokens": 59892418.0,
      "step": 26115
    },
    {
      "entropy": 5.141549015045166,
      "epoch": 2.509125840537944,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004372357146549664,
      "loss": 4.8531,
      "mean_token_accuracy": 0.23280137926340103,
      "num_tokens": 59902790.0,
      "step": 26120
    },
    {
      "entropy": 5.11729063987732,
      "epoch": 2.509606147934678,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043721195819294487,
      "loss": 4.8149,
      "mean_token_accuracy": 0.23124950975179673,
      "num_tokens": 59914982.0,
      "step": 26125
    },
    {
      "entropy": 5.187640380859375,
      "epoch": 2.510086455331412,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004371881979648433,
      "loss": 4.9358,
      "mean_token_accuracy": 0.21992879509925842,
      "num_tokens": 59925961.0,
      "step": 26130
    },
    {
      "entropy": 5.170394849777222,
      "epoch": 2.5105667627281463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004371644339712133,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22804148942232133,
      "num_tokens": 59938709.0,
      "step": 26135
    },
    {
      "entropy": 5.168030500411987,
      "epoch": 2.5110470701248797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004371406662126067,
      "loss": 4.7984,
      "mean_token_accuracy": 0.23032863140106202,
      "num_tokens": 59949981.0,
      "step": 26140
    },
    {
      "entropy": 5.098376321792602,
      "epoch": 2.511527377521614,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043711689468957534,
      "loss": 4.7555,
      "mean_token_accuracy": 0.23787154108285904,
      "num_tokens": 59961741.0,
      "step": 26145
    },
    {
      "entropy": 5.111446189880371,
      "epoch": 2.5120076849183475,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043709311940267107,
      "loss": 4.805,
      "mean_token_accuracy": 0.2280091568827629,
      "num_tokens": 59973399.0,
      "step": 26150
    },
    {
      "entropy": 5.12957501411438,
      "epoch": 2.512487992315082,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004370693403524458,
      "loss": 4.7852,
      "mean_token_accuracy": 0.23764416724443435,
      "num_tokens": 59985180.0,
      "step": 26155
    },
    {
      "entropy": 5.158020639419556,
      "epoch": 2.5129682997118157,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004370455575394518,
      "loss": 4.8066,
      "mean_token_accuracy": 0.23412394374608994,
      "num_tokens": 59995391.0,
      "step": 26160
    },
    {
      "entropy": 5.118181228637695,
      "epoch": 2.5134486071085496,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004370217709642411,
      "loss": 4.8332,
      "mean_token_accuracy": 0.22718574106693268,
      "num_tokens": 60006092.0,
      "step": 26165
    },
    {
      "entropy": 5.214171981811523,
      "epoch": 2.5139289145052834,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000436997980627366,
      "loss": 4.9113,
      "mean_token_accuracy": 0.23072213381528855,
      "num_tokens": 60017615.0,
      "step": 26170
    },
    {
      "entropy": 5.192780065536499,
      "epoch": 2.5144092219020173,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043697418652937877,
      "loss": 4.8314,
      "mean_token_accuracy": 0.22895765900611878,
      "num_tokens": 60030062.0,
      "step": 26175
    },
    {
      "entropy": 5.127854919433593,
      "epoch": 2.514889529298751,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004369503886708319,
      "loss": 4.7982,
      "mean_token_accuracy": 0.23160915672779084,
      "num_tokens": 60042188.0,
      "step": 26180
    },
    {
      "entropy": 5.166474962234497,
      "epoch": 2.515369836695485,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043692658705227796,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23338797837495803,
      "num_tokens": 60054532.0,
      "step": 26185
    },
    {
      "entropy": 5.249168920516968,
      "epoch": 2.515850144092219,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043690278167426945,
      "loss": 4.9675,
      "mean_token_accuracy": 0.22997674494981765,
      "num_tokens": 60066026.0,
      "step": 26190
    },
    {
      "entropy": 5.147393798828125,
      "epoch": 2.516330451488953,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004368789725373591,
      "loss": 4.7635,
      "mean_token_accuracy": 0.23659027367830276,
      "num_tokens": 60079679.0,
      "step": 26195
    },
    {
      "entropy": 5.17294659614563,
      "epoch": 2.5168107588856867,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043685515964209977,
      "loss": 4.8748,
      "mean_token_accuracy": 0.22779053300619126,
      "num_tokens": 60089971.0,
      "step": 26200
    },
    {
      "entropy": 5.106278944015503,
      "epoch": 2.5172910662824206,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004368313429890441,
      "loss": 4.779,
      "mean_token_accuracy": 0.23239662051200866,
      "num_tokens": 60100200.0,
      "step": 26205
    },
    {
      "entropy": 5.147910642623901,
      "epoch": 2.517771373679155,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004368075225787453,
      "loss": 4.8329,
      "mean_token_accuracy": 0.23109904825687408,
      "num_tokens": 60112031.0,
      "step": 26210
    },
    {
      "entropy": 5.169220352172852,
      "epoch": 2.5182516810758884,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004367836984117562,
      "loss": 4.7953,
      "mean_token_accuracy": 0.23918470591306687,
      "num_tokens": 60122233.0,
      "step": 26215
    },
    {
      "entropy": 5.149849700927734,
      "epoch": 2.5187319884726227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043675987048863,
      "loss": 4.7803,
      "mean_token_accuracy": 0.2361700043082237,
      "num_tokens": 60133144.0,
      "step": 26220
    },
    {
      "entropy": 5.100948429107666,
      "epoch": 2.519212295869356,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004367360388099201,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22915498167276382,
      "num_tokens": 60144052.0,
      "step": 26225
    },
    {
      "entropy": 5.20796308517456,
      "epoch": 2.5196926032660905,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004367122033761796,
      "loss": 4.8462,
      "mean_token_accuracy": 0.2271530285477638,
      "num_tokens": 60155333.0,
      "step": 26230
    },
    {
      "entropy": 5.1962498188018795,
      "epoch": 2.5201729106628243,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004366883641879618,
      "loss": 4.8195,
      "mean_token_accuracy": 0.22967554479837418,
      "num_tokens": 60166473.0,
      "step": 26235
    },
    {
      "entropy": 4.986312961578369,
      "epoch": 2.520653218059558,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043666452124582034,
      "loss": 4.6646,
      "mean_token_accuracy": 0.24425591230392457,
      "num_tokens": 60177914.0,
      "step": 26240
    },
    {
      "entropy": 5.115043354034424,
      "epoch": 2.521133525456292,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004366406745503088,
      "loss": 4.8874,
      "mean_token_accuracy": 0.2245262920856476,
      "num_tokens": 60190414.0,
      "step": 26245
    },
    {
      "entropy": 5.280806636810302,
      "epoch": 2.521613832853026,
      "grad_norm": 1.0,
      "learning_rate": 0.0004366168241019807,
      "loss": 5.0015,
      "mean_token_accuracy": 0.21616588681936263,
      "num_tokens": 60202686.0,
      "step": 26250
    },
    {
      "entropy": 5.229509973526001,
      "epoch": 2.52209414024976,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004365929699013899,
      "loss": 4.8285,
      "mean_token_accuracy": 0.22557784616947174,
      "num_tokens": 60214244.0,
      "step": 26255
    },
    {
      "entropy": 5.134967613220215,
      "epoch": 2.5225744476464937,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004365691119490902,
      "loss": 4.7996,
      "mean_token_accuracy": 0.2330076962709427,
      "num_tokens": 60225057.0,
      "step": 26260
    },
    {
      "entropy": 5.161644554138183,
      "epoch": 2.5230547550432276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004365452502456354,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22299009412527085,
      "num_tokens": 60236060.0,
      "step": 26265
    },
    {
      "entropy": 5.171703290939331,
      "epoch": 2.5235350624399615,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004365213847915796,
      "loss": 4.8267,
      "mean_token_accuracy": 0.22928389757871628,
      "num_tokens": 60247493.0,
      "step": 26270
    },
    {
      "entropy": 5.054736185073852,
      "epoch": 2.5240153698366954,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043649751558747695,
      "loss": 4.6916,
      "mean_token_accuracy": 0.23998323678970337,
      "num_tokens": 60258525.0,
      "step": 26275
    },
    {
      "entropy": 5.1592604637146,
      "epoch": 2.5244956772334293,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043647364263388143,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23278416991233825,
      "num_tokens": 60270466.0,
      "step": 26280
    },
    {
      "entropy": 5.149476718902588,
      "epoch": 2.524975984630163,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004364497659313475,
      "loss": 4.8054,
      "mean_token_accuracy": 0.2306404873728752,
      "num_tokens": 60282078.0,
      "step": 26285
    },
    {
      "entropy": 5.168886756896972,
      "epoch": 2.525456292026897,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004364258854804294,
      "loss": 4.8109,
      "mean_token_accuracy": 0.22479586005210878,
      "num_tokens": 60293569.0,
      "step": 26290
    },
    {
      "entropy": 5.079735136032104,
      "epoch": 2.5259365994236314,
      "grad_norm": 1.0,
      "learning_rate": 0.0004364020012816815,
      "loss": 4.7407,
      "mean_token_accuracy": 0.2377133384346962,
      "num_tokens": 60305840.0,
      "step": 26295
    },
    {
      "entropy": 5.122062873840332,
      "epoch": 2.526416906820365,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004363781133356584,
      "loss": 4.8425,
      "mean_token_accuracy": 0.23321947306394578,
      "num_tokens": 60318519.0,
      "step": 26300
    },
    {
      "entropy": 5.042409181594849,
      "epoch": 2.526897214217099,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004363542216429147,
      "loss": 4.7113,
      "mean_token_accuracy": 0.2417600616812706,
      "num_tokens": 60328984.0,
      "step": 26305
    },
    {
      "entropy": 5.230122423171997,
      "epoch": 2.527377521613833,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004363303262040051,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21927153319120407,
      "num_tokens": 60339846.0,
      "step": 26310
    },
    {
      "entropy": 5.143296718597412,
      "epoch": 2.527857829010567,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043630642701948446,
      "loss": 4.8466,
      "mean_token_accuracy": 0.23130675554275512,
      "num_tokens": 60351336.0,
      "step": 26315
    },
    {
      "entropy": 5.058671188354492,
      "epoch": 2.5283381364073008,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043628252408990756,
      "loss": 4.7406,
      "mean_token_accuracy": 0.24147887974977494,
      "num_tokens": 60362794.0,
      "step": 26320
    },
    {
      "entropy": 5.1445146083831785,
      "epoch": 2.5288184438040346,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043625861741582926,
      "loss": 4.8056,
      "mean_token_accuracy": 0.23520873934030534,
      "num_tokens": 60374176.0,
      "step": 26325
    },
    {
      "entropy": 5.079167079925537,
      "epoch": 2.5292987512007685,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043623470699780483,
      "loss": 4.7651,
      "mean_token_accuracy": 0.23610994964838028,
      "num_tokens": 60385414.0,
      "step": 26330
    },
    {
      "entropy": 5.127378225326538,
      "epoch": 2.5297790585975024,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004362107928363892,
      "loss": 4.7902,
      "mean_token_accuracy": 0.2372598797082901,
      "num_tokens": 60397738.0,
      "step": 26335
    },
    {
      "entropy": 5.075834608078003,
      "epoch": 2.5302593659942363,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004361868749321377,
      "loss": 4.7201,
      "mean_token_accuracy": 0.24069612771272658,
      "num_tokens": 60408554.0,
      "step": 26340
    },
    {
      "entropy": 5.071924018859863,
      "epoch": 2.53073967339097,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004361629532856055,
      "loss": 4.7764,
      "mean_token_accuracy": 0.23767312467098237,
      "num_tokens": 60420000.0,
      "step": 26345
    },
    {
      "entropy": 5.164826726913452,
      "epoch": 2.531219980787704,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00043613902789734816,
      "loss": 4.7343,
      "mean_token_accuracy": 0.23705534636974335,
      "num_tokens": 60431639.0,
      "step": 26350
    },
    {
      "entropy": 5.1856852054595945,
      "epoch": 2.531700288184438,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000436115098767921,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2302584692835808,
      "num_tokens": 60442401.0,
      "step": 26355
    },
    {
      "entropy": 5.2056872844696045,
      "epoch": 2.532180595581172,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043609116589787974,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22528564780950547,
      "num_tokens": 60455530.0,
      "step": 26360
    },
    {
      "entropy": 5.241705894470215,
      "epoch": 2.5326609029779057,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004360672292877799,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22274887859821318,
      "num_tokens": 60466380.0,
      "step": 26365
    },
    {
      "entropy": 5.135586786270141,
      "epoch": 2.53314121037464,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043604328893817726,
      "loss": 4.8389,
      "mean_token_accuracy": 0.23628002852201463,
      "num_tokens": 60477309.0,
      "step": 26370
    },
    {
      "entropy": 5.105282354354858,
      "epoch": 2.5336215177713735,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043601934484962775,
      "loss": 4.8658,
      "mean_token_accuracy": 0.2287624955177307,
      "num_tokens": 60489181.0,
      "step": 26375
    },
    {
      "entropy": 5.164829730987549,
      "epoch": 2.534101825168108,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004359953970226871,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23152091056108476,
      "num_tokens": 60501994.0,
      "step": 26380
    },
    {
      "entropy": 5.153964328765869,
      "epoch": 2.5345821325648417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043597144545791134,
      "loss": 4.7408,
      "mean_token_accuracy": 0.2331573948264122,
      "num_tokens": 60512769.0,
      "step": 26385
    },
    {
      "entropy": 5.21726884841919,
      "epoch": 2.5350624399615755,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004359474901558567,
      "loss": 4.8349,
      "mean_token_accuracy": 0.2222766399383545,
      "num_tokens": 60523965.0,
      "step": 26390
    },
    {
      "entropy": 5.116404008865357,
      "epoch": 2.5355427473583094,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004359235311170792,
      "loss": 4.8033,
      "mean_token_accuracy": 0.23426614701747894,
      "num_tokens": 60534892.0,
      "step": 26395
    },
    {
      "entropy": 5.178050947189331,
      "epoch": 2.5360230547550433,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004358995683421352,
      "loss": 4.8241,
      "mean_token_accuracy": 0.2367056503891945,
      "num_tokens": 60545744.0,
      "step": 26400
    },
    {
      "entropy": 5.013965463638305,
      "epoch": 2.536503362151777,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043587560183158095,
      "loss": 4.672,
      "mean_token_accuracy": 0.24330135732889174,
      "num_tokens": 60556830.0,
      "step": 26405
    },
    {
      "entropy": 5.169614362716675,
      "epoch": 2.536983669548511,
      "grad_norm": 1.03125,
      "learning_rate": 0.000435851631585973,
      "loss": 4.8454,
      "mean_token_accuracy": 0.2266107529401779,
      "num_tokens": 60567360.0,
      "step": 26410
    },
    {
      "entropy": 5.1638037204742435,
      "epoch": 2.537463976945245,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004358276576058677,
      "loss": 4.8206,
      "mean_token_accuracy": 0.22829893082380295,
      "num_tokens": 60577915.0,
      "step": 26415
    },
    {
      "entropy": 5.121583652496338,
      "epoch": 2.537944284341979,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004358036798918218,
      "loss": 4.8333,
      "mean_token_accuracy": 0.22700742185115813,
      "num_tokens": 60589160.0,
      "step": 26420
    },
    {
      "entropy": 5.142694044113159,
      "epoch": 2.5384245917387127,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004357796984443919,
      "loss": 4.8381,
      "mean_token_accuracy": 0.23110271394252777,
      "num_tokens": 60599573.0,
      "step": 26425
    },
    {
      "entropy": 5.165138626098633,
      "epoch": 2.5389048991354466,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043575571326413484,
      "loss": 4.8206,
      "mean_token_accuracy": 0.23648686110973358,
      "num_tokens": 60610886.0,
      "step": 26430
    },
    {
      "entropy": 5.20054874420166,
      "epoch": 2.5393852065321805,
      "grad_norm": 1.0,
      "learning_rate": 0.0004357317243516075,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2238464578986168,
      "num_tokens": 60623056.0,
      "step": 26435
    },
    {
      "entropy": 5.166807985305786,
      "epoch": 2.5398655139289144,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043570773170736676,
      "loss": 4.831,
      "mean_token_accuracy": 0.23089597672224044,
      "num_tokens": 60634358.0,
      "step": 26440
    },
    {
      "entropy": 5.125444316864014,
      "epoch": 2.5403458213256487,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043568373533196976,
      "loss": 4.8642,
      "mean_token_accuracy": 0.22884978353977203,
      "num_tokens": 60644611.0,
      "step": 26445
    },
    {
      "entropy": 5.154325342178344,
      "epoch": 2.540826128722382,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043565973522597344,
      "loss": 4.861,
      "mean_token_accuracy": 0.22964026182889938,
      "num_tokens": 60655873.0,
      "step": 26450
    },
    {
      "entropy": 5.199005842208862,
      "epoch": 2.5413064361191164,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043563573138993524,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2285619542002678,
      "num_tokens": 60668634.0,
      "step": 26455
    },
    {
      "entropy": 5.244334030151367,
      "epoch": 2.54178674351585,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004356117238244123,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22612323313951493,
      "num_tokens": 60679460.0,
      "step": 26460
    },
    {
      "entropy": 5.1368059635162355,
      "epoch": 2.542267050912584,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043558771252996204,
      "loss": 4.8262,
      "mean_token_accuracy": 0.2327822983264923,
      "num_tokens": 60689704.0,
      "step": 26465
    },
    {
      "entropy": 5.193216800689697,
      "epoch": 2.542747358309318,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000435563697507142,
      "loss": 4.8202,
      "mean_token_accuracy": 0.2289538264274597,
      "num_tokens": 60700838.0,
      "step": 26470
    },
    {
      "entropy": 5.168218946456909,
      "epoch": 2.543227665706052,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004355396787565096,
      "loss": 4.8222,
      "mean_token_accuracy": 0.22951307147741318,
      "num_tokens": 60712048.0,
      "step": 26475
    },
    {
      "entropy": 5.0935179710388185,
      "epoch": 2.543707973102786,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043551565627862257,
      "loss": 4.7368,
      "mean_token_accuracy": 0.24555462896823882,
      "num_tokens": 60723451.0,
      "step": 26480
    },
    {
      "entropy": 5.073282480239868,
      "epoch": 2.5441882804995197,
      "grad_norm": 1.0,
      "learning_rate": 0.0004354916300740387,
      "loss": 4.7934,
      "mean_token_accuracy": 0.24268673360347748,
      "num_tokens": 60735992.0,
      "step": 26485
    },
    {
      "entropy": 5.177085542678833,
      "epoch": 2.5446685878962536,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004354676001433157,
      "loss": 4.7972,
      "mean_token_accuracy": 0.23618430644273758,
      "num_tokens": 60746653.0,
      "step": 26490
    },
    {
      "entropy": 5.159961700439453,
      "epoch": 2.5451488952929875,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004354435664870116,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22932415455579758,
      "num_tokens": 60757884.0,
      "step": 26495
    },
    {
      "entropy": 5.126038885116577,
      "epoch": 2.5456292026897214,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043541952910568417,
      "loss": 4.7676,
      "mean_token_accuracy": 0.23382105380296708,
      "num_tokens": 60769243.0,
      "step": 26500
    },
    {
      "entropy": 5.1646137714385985,
      "epoch": 2.5461095100864553,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004353954879998916,
      "loss": 4.9369,
      "mean_token_accuracy": 0.2211565151810646,
      "num_tokens": 60781168.0,
      "step": 26505
    },
    {
      "entropy": 5.154367923736572,
      "epoch": 2.546589817483189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004353714431701922,
      "loss": 4.8033,
      "mean_token_accuracy": 0.23768406063318254,
      "num_tokens": 60792743.0,
      "step": 26510
    },
    {
      "entropy": 5.1362096786499025,
      "epoch": 2.547070124879923,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004353473946171441,
      "loss": 4.7411,
      "mean_token_accuracy": 0.23990954607725143,
      "num_tokens": 60804251.0,
      "step": 26515
    },
    {
      "entropy": 5.195781183242798,
      "epoch": 2.547550432276657,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043532334234130547,
      "loss": 4.8297,
      "mean_token_accuracy": 0.23189615905284883,
      "num_tokens": 60815567.0,
      "step": 26520
    },
    {
      "entropy": 5.113167238235474,
      "epoch": 2.5480307396733908,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043529928634323503,
      "loss": 4.8506,
      "mean_token_accuracy": 0.226038458943367,
      "num_tokens": 60827028.0,
      "step": 26525
    },
    {
      "entropy": 5.081631755828857,
      "epoch": 2.548511047070125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043527522662349113,
      "loss": 4.7589,
      "mean_token_accuracy": 0.23890237361192704,
      "num_tokens": 60839158.0,
      "step": 26530
    },
    {
      "entropy": 5.220319509506226,
      "epoch": 2.5489913544668585,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004352511631826324,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22058920711278915,
      "num_tokens": 60850022.0,
      "step": 26535
    },
    {
      "entropy": 5.0794469833374025,
      "epoch": 2.549471661863593,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004352270960212175,
      "loss": 4.7267,
      "mean_token_accuracy": 0.23991550356149674,
      "num_tokens": 60861259.0,
      "step": 26540
    },
    {
      "entropy": 5.212015628814697,
      "epoch": 2.5499519692603267,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004352030251398052,
      "loss": 4.8045,
      "mean_token_accuracy": 0.2313162162899971,
      "num_tokens": 60871838.0,
      "step": 26545
    },
    {
      "entropy": 5.116426420211792,
      "epoch": 2.5504322766570606,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043517895053895434,
      "loss": 4.7659,
      "mean_token_accuracy": 0.24484134167432786,
      "num_tokens": 60883939.0,
      "step": 26550
    },
    {
      "entropy": 5.176839303970337,
      "epoch": 2.5509125840537945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004351548722192239,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22933577597141266,
      "num_tokens": 60895799.0,
      "step": 26555
    },
    {
      "entropy": 5.165253686904907,
      "epoch": 2.5513928914505284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004351307901811729,
      "loss": 4.7584,
      "mean_token_accuracy": 0.2347305789589882,
      "num_tokens": 60906330.0,
      "step": 26560
    },
    {
      "entropy": 5.089956426620484,
      "epoch": 2.5518731988472623,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004351067044253605,
      "loss": 4.7777,
      "mean_token_accuracy": 0.23347077816724776,
      "num_tokens": 60917613.0,
      "step": 26565
    },
    {
      "entropy": 5.183178091049195,
      "epoch": 2.552353506243996,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043508261495234577,
      "loss": 4.9331,
      "mean_token_accuracy": 0.21920356750488282,
      "num_tokens": 60929116.0,
      "step": 26570
    },
    {
      "entropy": 5.116840410232544,
      "epoch": 2.55283381364073,
      "grad_norm": 0.859375,
      "learning_rate": 0.000435058521762688,
      "loss": 4.6846,
      "mean_token_accuracy": 0.2424318492412567,
      "num_tokens": 60940455.0,
      "step": 26575
    },
    {
      "entropy": 5.1432195663452145,
      "epoch": 2.553314121037464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004350344248569467,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22442502230405809,
      "num_tokens": 60953606.0,
      "step": 26580
    },
    {
      "entropy": 5.156888437271118,
      "epoch": 2.553794428434198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004350103242356813,
      "loss": 4.7996,
      "mean_token_accuracy": 0.2322482630610466,
      "num_tokens": 60964487.0,
      "step": 26585
    },
    {
      "entropy": 5.2700526237487795,
      "epoch": 2.5542747358309317,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004349862198994513,
      "loss": 4.9632,
      "mean_token_accuracy": 0.2225348472595215,
      "num_tokens": 60976249.0,
      "step": 26590
    },
    {
      "entropy": 5.206593132019043,
      "epoch": 2.5547550432276656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004349621118488163,
      "loss": 4.8159,
      "mean_token_accuracy": 0.23790787607431413,
      "num_tokens": 60987581.0,
      "step": 26595
    },
    {
      "entropy": 5.104147291183471,
      "epoch": 2.5552353506243994,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004349380000843361,
      "loss": 4.7979,
      "mean_token_accuracy": 0.23061565607786177,
      "num_tokens": 60999286.0,
      "step": 26600
    },
    {
      "entropy": 5.106215095520019,
      "epoch": 2.5557156580211338,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004349138846065704,
      "loss": 4.7375,
      "mean_token_accuracy": 0.23778731673955916,
      "num_tokens": 61010207.0,
      "step": 26605
    },
    {
      "entropy": 5.112382507324218,
      "epoch": 2.556195965417867,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004348897654160791,
      "loss": 4.8224,
      "mean_token_accuracy": 0.23485430628061293,
      "num_tokens": 61021842.0,
      "step": 26610
    },
    {
      "entropy": 5.149652147293091,
      "epoch": 2.5566762728146015,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004348656425134223,
      "loss": 4.8708,
      "mean_token_accuracy": 0.22591934055089952,
      "num_tokens": 61033344.0,
      "step": 26615
    },
    {
      "entropy": 5.074617624282837,
      "epoch": 2.5571565802113354,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043484151589916,
      "loss": 4.7801,
      "mean_token_accuracy": 0.2354456290602684,
      "num_tokens": 61044920.0,
      "step": 26620
    },
    {
      "entropy": 5.15938892364502,
      "epoch": 2.5576368876080693,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004348173855738523,
      "loss": 4.7419,
      "mean_token_accuracy": 0.23890597969293595,
      "num_tokens": 61056132.0,
      "step": 26625
    },
    {
      "entropy": 5.103297185897827,
      "epoch": 2.558117195004803,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004347932515380595,
      "loss": 4.7877,
      "mean_token_accuracy": 0.23695410788059235,
      "num_tokens": 61068590.0,
      "step": 26630
    },
    {
      "entropy": 5.1661652565002445,
      "epoch": 2.558597502401537,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004347691137923418,
      "loss": 4.8521,
      "mean_token_accuracy": 0.23301667273044585,
      "num_tokens": 61080011.0,
      "step": 26635
    },
    {
      "entropy": 5.13643741607666,
      "epoch": 2.559077809798271,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004347449723372598,
      "loss": 4.7316,
      "mean_token_accuracy": 0.2382487416267395,
      "num_tokens": 61091448.0,
      "step": 26640
    },
    {
      "entropy": 5.0812522888183596,
      "epoch": 2.559558117195005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004347208271733738,
      "loss": 4.6938,
      "mean_token_accuracy": 0.24659420996904374,
      "num_tokens": 61102507.0,
      "step": 26645
    },
    {
      "entropy": 5.101203346252442,
      "epoch": 2.5600384245917387,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004346966783012445,
      "loss": 4.8092,
      "mean_token_accuracy": 0.227750001847744,
      "num_tokens": 61114577.0,
      "step": 26650
    },
    {
      "entropy": 5.113767433166504,
      "epoch": 2.5605187319884726,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043467252572143247,
      "loss": 4.6752,
      "mean_token_accuracy": 0.243193319439888,
      "num_tokens": 61126044.0,
      "step": 26655
    },
    {
      "entropy": 5.171478033065796,
      "epoch": 2.5609990393852065,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043464836943449866,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2197179153561592,
      "num_tokens": 61137880.0,
      "step": 26660
    },
    {
      "entropy": 5.125174474716187,
      "epoch": 2.5614793467819403,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004346242094410036,
      "loss": 4.8674,
      "mean_token_accuracy": 0.22778966426849365,
      "num_tokens": 61149103.0,
      "step": 26665
    },
    {
      "entropy": 5.047973871231079,
      "epoch": 2.561959654178674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004346000457415085,
      "loss": 4.7611,
      "mean_token_accuracy": 0.23215031176805495,
      "num_tokens": 61161235.0,
      "step": 26670
    },
    {
      "entropy": 5.148712825775147,
      "epoch": 2.562439961575408,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043457587833657424,
      "loss": 4.8389,
      "mean_token_accuracy": 0.23094941377639772,
      "num_tokens": 61172497.0,
      "step": 26675
    },
    {
      "entropy": 5.293687057495117,
      "epoch": 2.5629202689721424,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043455170722676194,
      "loss": 4.9814,
      "mean_token_accuracy": 0.21550409346818925,
      "num_tokens": 61183795.0,
      "step": 26680
    },
    {
      "entropy": 5.136210823059082,
      "epoch": 2.563400576368876,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004345275324126328,
      "loss": 4.7662,
      "mean_token_accuracy": 0.24045863300561904,
      "num_tokens": 61195909.0,
      "step": 26685
    },
    {
      "entropy": 5.095781564712524,
      "epoch": 2.56388088376561,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043450335389474796,
      "loss": 4.7796,
      "mean_token_accuracy": 0.23264581710100174,
      "num_tokens": 61207017.0,
      "step": 26690
    },
    {
      "entropy": 5.148272514343262,
      "epoch": 2.5643611911623436,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004344791716736689,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22744334042072295,
      "num_tokens": 61219172.0,
      "step": 26695
    },
    {
      "entropy": 5.20787935256958,
      "epoch": 2.564841498559078,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043445498574995705,
      "loss": 4.8738,
      "mean_token_accuracy": 0.23525165170431137,
      "num_tokens": 61232612.0,
      "step": 26700
    },
    {
      "entropy": 5.168517827987671,
      "epoch": 2.565321805955812,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043443079612417394,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22835026532411576,
      "num_tokens": 61243627.0,
      "step": 26705
    },
    {
      "entropy": 5.135371541976928,
      "epoch": 2.5658021133525457,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004344066027968812,
      "loss": 4.8754,
      "mean_token_accuracy": 0.22621981352567672,
      "num_tokens": 61254516.0,
      "step": 26710
    },
    {
      "entropy": 5.228228425979614,
      "epoch": 2.5662824207492796,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043438240576864034,
      "loss": 4.8308,
      "mean_token_accuracy": 0.2297377347946167,
      "num_tokens": 61266764.0,
      "step": 26715
    },
    {
      "entropy": 5.091504812240601,
      "epoch": 2.5667627281460135,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004343582050400134,
      "loss": 4.7337,
      "mean_token_accuracy": 0.24118281453847884,
      "num_tokens": 61278088.0,
      "step": 26720
    },
    {
      "entropy": 5.066771554946899,
      "epoch": 2.5672430355427474,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004343340006115621,
      "loss": 4.8193,
      "mean_token_accuracy": 0.23274065256118776,
      "num_tokens": 61289558.0,
      "step": 26725
    },
    {
      "entropy": 5.020469808578492,
      "epoch": 2.5677233429394812,
      "grad_norm": 1.0,
      "learning_rate": 0.0004343097924838483,
      "loss": 4.748,
      "mean_token_accuracy": 0.24121522456407546,
      "num_tokens": 61301815.0,
      "step": 26730
    },
    {
      "entropy": 5.218504858016968,
      "epoch": 2.568203650336215,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004342855806574344,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22506242543458937,
      "num_tokens": 61312216.0,
      "step": 26735
    },
    {
      "entropy": 5.15383734703064,
      "epoch": 2.568683957732949,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004342613651328822,
      "loss": 4.8162,
      "mean_token_accuracy": 0.23086303919553758,
      "num_tokens": 61322485.0,
      "step": 26740
    },
    {
      "entropy": 5.180078125,
      "epoch": 2.569164265129683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000434237145910754,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22608800381422042,
      "num_tokens": 61334654.0,
      "step": 26745
    },
    {
      "entropy": 5.077005767822266,
      "epoch": 2.5696445725264168,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043421292299161213,
      "loss": 4.7551,
      "mean_token_accuracy": 0.24006779342889786,
      "num_tokens": 61346563.0,
      "step": 26750
    },
    {
      "entropy": 5.130997228622436,
      "epoch": 2.570124879923151,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043418869637601887,
      "loss": 4.81,
      "mean_token_accuracy": 0.22727591842412947,
      "num_tokens": 61360121.0,
      "step": 26755
    },
    {
      "entropy": 5.185331392288208,
      "epoch": 2.5706051873198845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043416446606453686,
      "loss": 4.7735,
      "mean_token_accuracy": 0.23791826367378235,
      "num_tokens": 61371340.0,
      "step": 26760
    },
    {
      "entropy": 5.08425874710083,
      "epoch": 2.571085494716619,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004341402320577285,
      "loss": 4.7724,
      "mean_token_accuracy": 0.2332771047949791,
      "num_tokens": 61382571.0,
      "step": 26765
    },
    {
      "entropy": 5.034887886047363,
      "epoch": 2.5715658021133523,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004341159943561566,
      "loss": 4.6797,
      "mean_token_accuracy": 0.24153711646795273,
      "num_tokens": 61394548.0,
      "step": 26770
    },
    {
      "entropy": 5.106968975067138,
      "epoch": 2.5720461095100866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004340917529603837,
      "loss": 4.753,
      "mean_token_accuracy": 0.23398763090372085,
      "num_tokens": 61405668.0,
      "step": 26775
    },
    {
      "entropy": 5.15061616897583,
      "epoch": 2.5725264169068205,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004340675078709728,
      "loss": 4.8409,
      "mean_token_accuracy": 0.22531768530607224,
      "num_tokens": 61417396.0,
      "step": 26780
    },
    {
      "entropy": 5.0940203189849855,
      "epoch": 2.5730067243035544,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004340432590884866,
      "loss": 4.7227,
      "mean_token_accuracy": 0.24593927413225175,
      "num_tokens": 61427981.0,
      "step": 26785
    },
    {
      "entropy": 5.127006959915161,
      "epoch": 2.5734870317002883,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043401900661348825,
      "loss": 4.804,
      "mean_token_accuracy": 0.23090792447328568,
      "num_tokens": 61439011.0,
      "step": 26790
    },
    {
      "entropy": 5.174531173706055,
      "epoch": 2.573967339097022,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043399475044654073,
      "loss": 4.8846,
      "mean_token_accuracy": 0.23367905616760254,
      "num_tokens": 61450567.0,
      "step": 26795
    },
    {
      "entropy": 5.146568870544433,
      "epoch": 2.574447646493756,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043397049058820726,
      "loss": 4.7896,
      "mean_token_accuracy": 0.23230030685663222,
      "num_tokens": 61461563.0,
      "step": 26800
    },
    {
      "entropy": 5.1396583080291744,
      "epoch": 2.57492795389049,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000433946227039051,
      "loss": 4.8104,
      "mean_token_accuracy": 0.2307019904255867,
      "num_tokens": 61472697.0,
      "step": 26805
    },
    {
      "entropy": 5.112583160400391,
      "epoch": 2.5754082612872238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004339219597996353,
      "loss": 4.8138,
      "mean_token_accuracy": 0.22455327808856965,
      "num_tokens": 61485877.0,
      "step": 26810
    },
    {
      "entropy": 5.102473783493042,
      "epoch": 2.5758885686839577,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004338976888705236,
      "loss": 4.7882,
      "mean_token_accuracy": 0.22976726293563843,
      "num_tokens": 61499007.0,
      "step": 26815
    },
    {
      "entropy": 5.183072948455811,
      "epoch": 2.5763688760806915,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043387341425227944,
      "loss": 4.8512,
      "mean_token_accuracy": 0.22983447462320328,
      "num_tokens": 61510901.0,
      "step": 26820
    },
    {
      "entropy": 5.10213713645935,
      "epoch": 2.5768491834774254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004338491359454664,
      "loss": 4.7725,
      "mean_token_accuracy": 0.2361150398850441,
      "num_tokens": 61523628.0,
      "step": 26825
    },
    {
      "entropy": 5.122937488555908,
      "epoch": 2.5773294908741593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043382485395064796,
      "loss": 4.8226,
      "mean_token_accuracy": 0.2403786525130272,
      "num_tokens": 61534765.0,
      "step": 26830
    },
    {
      "entropy": 5.126094341278076,
      "epoch": 2.577809798270893,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004338005682683881,
      "loss": 4.8326,
      "mean_token_accuracy": 0.23015301376581193,
      "num_tokens": 61547009.0,
      "step": 26835
    },
    {
      "entropy": 5.241258001327514,
      "epoch": 2.5782901056676275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043377627889925057,
      "loss": 4.9402,
      "mean_token_accuracy": 0.21959612220525743,
      "num_tokens": 61558713.0,
      "step": 26840
    },
    {
      "entropy": 5.212319040298462,
      "epoch": 2.578770413064361,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004337519858437994,
      "loss": 4.882,
      "mean_token_accuracy": 0.23149473965168,
      "num_tokens": 61569841.0,
      "step": 26845
    },
    {
      "entropy": 5.102365207672119,
      "epoch": 2.5792507204610953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004337276891025984,
      "loss": 4.7432,
      "mean_token_accuracy": 0.2373405620455742,
      "num_tokens": 61580878.0,
      "step": 26850
    },
    {
      "entropy": 5.204611825942993,
      "epoch": 2.579731027857829,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043370338867621184,
      "loss": 4.9351,
      "mean_token_accuracy": 0.21650485545396805,
      "num_tokens": 61593328.0,
      "step": 26855
    },
    {
      "entropy": 5.188858985900879,
      "epoch": 2.580211335254563,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043367908456520387,
      "loss": 4.8465,
      "mean_token_accuracy": 0.22644302248954773,
      "num_tokens": 61605085.0,
      "step": 26860
    },
    {
      "entropy": 5.171852254867554,
      "epoch": 2.580691642651297,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004336547767701387,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2224779948592186,
      "num_tokens": 61618410.0,
      "step": 26865
    },
    {
      "entropy": 5.180705785751343,
      "epoch": 2.581171950048031,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043363046529158077,
      "loss": 4.7961,
      "mean_token_accuracy": 0.23887282907962798,
      "num_tokens": 61628795.0,
      "step": 26870
    },
    {
      "entropy": 5.181337070465088,
      "epoch": 2.5816522574447647,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004336061501300944,
      "loss": 4.8121,
      "mean_token_accuracy": 0.2355806350708008,
      "num_tokens": 61638769.0,
      "step": 26875
    },
    {
      "entropy": 5.104330444335938,
      "epoch": 2.5821325648414986,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004335818312862442,
      "loss": 4.7818,
      "mean_token_accuracy": 0.23897703886032104,
      "num_tokens": 61648748.0,
      "step": 26880
    },
    {
      "entropy": 5.0765848636627195,
      "epoch": 2.5826128722382324,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043355750876059485,
      "loss": 4.6964,
      "mean_token_accuracy": 0.23791560977697374,
      "num_tokens": 61658733.0,
      "step": 26885
    },
    {
      "entropy": 5.1194089412689205,
      "epoch": 2.5830931796349663,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043353318255371086,
      "loss": 4.82,
      "mean_token_accuracy": 0.22568898499011994,
      "num_tokens": 61669192.0,
      "step": 26890
    },
    {
      "entropy": 5.259398126602173,
      "epoch": 2.5835734870317,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004335088526661571,
      "loss": 4.9003,
      "mean_token_accuracy": 0.22575218975543976,
      "num_tokens": 61679091.0,
      "step": 26895
    },
    {
      "entropy": 5.198182106018066,
      "epoch": 2.584053794428434,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043348451909849855,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23573306649923326,
      "num_tokens": 61691143.0,
      "step": 26900
    },
    {
      "entropy": 5.250957489013672,
      "epoch": 2.584534101825168,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043346018185130006,
      "loss": 4.9121,
      "mean_token_accuracy": 0.23190910816192628,
      "num_tokens": 61703549.0,
      "step": 26905
    },
    {
      "entropy": 5.078440809249878,
      "epoch": 2.585014409221902,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043343584092512665,
      "loss": 4.7416,
      "mean_token_accuracy": 0.2375257357954979,
      "num_tokens": 61713447.0,
      "step": 26910
    },
    {
      "entropy": 5.182172060012817,
      "epoch": 2.585494716618636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004334114963205435,
      "loss": 4.8334,
      "mean_token_accuracy": 0.22358786314725876,
      "num_tokens": 61724848.0,
      "step": 26915
    },
    {
      "entropy": 5.103286790847778,
      "epoch": 2.5859750240153696,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004333871480381158,
      "loss": 4.8059,
      "mean_token_accuracy": 0.23032637685537338,
      "num_tokens": 61734624.0,
      "step": 26920
    },
    {
      "entropy": 5.179016542434693,
      "epoch": 2.586455331412104,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004333627960784088,
      "loss": 4.8131,
      "mean_token_accuracy": 0.2354464814066887,
      "num_tokens": 61745627.0,
      "step": 26925
    },
    {
      "entropy": 5.122391033172607,
      "epoch": 2.586935638808838,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004333384404419879,
      "loss": 4.8312,
      "mean_token_accuracy": 0.2293297082185745,
      "num_tokens": 61757050.0,
      "step": 26930
    },
    {
      "entropy": 5.2235105514526365,
      "epoch": 2.5874159462055717,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004333140811294186,
      "loss": 4.95,
      "mean_token_accuracy": 0.2239794909954071,
      "num_tokens": 61768896.0,
      "step": 26935
    },
    {
      "entropy": 5.237671279907227,
      "epoch": 2.5878962536023056,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004332897181412664,
      "loss": 4.8931,
      "mean_token_accuracy": 0.2212544247508049,
      "num_tokens": 61779809.0,
      "step": 26940
    },
    {
      "entropy": 5.085165357589721,
      "epoch": 2.5883765609990395,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043326535147809696,
      "loss": 4.7366,
      "mean_token_accuracy": 0.24002355635166167,
      "num_tokens": 61791737.0,
      "step": 26945
    },
    {
      "entropy": 5.128943300247192,
      "epoch": 2.5888568683957733,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043324098114047604,
      "loss": 4.7482,
      "mean_token_accuracy": 0.2351958766579628,
      "num_tokens": 61804193.0,
      "step": 26950
    },
    {
      "entropy": 5.092439270019531,
      "epoch": 2.589337175792507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004332166071289695,
      "loss": 4.7278,
      "mean_token_accuracy": 0.2379429116845131,
      "num_tokens": 61815735.0,
      "step": 26955
    },
    {
      "entropy": 5.124590587615967,
      "epoch": 2.589817483189241,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043319222944414304,
      "loss": 4.8,
      "mean_token_accuracy": 0.2403637334704399,
      "num_tokens": 61827263.0,
      "step": 26960
    },
    {
      "entropy": 5.132837915420533,
      "epoch": 2.590297790585975,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043316784808656276,
      "loss": 4.7149,
      "mean_token_accuracy": 0.23597896695137024,
      "num_tokens": 61838734.0,
      "step": 26965
    },
    {
      "entropy": 5.16165452003479,
      "epoch": 2.590778097982709,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043314346305679477,
      "loss": 4.8289,
      "mean_token_accuracy": 0.22982044219970704,
      "num_tokens": 61849935.0,
      "step": 26970
    },
    {
      "entropy": 5.064759159088135,
      "epoch": 2.5912584053794427,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043311907435540517,
      "loss": 4.7125,
      "mean_token_accuracy": 0.2451219752430916,
      "num_tokens": 61861849.0,
      "step": 26975
    },
    {
      "entropy": 5.163621139526367,
      "epoch": 2.5917387127761766,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004330946819829601,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23776889145374297,
      "num_tokens": 61873017.0,
      "step": 26980
    },
    {
      "entropy": 5.108000135421753,
      "epoch": 2.5922190201729105,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043307028594002597,
      "loss": 4.7821,
      "mean_token_accuracy": 0.23048410266637803,
      "num_tokens": 61884730.0,
      "step": 26985
    },
    {
      "entropy": 5.110158395767212,
      "epoch": 2.592699327569645,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043304588622716924,
      "loss": 4.807,
      "mean_token_accuracy": 0.2381896659731865,
      "num_tokens": 61896726.0,
      "step": 26990
    },
    {
      "entropy": 5.107672119140625,
      "epoch": 2.5931796349663783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004330214828449563,
      "loss": 4.7681,
      "mean_token_accuracy": 0.23335302919149398,
      "num_tokens": 61907931.0,
      "step": 26995
    },
    {
      "entropy": 5.135706806182862,
      "epoch": 2.5936599423631126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043299707579395365,
      "loss": 4.7407,
      "mean_token_accuracy": 0.24161033630371093,
      "num_tokens": 61919001.0,
      "step": 27000
    },
    {
      "epoch": 2.5936599423631126,
      "eval_entropy": 4.9524550884308205,
      "eval_loss": 4.913448333740234,
      "eval_mean_token_accuracy": 0.2359061943478828,
      "eval_num_tokens": 61919001.0,
      "eval_runtime": 26.6153,
      "eval_samples_per_second": 1232.937,
      "eval_steps_per_second": 154.122,
      "step": 27000
    },
    {
      "entropy": 5.222839832305908,
      "epoch": 2.594140249759846,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004329726650747282,
      "loss": 4.938,
      "mean_token_accuracy": 0.22010616362094879,
      "num_tokens": 61931948.0,
      "step": 27005
    },
    {
      "entropy": 5.198479509353637,
      "epoch": 2.5946205571565804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004329482506878465,
      "loss": 4.9239,
      "mean_token_accuracy": 0.21551909893751145,
      "num_tokens": 61942781.0,
      "step": 27010
    },
    {
      "entropy": 5.1219532012939455,
      "epoch": 2.5951008645533142,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043292383263387536,
      "loss": 4.792,
      "mean_token_accuracy": 0.23862185925245286,
      "num_tokens": 61953981.0,
      "step": 27015
    },
    {
      "entropy": 5.113706254959107,
      "epoch": 2.595581171950048,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043289941091338187,
      "loss": 4.7588,
      "mean_token_accuracy": 0.24137406349182128,
      "num_tokens": 61965328.0,
      "step": 27020
    },
    {
      "entropy": 5.164354133605957,
      "epoch": 2.596061479346782,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004328749855269329,
      "loss": 4.7658,
      "mean_token_accuracy": 0.23660996556282043,
      "num_tokens": 61975427.0,
      "step": 27025
    },
    {
      "entropy": 5.14285979270935,
      "epoch": 2.596541786743516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004328505564750955,
      "loss": 4.8269,
      "mean_token_accuracy": 0.22698781341314317,
      "num_tokens": 61985724.0,
      "step": 27030
    },
    {
      "entropy": 5.250479698181152,
      "epoch": 2.5970220941402498,
      "grad_norm": 1.015625,
      "learning_rate": 0.000432826123758437,
      "loss": 4.9889,
      "mean_token_accuracy": 0.2224618077278137,
      "num_tokens": 61996382.0,
      "step": 27035
    },
    {
      "entropy": 5.200682497024536,
      "epoch": 2.5975024015369836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004328016873775244,
      "loss": 4.8699,
      "mean_token_accuracy": 0.23141425400972365,
      "num_tokens": 62008859.0,
      "step": 27040
    },
    {
      "entropy": 5.098199415206909,
      "epoch": 2.5979827089337175,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043277724733292527,
      "loss": 4.6926,
      "mean_token_accuracy": 0.24305624812841414,
      "num_tokens": 62019096.0,
      "step": 27045
    },
    {
      "entropy": 5.149641227722168,
      "epoch": 2.5984630163304514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004327528036252071,
      "loss": 4.8302,
      "mean_token_accuracy": 0.2346649155020714,
      "num_tokens": 62030874.0,
      "step": 27050
    },
    {
      "entropy": 5.092842292785645,
      "epoch": 2.5989433237271853,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004327283562549371,
      "loss": 4.7423,
      "mean_token_accuracy": 0.23982830345630646,
      "num_tokens": 62042620.0,
      "step": 27055
    },
    {
      "entropy": 5.134811496734619,
      "epoch": 2.599423631123919,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000432703905222683,
      "loss": 4.8308,
      "mean_token_accuracy": 0.23652447611093522,
      "num_tokens": 62054124.0,
      "step": 27060
    },
    {
      "entropy": 5.135378837585449,
      "epoch": 2.5999039385206535,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043267945052901264,
      "loss": 4.7671,
      "mean_token_accuracy": 0.23503005206584932,
      "num_tokens": 62064647.0,
      "step": 27065
    },
    {
      "entropy": 5.143292808532715,
      "epoch": 2.600384245917387,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004326549921744935,
      "loss": 4.778,
      "mean_token_accuracy": 0.24063572883605958,
      "num_tokens": 62075298.0,
      "step": 27070
    },
    {
      "entropy": 5.18067626953125,
      "epoch": 2.6008645533141213,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004326305301596937,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22134626805782318,
      "num_tokens": 62087875.0,
      "step": 27075
    },
    {
      "entropy": 5.108724784851074,
      "epoch": 2.6013448607108547,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043260606448518096,
      "loss": 4.6746,
      "mean_token_accuracy": 0.24498000591993332,
      "num_tokens": 62098538.0,
      "step": 27080
    },
    {
      "entropy": 5.080880641937256,
      "epoch": 2.601825168107589,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043258159515152347,
      "loss": 4.7583,
      "mean_token_accuracy": 0.23256036639213562,
      "num_tokens": 62109637.0,
      "step": 27085
    },
    {
      "entropy": 5.078181886672974,
      "epoch": 2.602305475504323,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004325571221592892,
      "loss": 4.7225,
      "mean_token_accuracy": 0.2366447389125824,
      "num_tokens": 62120471.0,
      "step": 27090
    },
    {
      "entropy": 5.125346422195435,
      "epoch": 2.6027857829010568,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043253264550904646,
      "loss": 4.8044,
      "mean_token_accuracy": 0.2362074300646782,
      "num_tokens": 62130150.0,
      "step": 27095
    },
    {
      "entropy": 5.16070523262024,
      "epoch": 2.6032660902977907,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004325081652013635,
      "loss": 4.8679,
      "mean_token_accuracy": 0.2289569139480591,
      "num_tokens": 62142382.0,
      "step": 27100
    },
    {
      "entropy": 5.133693838119507,
      "epoch": 2.6037463976945245,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043248368123680855,
      "loss": 4.8018,
      "mean_token_accuracy": 0.23536351919174195,
      "num_tokens": 62155398.0,
      "step": 27105
    },
    {
      "entropy": 5.086481666564941,
      "epoch": 2.6042267050912584,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043245919361595026,
      "loss": 4.7332,
      "mean_token_accuracy": 0.2377217933535576,
      "num_tokens": 62167694.0,
      "step": 27110
    },
    {
      "entropy": 5.059261846542358,
      "epoch": 2.6047070124879923,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043243470233935696,
      "loss": 4.7335,
      "mean_token_accuracy": 0.2396539956331253,
      "num_tokens": 62179420.0,
      "step": 27115
    },
    {
      "entropy": 5.141218662261963,
      "epoch": 2.605187319884726,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004324102074075973,
      "loss": 4.7651,
      "mean_token_accuracy": 0.23624598681926728,
      "num_tokens": 62190881.0,
      "step": 27120
    },
    {
      "entropy": 5.057152891159058,
      "epoch": 2.60566762728146,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004323857088212402,
      "loss": 4.6887,
      "mean_token_accuracy": 0.23906176388263703,
      "num_tokens": 62201513.0,
      "step": 27125
    },
    {
      "entropy": 5.066849613189698,
      "epoch": 2.606147934678194,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004323612065808541,
      "loss": 4.7447,
      "mean_token_accuracy": 0.23588161766529084,
      "num_tokens": 62212921.0,
      "step": 27130
    },
    {
      "entropy": 5.072431325912476,
      "epoch": 2.606628242074928,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043233670068700827,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23682795464992523,
      "num_tokens": 62224579.0,
      "step": 27135
    },
    {
      "entropy": 5.088177490234375,
      "epoch": 2.6071085494716617,
      "grad_norm": 1.0,
      "learning_rate": 0.0004323121911402713,
      "loss": 4.8332,
      "mean_token_accuracy": 0.2302288681268692,
      "num_tokens": 62236181.0,
      "step": 27140
    },
    {
      "entropy": 5.209999418258667,
      "epoch": 2.6075888568683956,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043228767794121245,
      "loss": 4.9181,
      "mean_token_accuracy": 0.23004982322454454,
      "num_tokens": 62247607.0,
      "step": 27145
    },
    {
      "entropy": 5.209290075302124,
      "epoch": 2.60806916426513,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004322631610904006,
      "loss": 4.8173,
      "mean_token_accuracy": 0.2348630651831627,
      "num_tokens": 62259171.0,
      "step": 27150
    },
    {
      "entropy": 5.163558721542358,
      "epoch": 2.6085494716618634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043223864058840534,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23147836029529573,
      "num_tokens": 62271105.0,
      "step": 27155
    },
    {
      "entropy": 5.128911876678467,
      "epoch": 2.6090297790585977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043221411643579557,
      "loss": 4.8049,
      "mean_token_accuracy": 0.2373675972223282,
      "num_tokens": 62282818.0,
      "step": 27160
    },
    {
      "entropy": 5.121401596069336,
      "epoch": 2.6095100864553316,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043218958863314096,
      "loss": 4.8488,
      "mean_token_accuracy": 0.22862073481082917,
      "num_tokens": 62294604.0,
      "step": 27165
    },
    {
      "entropy": 5.1859150409698485,
      "epoch": 2.6099903938520654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004321650571810109,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2351876512169838,
      "num_tokens": 62305871.0,
      "step": 27170
    },
    {
      "entropy": 5.120514726638794,
      "epoch": 2.6104707012487993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004321405220799747,
      "loss": 4.7751,
      "mean_token_accuracy": 0.23848947435617446,
      "num_tokens": 62317100.0,
      "step": 27175
    },
    {
      "entropy": 5.118195581436157,
      "epoch": 2.610951008645533,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004321159833306024,
      "loss": 4.78,
      "mean_token_accuracy": 0.2301449492573738,
      "num_tokens": 62328934.0,
      "step": 27180
    },
    {
      "entropy": 5.191256666183472,
      "epoch": 2.611431316042267,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004320914409334634,
      "loss": 4.8411,
      "mean_token_accuracy": 0.23135278671979903,
      "num_tokens": 62340812.0,
      "step": 27185
    },
    {
      "entropy": 5.120903205871582,
      "epoch": 2.611911623439001,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004320668948891276,
      "loss": 4.788,
      "mean_token_accuracy": 0.23665708899497986,
      "num_tokens": 62352753.0,
      "step": 27190
    },
    {
      "entropy": 5.175097370147705,
      "epoch": 2.612391930835735,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043204234519816486,
      "loss": 4.8062,
      "mean_token_accuracy": 0.23216795325279235,
      "num_tokens": 62362731.0,
      "step": 27195
    },
    {
      "entropy": 5.2038475513458256,
      "epoch": 2.6128722382324687,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004320177918611453,
      "loss": 4.8446,
      "mean_token_accuracy": 0.2300620272755623,
      "num_tokens": 62374299.0,
      "step": 27200
    },
    {
      "entropy": 5.134019088745117,
      "epoch": 2.6133525456292026,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043199323487863876,
      "loss": 4.844,
      "mean_token_accuracy": 0.22377394884824753,
      "num_tokens": 62387708.0,
      "step": 27205
    },
    {
      "entropy": 5.15350112915039,
      "epoch": 2.6138328530259365,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043196867425121554,
      "loss": 4.7948,
      "mean_token_accuracy": 0.2377636179327965,
      "num_tokens": 62399139.0,
      "step": 27210
    },
    {
      "entropy": 5.1719715118408205,
      "epoch": 2.6143131604226704,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043194410997944577,
      "loss": 4.7965,
      "mean_token_accuracy": 0.228760202229023,
      "num_tokens": 62411581.0,
      "step": 27215
    },
    {
      "entropy": 5.166399097442627,
      "epoch": 2.6147934678194042,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043191954206389985,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22237877398729325,
      "num_tokens": 62423768.0,
      "step": 27220
    },
    {
      "entropy": 5.140848636627197,
      "epoch": 2.6152737752161386,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004318949705051481,
      "loss": 4.8271,
      "mean_token_accuracy": 0.23395203799009323,
      "num_tokens": 62435716.0,
      "step": 27225
    },
    {
      "entropy": 5.121417474746704,
      "epoch": 2.615754082612872,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004318703953037609,
      "loss": 4.7455,
      "mean_token_accuracy": 0.22937429398298265,
      "num_tokens": 62447409.0,
      "step": 27230
    },
    {
      "entropy": 5.098790550231934,
      "epoch": 2.6162343900096063,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004318458164603091,
      "loss": 4.7787,
      "mean_token_accuracy": 0.23898401558399202,
      "num_tokens": 62458900.0,
      "step": 27235
    },
    {
      "entropy": 5.155187559127808,
      "epoch": 2.61671469740634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004318212339753631,
      "loss": 4.8354,
      "mean_token_accuracy": 0.233867084980011,
      "num_tokens": 62471433.0,
      "step": 27240
    },
    {
      "entropy": 5.144567155838013,
      "epoch": 2.617195004803074,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043179664784949375,
      "loss": 4.8574,
      "mean_token_accuracy": 0.2291429102420807,
      "num_tokens": 62482633.0,
      "step": 27245
    },
    {
      "entropy": 5.1346252918243405,
      "epoch": 2.617675312199808,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004317720580832717,
      "loss": 4.6748,
      "mean_token_accuracy": 0.24546798914670945,
      "num_tokens": 62492661.0,
      "step": 27250
    },
    {
      "entropy": 5.101106119155884,
      "epoch": 2.618155619596542,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004317474646772681,
      "loss": 4.7923,
      "mean_token_accuracy": 0.23491035848855973,
      "num_tokens": 62504470.0,
      "step": 27255
    },
    {
      "entropy": 5.254567241668701,
      "epoch": 2.6186359269932757,
      "grad_norm": 1.0,
      "learning_rate": 0.0004317228676320539,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22066180408000946,
      "num_tokens": 62514711.0,
      "step": 27260
    },
    {
      "entropy": 5.234379005432129,
      "epoch": 2.6191162343900096,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043169826694819987,
      "loss": 4.8247,
      "mean_token_accuracy": 0.23570848256349564,
      "num_tokens": 62525939.0,
      "step": 27265
    },
    {
      "entropy": 5.084956693649292,
      "epoch": 2.6195965417867435,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004316736626262775,
      "loss": 4.7877,
      "mean_token_accuracy": 0.23713348954916,
      "num_tokens": 62536643.0,
      "step": 27270
    },
    {
      "entropy": 5.118238019943237,
      "epoch": 2.6200768491834774,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004316490546668579,
      "loss": 4.8024,
      "mean_token_accuracy": 0.23492947220802307,
      "num_tokens": 62547699.0,
      "step": 27275
    },
    {
      "entropy": 5.167779397964478,
      "epoch": 2.6205571565802113,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004316244430705124,
      "loss": 4.8196,
      "mean_token_accuracy": 0.22624894231557846,
      "num_tokens": 62557756.0,
      "step": 27280
    },
    {
      "entropy": 5.119980573654175,
      "epoch": 2.621037463976945,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004315998278378123,
      "loss": 4.824,
      "mean_token_accuracy": 0.2320847451686859,
      "num_tokens": 62570046.0,
      "step": 27285
    },
    {
      "entropy": 5.180103540420532,
      "epoch": 2.621517771373679,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043157520896932943,
      "loss": 4.8691,
      "mean_token_accuracy": 0.22879524230957032,
      "num_tokens": 62580037.0,
      "step": 27290
    },
    {
      "entropy": 5.115751028060913,
      "epoch": 2.621998078770413,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000431550586465635,
      "loss": 4.8184,
      "mean_token_accuracy": 0.24122264236211777,
      "num_tokens": 62591442.0,
      "step": 27295
    },
    {
      "entropy": 5.20201096534729,
      "epoch": 2.6224783861671472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043152596032730085,
      "loss": 4.8556,
      "mean_token_accuracy": 0.2261336237192154,
      "num_tokens": 62603367.0,
      "step": 27300
    },
    {
      "entropy": 5.115754842758179,
      "epoch": 2.6229586935638807,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043150133055489865,
      "loss": 4.7214,
      "mean_token_accuracy": 0.24158381372690202,
      "num_tokens": 62613506.0,
      "step": 27305
    },
    {
      "entropy": 5.057960462570191,
      "epoch": 2.623439000960615,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004314766971490003,
      "loss": 4.7425,
      "mean_token_accuracy": 0.23489145636558534,
      "num_tokens": 62625389.0,
      "step": 27310
    },
    {
      "entropy": 5.167988395690918,
      "epoch": 2.6239193083573484,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004314520601101776,
      "loss": 4.8162,
      "mean_token_accuracy": 0.23584689050912858,
      "num_tokens": 62637949.0,
      "step": 27315
    },
    {
      "entropy": 5.108074426651001,
      "epoch": 2.6243996157540828,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043142741943900275,
      "loss": 4.7479,
      "mean_token_accuracy": 0.23094150722026824,
      "num_tokens": 62649952.0,
      "step": 27320
    },
    {
      "entropy": 5.153662300109863,
      "epoch": 2.6248799231508166,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00043140277513604763,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22415365129709244,
      "num_tokens": 62661152.0,
      "step": 27325
    },
    {
      "entropy": 5.142730569839477,
      "epoch": 2.6253602305475505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004313781272018845,
      "loss": 4.86,
      "mean_token_accuracy": 0.2219841405749321,
      "num_tokens": 62671648.0,
      "step": 27330
    },
    {
      "entropy": 5.142015600204468,
      "epoch": 2.6258405379442844,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004313534756370857,
      "loss": 4.786,
      "mean_token_accuracy": 0.23029768615961074,
      "num_tokens": 62683114.0,
      "step": 27335
    },
    {
      "entropy": 5.126420640945435,
      "epoch": 2.6263208453410183,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043132882044222336,
      "loss": 4.7725,
      "mean_token_accuracy": 0.23863262683153152,
      "num_tokens": 62693921.0,
      "step": 27340
    },
    {
      "entropy": 5.176171541213989,
      "epoch": 2.626801152737752,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043130416161787005,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22429923415184022,
      "num_tokens": 62705342.0,
      "step": 27345
    },
    {
      "entropy": 5.202436923980713,
      "epoch": 2.627281460134486,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043127949916459823,
      "loss": 4.9046,
      "mean_token_accuracy": 0.2280184343457222,
      "num_tokens": 62717359.0,
      "step": 27350
    },
    {
      "entropy": 5.197853231430054,
      "epoch": 2.62776176753122,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043125483308298053,
      "loss": 4.8536,
      "mean_token_accuracy": 0.2288282632827759,
      "num_tokens": 62728824.0,
      "step": 27355
    },
    {
      "entropy": 5.056105756759644,
      "epoch": 2.628242074927954,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004312301633735895,
      "loss": 4.7068,
      "mean_token_accuracy": 0.23864805847406387,
      "num_tokens": 62740909.0,
      "step": 27360
    },
    {
      "entropy": 5.095552539825439,
      "epoch": 2.6287223823246877,
      "grad_norm": 1.0,
      "learning_rate": 0.0004312054900369981,
      "loss": 4.7431,
      "mean_token_accuracy": 0.23609407097101212,
      "num_tokens": 62751934.0,
      "step": 27365
    },
    {
      "entropy": 5.158689165115357,
      "epoch": 2.6292026897214216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000431180813073779,
      "loss": 4.809,
      "mean_token_accuracy": 0.23032389879226683,
      "num_tokens": 62762975.0,
      "step": 27370
    },
    {
      "entropy": 5.20560131072998,
      "epoch": 2.629682997118156,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004311561324845051,
      "loss": 4.8328,
      "mean_token_accuracy": 0.2272477477788925,
      "num_tokens": 62774880.0,
      "step": 27375
    },
    {
      "entropy": 5.127796459197998,
      "epoch": 2.6301633045148893,
      "grad_norm": 1.0,
      "learning_rate": 0.0004311314482697496,
      "loss": 4.7608,
      "mean_token_accuracy": 0.2422165185213089,
      "num_tokens": 62787372.0,
      "step": 27380
    },
    {
      "entropy": 5.125594663619995,
      "epoch": 2.6306436119116237,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004311067604300855,
      "loss": 4.8105,
      "mean_token_accuracy": 0.23594184517860411,
      "num_tokens": 62798437.0,
      "step": 27385
    },
    {
      "entropy": 5.087652683258057,
      "epoch": 2.631123919308357,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004310820689660859,
      "loss": 4.7559,
      "mean_token_accuracy": 0.23274718075990677,
      "num_tokens": 62810406.0,
      "step": 27390
    },
    {
      "entropy": 5.1609093189239506,
      "epoch": 2.6316042267050914,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004310573738783242,
      "loss": 4.8324,
      "mean_token_accuracy": 0.2313787519931793,
      "num_tokens": 62822334.0,
      "step": 27395
    },
    {
      "entropy": 5.16923828125,
      "epoch": 2.6320845341018253,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004310326751673736,
      "loss": 4.8319,
      "mean_token_accuracy": 0.23069619238376618,
      "num_tokens": 62834601.0,
      "step": 27400
    },
    {
      "entropy": 5.182215690612793,
      "epoch": 2.632564841498559,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043100797283380756,
      "loss": 4.8315,
      "mean_token_accuracy": 0.23372606188058853,
      "num_tokens": 62847086.0,
      "step": 27405
    },
    {
      "entropy": 5.14840259552002,
      "epoch": 2.633045148895293,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043098326687819973,
      "loss": 4.7451,
      "mean_token_accuracy": 0.2347632095217705,
      "num_tokens": 62858658.0,
      "step": 27410
    },
    {
      "entropy": 5.139940977096558,
      "epoch": 2.633525456292027,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004309585573011236,
      "loss": 4.792,
      "mean_token_accuracy": 0.23298866897821427,
      "num_tokens": 62870731.0,
      "step": 27415
    },
    {
      "entropy": 5.152122688293457,
      "epoch": 2.634005763688761,
      "grad_norm": 1.0,
      "learning_rate": 0.0004309338441031528,
      "loss": 4.8865,
      "mean_token_accuracy": 0.23030110597610473,
      "num_tokens": 62880848.0,
      "step": 27420
    },
    {
      "entropy": 5.135837745666504,
      "epoch": 2.6344860710854947,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043090912728486135,
      "loss": 4.8012,
      "mean_token_accuracy": 0.23687166273593901,
      "num_tokens": 62892367.0,
      "step": 27425
    },
    {
      "entropy": 5.1409564971923825,
      "epoch": 2.6349663784822286,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004308844068468228,
      "loss": 4.828,
      "mean_token_accuracy": 0.23161530792713164,
      "num_tokens": 62903764.0,
      "step": 27430
    },
    {
      "entropy": 5.214621448516846,
      "epoch": 2.6354466858789625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043085968278961116,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22243367731571198,
      "num_tokens": 62914502.0,
      "step": 27435
    },
    {
      "entropy": 5.221885871887207,
      "epoch": 2.6359269932756964,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043083495511380055,
      "loss": 4.9822,
      "mean_token_accuracy": 0.22361674159765244,
      "num_tokens": 62924793.0,
      "step": 27440
    },
    {
      "entropy": 5.092361497879028,
      "epoch": 2.6364073006724302,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043081022381996506,
      "loss": 4.7761,
      "mean_token_accuracy": 0.23522073924541473,
      "num_tokens": 62936806.0,
      "step": 27445
    },
    {
      "entropy": 5.109061527252197,
      "epoch": 2.636887608069164,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004307854889086787,
      "loss": 4.7854,
      "mean_token_accuracy": 0.2387601062655449,
      "num_tokens": 62948957.0,
      "step": 27450
    },
    {
      "entropy": 5.147073650360108,
      "epoch": 2.637367915465898,
      "grad_norm": 0.890625,
      "learning_rate": 0.00043076075038051605,
      "loss": 4.8006,
      "mean_token_accuracy": 0.23465954810380935,
      "num_tokens": 62960983.0,
      "step": 27455
    },
    {
      "entropy": 5.056014633178711,
      "epoch": 2.6378482228626323,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004307360082360512,
      "loss": 4.7152,
      "mean_token_accuracy": 0.24575963467359543,
      "num_tokens": 62972155.0,
      "step": 27460
    },
    {
      "entropy": 5.15306715965271,
      "epoch": 2.6383285302593658,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043071126247585866,
      "loss": 4.8381,
      "mean_token_accuracy": 0.2345609575510025,
      "num_tokens": 62982928.0,
      "step": 27465
    },
    {
      "entropy": 5.1249596118927006,
      "epoch": 2.6388088376561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000430686513100513,
      "loss": 4.7367,
      "mean_token_accuracy": 0.2279765084385872,
      "num_tokens": 62993326.0,
      "step": 27470
    },
    {
      "entropy": 5.247874021530151,
      "epoch": 2.639289145052834,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043066176011058877,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2185593292117119,
      "num_tokens": 63006512.0,
      "step": 27475
    },
    {
      "entropy": 5.21947832107544,
      "epoch": 2.639769452449568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043063700350666066,
      "loss": 4.8589,
      "mean_token_accuracy": 0.234561587870121,
      "num_tokens": 63017439.0,
      "step": 27480
    },
    {
      "entropy": 5.158090257644654,
      "epoch": 2.6402497598463017,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004306122432893036,
      "loss": 4.7508,
      "mean_token_accuracy": 0.2437572345137596,
      "num_tokens": 63028186.0,
      "step": 27485
    },
    {
      "entropy": 5.066832256317139,
      "epoch": 2.6407300672430356,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043058747945909224,
      "loss": 4.8053,
      "mean_token_accuracy": 0.23804315477609633,
      "num_tokens": 63039864.0,
      "step": 27490
    },
    {
      "entropy": 5.1401807308197025,
      "epoch": 2.6412103746397695,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043056271201660166,
      "loss": 4.8073,
      "mean_token_accuracy": 0.2341104969382286,
      "num_tokens": 63051965.0,
      "step": 27495
    },
    {
      "entropy": 5.13367919921875,
      "epoch": 2.6416906820365034,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004305379409624068,
      "loss": 4.8269,
      "mean_token_accuracy": 0.23455562442541122,
      "num_tokens": 63063116.0,
      "step": 27500
    },
    {
      "entropy": 5.140715646743774,
      "epoch": 2.6421709894332372,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004305131662970827,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22875402718782425,
      "num_tokens": 63074680.0,
      "step": 27505
    },
    {
      "entropy": 5.128168678283691,
      "epoch": 2.642651296829971,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004304883880212048,
      "loss": 4.7782,
      "mean_token_accuracy": 0.23594411611557006,
      "num_tokens": 63086470.0,
      "step": 27510
    },
    {
      "entropy": 5.215835905075073,
      "epoch": 2.643131604226705,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004304636061353482,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22843779921531676,
      "num_tokens": 63097616.0,
      "step": 27515
    },
    {
      "entropy": 5.158534574508667,
      "epoch": 2.643611911623439,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004304388206400883,
      "loss": 4.8392,
      "mean_token_accuracy": 0.22980901300907136,
      "num_tokens": 63109792.0,
      "step": 27520
    },
    {
      "entropy": 5.199593830108642,
      "epoch": 2.6440922190201728,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004304140315360005,
      "loss": 4.9433,
      "mean_token_accuracy": 0.2176576405763626,
      "num_tokens": 63122491.0,
      "step": 27525
    },
    {
      "entropy": 5.22573037147522,
      "epoch": 2.6445725264169067,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004303892388236604,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22074507921934128,
      "num_tokens": 63134042.0,
      "step": 27530
    },
    {
      "entropy": 5.1083780288696286,
      "epoch": 2.645052833813641,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004303644425036436,
      "loss": 4.6809,
      "mean_token_accuracy": 0.239445661008358,
      "num_tokens": 63145885.0,
      "step": 27535
    },
    {
      "entropy": 5.103698873519898,
      "epoch": 2.6455331412103744,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043033964257652575,
      "loss": 4.7641,
      "mean_token_accuracy": 0.23355616480112076,
      "num_tokens": 63156995.0,
      "step": 27540
    },
    {
      "entropy": 5.120297431945801,
      "epoch": 2.6460134486071087,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004303148390428827,
      "loss": 4.872,
      "mean_token_accuracy": 0.22660018652677535,
      "num_tokens": 63168604.0,
      "step": 27545
    },
    {
      "entropy": 5.212206888198852,
      "epoch": 2.6464937560038426,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043029003190329023,
      "loss": 4.7775,
      "mean_token_accuracy": 0.23038243353366852,
      "num_tokens": 63181284.0,
      "step": 27550
    },
    {
      "entropy": 5.0840739727020265,
      "epoch": 2.6469740634005765,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004302652211583244,
      "loss": 4.722,
      "mean_token_accuracy": 0.23690662831068038,
      "num_tokens": 63193760.0,
      "step": 27555
    },
    {
      "entropy": 5.159840440750122,
      "epoch": 2.6474543707973104,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004302404068085612,
      "loss": 4.8156,
      "mean_token_accuracy": 0.24142950177192687,
      "num_tokens": 63205280.0,
      "step": 27560
    },
    {
      "entropy": 5.0943375587463375,
      "epoch": 2.6479346781940443,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004302155888545766,
      "loss": 4.6674,
      "mean_token_accuracy": 0.24367872029542922,
      "num_tokens": 63216253.0,
      "step": 27565
    },
    {
      "entropy": 5.081588935852051,
      "epoch": 2.648414985590778,
      "grad_norm": 1.03125,
      "learning_rate": 0.000430190767296947,
      "loss": 4.8404,
      "mean_token_accuracy": 0.23172563314437866,
      "num_tokens": 63227341.0,
      "step": 27570
    },
    {
      "entropy": 5.131738615036011,
      "epoch": 2.648895292987512,
      "grad_norm": 1.0,
      "learning_rate": 0.0004301659421362486,
      "loss": 4.8067,
      "mean_token_accuracy": 0.23232870697975158,
      "num_tokens": 63239090.0,
      "step": 27575
    },
    {
      "entropy": 5.152169942855835,
      "epoch": 2.649375600384246,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004301411133730578,
      "loss": 4.7774,
      "mean_token_accuracy": 0.24474380910396576,
      "num_tokens": 63250077.0,
      "step": 27580
    },
    {
      "entropy": 5.098544359207153,
      "epoch": 2.64985590778098,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043011628100795093,
      "loss": 4.7429,
      "mean_token_accuracy": 0.2302865743637085,
      "num_tokens": 63261177.0,
      "step": 27585
    },
    {
      "entropy": 5.071889495849609,
      "epoch": 2.6503362151777137,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004300914450415047,
      "loss": 4.7744,
      "mean_token_accuracy": 0.24236011058092116,
      "num_tokens": 63273176.0,
      "step": 27590
    },
    {
      "entropy": 5.135448980331421,
      "epoch": 2.6508165225744476,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043006660547429565,
      "loss": 4.8453,
      "mean_token_accuracy": 0.2319052442908287,
      "num_tokens": 63284529.0,
      "step": 27595
    },
    {
      "entropy": 5.04141058921814,
      "epoch": 2.6512968299711814,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004300417623069005,
      "loss": 4.6417,
      "mean_token_accuracy": 0.24780309200286865,
      "num_tokens": 63295824.0,
      "step": 27600
    },
    {
      "entropy": 5.07869668006897,
      "epoch": 2.6517771373679153,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004300169155398959,
      "loss": 4.792,
      "mean_token_accuracy": 0.23743869960308076,
      "num_tokens": 63307267.0,
      "step": 27605
    },
    {
      "entropy": 5.222121620178223,
      "epoch": 2.6522574447646496,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042999206517385885,
      "loss": 4.8515,
      "mean_token_accuracy": 0.23492206931114196,
      "num_tokens": 63317703.0,
      "step": 27610
    },
    {
      "entropy": 5.1220865726470945,
      "epoch": 2.652737752161383,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004299672112093663,
      "loss": 4.8652,
      "mean_token_accuracy": 0.22955939471721648,
      "num_tokens": 63328677.0,
      "step": 27615
    },
    {
      "entropy": 5.221644353866577,
      "epoch": 2.6532180595581174,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042994235364699526,
      "loss": 4.9374,
      "mean_token_accuracy": 0.2230439603328705,
      "num_tokens": 63339604.0,
      "step": 27620
    },
    {
      "entropy": 5.20279974937439,
      "epoch": 2.653698366954851,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004299174924873229,
      "loss": 4.8677,
      "mean_token_accuracy": 0.22564025074243546,
      "num_tokens": 63351428.0,
      "step": 27625
    },
    {
      "entropy": 5.193209600448609,
      "epoch": 2.654178674351585,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004298926277309263,
      "loss": 4.9035,
      "mean_token_accuracy": 0.226796393096447,
      "num_tokens": 63364545.0,
      "step": 27630
    },
    {
      "entropy": 5.145886182785034,
      "epoch": 2.654658981748319,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042986775937838283,
      "loss": 4.7487,
      "mean_token_accuracy": 0.23875180035829544,
      "num_tokens": 63376063.0,
      "step": 27635
    },
    {
      "entropy": 5.106722402572632,
      "epoch": 2.655139289145053,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004298428874302699,
      "loss": 4.7297,
      "mean_token_accuracy": 0.24668105691671371,
      "num_tokens": 63387448.0,
      "step": 27640
    },
    {
      "entropy": 5.092085647583008,
      "epoch": 2.655619596541787,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004298180118871649,
      "loss": 4.8585,
      "mean_token_accuracy": 0.22762174308300018,
      "num_tokens": 63399674.0,
      "step": 27645
    },
    {
      "entropy": 5.11650538444519,
      "epoch": 2.6560999039385207,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042979313274964535,
      "loss": 4.8366,
      "mean_token_accuracy": 0.23271109908819199,
      "num_tokens": 63410622.0,
      "step": 27650
    },
    {
      "entropy": 5.17648868560791,
      "epoch": 2.6565802113352546,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042976825001828897,
      "loss": 4.8295,
      "mean_token_accuracy": 0.23163264989852905,
      "num_tokens": 63422354.0,
      "step": 27655
    },
    {
      "entropy": 5.162672472000122,
      "epoch": 2.6570605187319885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042974336369367333,
      "loss": 4.8009,
      "mean_token_accuracy": 0.23334124833345413,
      "num_tokens": 63433232.0,
      "step": 27660
    },
    {
      "entropy": 5.028604507446289,
      "epoch": 2.6575408261287223,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004297184737763763,
      "loss": 4.7152,
      "mean_token_accuracy": 0.24105548113584518,
      "num_tokens": 63444998.0,
      "step": 27665
    },
    {
      "entropy": 5.121984481811523,
      "epoch": 2.658021133525456,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042969358026697567,
      "loss": 4.8586,
      "mean_token_accuracy": 0.22442815601825714,
      "num_tokens": 63455846.0,
      "step": 27670
    },
    {
      "entropy": 5.193108797073364,
      "epoch": 2.65850144092219,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004296686831660495,
      "loss": 4.8172,
      "mean_token_accuracy": 0.23135393410921096,
      "num_tokens": 63468353.0,
      "step": 27675
    },
    {
      "entropy": 5.178035068511963,
      "epoch": 2.658981748318924,
      "grad_norm": 1.0,
      "learning_rate": 0.0004296437824741758,
      "loss": 4.7757,
      "mean_token_accuracy": 0.2371852308511734,
      "num_tokens": 63479802.0,
      "step": 27680
    },
    {
      "entropy": 5.1749382495880125,
      "epoch": 2.659462055715658,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042961887819193263,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22459405958652495,
      "num_tokens": 63490856.0,
      "step": 27685
    },
    {
      "entropy": 5.210006046295166,
      "epoch": 2.6599423631123917,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004295939703198983,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22611370384693147,
      "num_tokens": 63502294.0,
      "step": 27690
    },
    {
      "entropy": 5.227022790908814,
      "epoch": 2.660422670509126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000429569058858651,
      "loss": 4.8795,
      "mean_token_accuracy": 0.2287968710064888,
      "num_tokens": 63513466.0,
      "step": 27695
    },
    {
      "entropy": 5.124080801010132,
      "epoch": 2.6609029779058595,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042954414380876906,
      "loss": 4.7124,
      "mean_token_accuracy": 0.24017832726240157,
      "num_tokens": 63523596.0,
      "step": 27700
    },
    {
      "entropy": 5.076009368896484,
      "epoch": 2.661383285302594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042951922517083104,
      "loss": 4.7758,
      "mean_token_accuracy": 0.23295149058103562,
      "num_tokens": 63534468.0,
      "step": 27705
    },
    {
      "entropy": 5.124093818664551,
      "epoch": 2.6618635926993277,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004294943029454155,
      "loss": 4.7819,
      "mean_token_accuracy": 0.23766618669033052,
      "num_tokens": 63546273.0,
      "step": 27710
    },
    {
      "entropy": 5.100502681732178,
      "epoch": 2.6623439000960616,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042946937713310093,
      "loss": 4.794,
      "mean_token_accuracy": 0.22918396741151809,
      "num_tokens": 63557032.0,
      "step": 27715
    },
    {
      "entropy": 5.155602169036865,
      "epoch": 2.6628242074927955,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004294444477344661,
      "loss": 4.825,
      "mean_token_accuracy": 0.23825148344039918,
      "num_tokens": 63568358.0,
      "step": 27720
    },
    {
      "entropy": 5.241380500793457,
      "epoch": 2.6633045148895294,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004294195147500898,
      "loss": 4.9205,
      "mean_token_accuracy": 0.23008621335029603,
      "num_tokens": 63579716.0,
      "step": 27725
    },
    {
      "entropy": 5.063503122329712,
      "epoch": 2.6637848222862632,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042939457818055095,
      "loss": 4.7597,
      "mean_token_accuracy": 0.2394974187016487,
      "num_tokens": 63590871.0,
      "step": 27730
    },
    {
      "entropy": 5.189012956619263,
      "epoch": 2.664265129682997,
      "grad_norm": 1.0,
      "learning_rate": 0.00042936963802642843,
      "loss": 4.8416,
      "mean_token_accuracy": 0.23244602382183074,
      "num_tokens": 63600813.0,
      "step": 27735
    },
    {
      "entropy": 5.059493112564087,
      "epoch": 2.664745437079731,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004293446942883013,
      "loss": 4.7064,
      "mean_token_accuracy": 0.23645571917295455,
      "num_tokens": 63611141.0,
      "step": 27740
    },
    {
      "entropy": 5.068724536895752,
      "epoch": 2.665225744476465,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042931974696674866,
      "loss": 4.7188,
      "mean_token_accuracy": 0.2382852092385292,
      "num_tokens": 63622172.0,
      "step": 27745
    },
    {
      "entropy": 5.110167169570923,
      "epoch": 2.6657060518731988,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004292947960623497,
      "loss": 4.8337,
      "mean_token_accuracy": 0.22549981325864793,
      "num_tokens": 63634369.0,
      "step": 27750
    },
    {
      "entropy": 5.07058539390564,
      "epoch": 2.6661863592699326,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042926984157568384,
      "loss": 4.7529,
      "mean_token_accuracy": 0.24388787448406218,
      "num_tokens": 63645142.0,
      "step": 27755
    },
    {
      "entropy": 5.096399402618408,
      "epoch": 2.6666666666666665,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042924488350733024,
      "loss": 4.7579,
      "mean_token_accuracy": 0.2409048020839691,
      "num_tokens": 63656497.0,
      "step": 27760
    },
    {
      "entropy": 5.207997417449951,
      "epoch": 2.6671469740634004,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042921992185786847,
      "loss": 4.8705,
      "mean_token_accuracy": 0.23002219796180726,
      "num_tokens": 63667795.0,
      "step": 27765
    },
    {
      "entropy": 5.093822860717774,
      "epoch": 2.6676272814601347,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042919495662787813,
      "loss": 4.8463,
      "mean_token_accuracy": 0.2357943043112755,
      "num_tokens": 63679676.0,
      "step": 27770
    },
    {
      "entropy": 5.210189437866211,
      "epoch": 2.668107588856868,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004291699878179387,
      "loss": 4.8066,
      "mean_token_accuracy": 0.22515212297439574,
      "num_tokens": 63690749.0,
      "step": 27775
    },
    {
      "entropy": 5.109786081314087,
      "epoch": 2.6685878962536025,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004291450154286299,
      "loss": 4.7285,
      "mean_token_accuracy": 0.23221942782402039,
      "num_tokens": 63702583.0,
      "step": 27780
    },
    {
      "entropy": 5.167418193817139,
      "epoch": 2.6690682036503364,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004291200394605317,
      "loss": 4.8089,
      "mean_token_accuracy": 0.23035950511693953,
      "num_tokens": 63715296.0,
      "step": 27785
    },
    {
      "entropy": 5.131178283691407,
      "epoch": 2.6695485110470702,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004290950599142237,
      "loss": 4.7323,
      "mean_token_accuracy": 0.23914785087108612,
      "num_tokens": 63726261.0,
      "step": 27790
    },
    {
      "entropy": 5.173887395858765,
      "epoch": 2.670028818443804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000429070076790286,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2235393851995468,
      "num_tokens": 63738482.0,
      "step": 27795
    },
    {
      "entropy": 5.156417465209961,
      "epoch": 2.670509125840538,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042904509008929873,
      "loss": 4.8955,
      "mean_token_accuracy": 0.22595921903848648,
      "num_tokens": 63748899.0,
      "step": 27800
    },
    {
      "entropy": 5.239306449890137,
      "epoch": 2.670989433237272,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004290200998118417,
      "loss": 4.8452,
      "mean_token_accuracy": 0.23215727955102922,
      "num_tokens": 63760427.0,
      "step": 27805
    },
    {
      "entropy": 5.143859958648681,
      "epoch": 2.6714697406340058,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042899510595849544,
      "loss": 4.8022,
      "mean_token_accuracy": 0.2422550991177559,
      "num_tokens": 63772226.0,
      "step": 27810
    },
    {
      "entropy": 5.1988890171051025,
      "epoch": 2.6719500480307397,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042897010852984004,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22931375950574875,
      "num_tokens": 63782918.0,
      "step": 27815
    },
    {
      "entropy": 5.1696959972381595,
      "epoch": 2.6724303554274735,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042894510752645586,
      "loss": 4.7693,
      "mean_token_accuracy": 0.22985336035490037,
      "num_tokens": 63793461.0,
      "step": 27820
    },
    {
      "entropy": 5.070625352859497,
      "epoch": 2.6729106628242074,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004289201029489235,
      "loss": 4.7295,
      "mean_token_accuracy": 0.24263384342193603,
      "num_tokens": 63804648.0,
      "step": 27825
    },
    {
      "entropy": 5.136661434173584,
      "epoch": 2.6733909702209413,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004288950947978234,
      "loss": 4.7966,
      "mean_token_accuracy": 0.23035785257816316,
      "num_tokens": 63815809.0,
      "step": 27830
    },
    {
      "entropy": 5.109915494918823,
      "epoch": 2.673871277617675,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004288700830737361,
      "loss": 4.7604,
      "mean_token_accuracy": 0.23907267451286315,
      "num_tokens": 63828310.0,
      "step": 27835
    },
    {
      "entropy": 5.096502494812012,
      "epoch": 2.674351585014409,
      "grad_norm": 0.90625,
      "learning_rate": 0.00042884506777724244,
      "loss": 4.7546,
      "mean_token_accuracy": 0.23628678768873215,
      "num_tokens": 63840546.0,
      "step": 27840
    },
    {
      "entropy": 5.04342360496521,
      "epoch": 2.6748318924111434,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004288200489089231,
      "loss": 4.681,
      "mean_token_accuracy": 0.2447728157043457,
      "num_tokens": 63852049.0,
      "step": 27845
    },
    {
      "entropy": 5.086188554763794,
      "epoch": 2.675312199807877,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000428795026469359,
      "loss": 4.7728,
      "mean_token_accuracy": 0.23771732151508332,
      "num_tokens": 63863474.0,
      "step": 27850
    },
    {
      "entropy": 5.0936637878417965,
      "epoch": 2.675792507204611,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000428770000459131,
      "loss": 4.7863,
      "mean_token_accuracy": 0.23430461585521697,
      "num_tokens": 63874906.0,
      "step": 27855
    },
    {
      "entropy": 5.115778684616089,
      "epoch": 2.6762728146013446,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004287449708788202,
      "loss": 4.7852,
      "mean_token_accuracy": 0.2293238863348961,
      "num_tokens": 63885443.0,
      "step": 27860
    },
    {
      "entropy": 5.12390513420105,
      "epoch": 2.676753121998079,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004287199377290077,
      "loss": 4.8407,
      "mean_token_accuracy": 0.23341258466243744,
      "num_tokens": 63897772.0,
      "step": 27865
    },
    {
      "entropy": 5.252535963058472,
      "epoch": 2.677233429394813,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004286949010102748,
      "loss": 4.8515,
      "mean_token_accuracy": 0.2305739030241966,
      "num_tokens": 63910409.0,
      "step": 27870
    },
    {
      "entropy": 5.122838735580444,
      "epoch": 2.6777137367915467,
      "grad_norm": 1.0,
      "learning_rate": 0.0004286698607232026,
      "loss": 4.7894,
      "mean_token_accuracy": 0.23183362782001496,
      "num_tokens": 63922671.0,
      "step": 27875
    },
    {
      "entropy": 5.164465856552124,
      "epoch": 2.6781940441882806,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00042864481686837253,
      "loss": 4.8111,
      "mean_token_accuracy": 0.23661570250988007,
      "num_tokens": 63933964.0,
      "step": 27880
    },
    {
      "entropy": 5.190036678314209,
      "epoch": 2.6786743515850144,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042861976944636604,
      "loss": 4.81,
      "mean_token_accuracy": 0.2309420019388199,
      "num_tokens": 63945347.0,
      "step": 27885
    },
    {
      "entropy": 5.187971544265747,
      "epoch": 2.6791546589817483,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004285947184577647,
      "loss": 4.8236,
      "mean_token_accuracy": 0.23167232275009156,
      "num_tokens": 63957573.0,
      "step": 27890
    },
    {
      "entropy": 5.132611799240112,
      "epoch": 2.679634966378482,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042856966390315013,
      "loss": 4.7855,
      "mean_token_accuracy": 0.24109681993722915,
      "num_tokens": 63968666.0,
      "step": 27895
    },
    {
      "entropy": 5.060193395614624,
      "epoch": 2.680115273775216,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004285446057831039,
      "loss": 4.7711,
      "mean_token_accuracy": 0.23346416056156158,
      "num_tokens": 63981147.0,
      "step": 27900
    },
    {
      "entropy": 5.204530811309814,
      "epoch": 2.68059558117195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004285195440982078,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2187408834695816,
      "num_tokens": 63992474.0,
      "step": 27905
    },
    {
      "entropy": 5.178922986984253,
      "epoch": 2.681075888568684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004284944788490439,
      "loss": 4.783,
      "mean_token_accuracy": 0.2314511403441429,
      "num_tokens": 64003804.0,
      "step": 27910
    },
    {
      "entropy": 5.107702970504761,
      "epoch": 2.6815561959654177,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004284694100361938,
      "loss": 4.7899,
      "mean_token_accuracy": 0.2403019443154335,
      "num_tokens": 64015587.0,
      "step": 27915
    },
    {
      "entropy": 5.1991767406463625,
      "epoch": 2.682036503362152,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004284443376602398,
      "loss": 4.895,
      "mean_token_accuracy": 0.23047112077474594,
      "num_tokens": 64027088.0,
      "step": 27920
    },
    {
      "entropy": 5.172715902328491,
      "epoch": 2.6825168107588855,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004284192617217639,
      "loss": 4.7549,
      "mean_token_accuracy": 0.24177204072475433,
      "num_tokens": 64037050.0,
      "step": 27925
    },
    {
      "entropy": 5.06893949508667,
      "epoch": 2.68299711815562,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004283941822213484,
      "loss": 4.8178,
      "mean_token_accuracy": 0.23165417611598968,
      "num_tokens": 64049287.0,
      "step": 27930
    },
    {
      "entropy": 5.168262004852295,
      "epoch": 2.6834774255523532,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004283690991595754,
      "loss": 4.7988,
      "mean_token_accuracy": 0.23691990971565247,
      "num_tokens": 64060578.0,
      "step": 27935
    },
    {
      "entropy": 5.202227163314819,
      "epoch": 2.6839577329490876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042834401253702734,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22820103764533997,
      "num_tokens": 64071540.0,
      "step": 27940
    },
    {
      "entropy": 5.1628419876098635,
      "epoch": 2.6844380403458215,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004283189223542866,
      "loss": 4.8167,
      "mean_token_accuracy": 0.23137460201978682,
      "num_tokens": 64084014.0,
      "step": 27945
    },
    {
      "entropy": 5.156031131744385,
      "epoch": 2.6849183477425553,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042829382861193585,
      "loss": 4.8491,
      "mean_token_accuracy": 0.233239109814167,
      "num_tokens": 64096735.0,
      "step": 27950
    },
    {
      "entropy": 5.220812845230102,
      "epoch": 2.685398655139289,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004282687313105575,
      "loss": 4.8932,
      "mean_token_accuracy": 0.22446609139442444,
      "num_tokens": 64107548.0,
      "step": 27955
    },
    {
      "entropy": 5.146296882629395,
      "epoch": 2.685878962536023,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042824363045073434,
      "loss": 4.8258,
      "mean_token_accuracy": 0.23478048294782639,
      "num_tokens": 64118433.0,
      "step": 27960
    },
    {
      "entropy": 5.138224935531616,
      "epoch": 2.686359269932757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004282185260330491,
      "loss": 4.8021,
      "mean_token_accuracy": 0.23732737451791763,
      "num_tokens": 64129883.0,
      "step": 27965
    },
    {
      "entropy": 5.152202796936035,
      "epoch": 2.686839577329491,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042819341805808473,
      "loss": 4.8579,
      "mean_token_accuracy": 0.23539561182260513,
      "num_tokens": 64141827.0,
      "step": 27970
    },
    {
      "entropy": 5.141028213500976,
      "epoch": 2.6873198847262247,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042816830652642396,
      "loss": 4.7696,
      "mean_token_accuracy": 0.23352108299732208,
      "num_tokens": 64152957.0,
      "step": 27975
    },
    {
      "entropy": 5.0870774269104,
      "epoch": 2.6878001921229586,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004281431914386501,
      "loss": 4.7483,
      "mean_token_accuracy": 0.23787833601236344,
      "num_tokens": 64164800.0,
      "step": 27980
    },
    {
      "entropy": 5.186512517929077,
      "epoch": 2.6882804995196925,
      "grad_norm": 1.078125,
      "learning_rate": 0.000428118072795346,
      "loss": 4.8277,
      "mean_token_accuracy": 0.23269503712654113,
      "num_tokens": 64175941.0,
      "step": 27985
    },
    {
      "entropy": 5.147772789001465,
      "epoch": 2.6887608069164264,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042809295059709483,
      "loss": 4.7684,
      "mean_token_accuracy": 0.2378661721944809,
      "num_tokens": 64187229.0,
      "step": 27990
    },
    {
      "entropy": 5.1485682964324955,
      "epoch": 2.6892411143131603,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042806782484448,
      "loss": 4.8194,
      "mean_token_accuracy": 0.2367240995168686,
      "num_tokens": 64198328.0,
      "step": 27995
    },
    {
      "entropy": 5.132372045516968,
      "epoch": 2.689721421709894,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004280426955380848,
      "loss": 4.7268,
      "mean_token_accuracy": 0.2322726473212242,
      "num_tokens": 64207434.0,
      "step": 28000
    },
    {
      "entropy": 5.064221286773682,
      "epoch": 2.6902017291066285,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00042801756267849266,
      "loss": 4.7536,
      "mean_token_accuracy": 0.24125337451696396,
      "num_tokens": 64219128.0,
      "step": 28005
    },
    {
      "entropy": 5.104213905334473,
      "epoch": 2.690682036503362,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004279924262662871,
      "loss": 4.7908,
      "mean_token_accuracy": 0.23313400596380235,
      "num_tokens": 64230664.0,
      "step": 28010
    },
    {
      "entropy": 5.153283023834229,
      "epoch": 2.6911623439000962,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004279672863020517,
      "loss": 4.7625,
      "mean_token_accuracy": 0.24451308846473693,
      "num_tokens": 64242339.0,
      "step": 28015
    },
    {
      "entropy": 5.134478998184204,
      "epoch": 2.69164265129683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042794214278637013,
      "loss": 4.8173,
      "mean_token_accuracy": 0.22874912172555922,
      "num_tokens": 64253005.0,
      "step": 28020
    },
    {
      "entropy": 5.103502464294434,
      "epoch": 2.692122958693564,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042791699571982606,
      "loss": 4.8399,
      "mean_token_accuracy": 0.23253692239522933,
      "num_tokens": 64264791.0,
      "step": 28025
    },
    {
      "entropy": 5.1416090965271,
      "epoch": 2.692603266090298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004278918451030035,
      "loss": 4.761,
      "mean_token_accuracy": 0.23730180114507676,
      "num_tokens": 64274501.0,
      "step": 28030
    },
    {
      "entropy": 5.1028289794921875,
      "epoch": 2.6930835734870318,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004278666909364863,
      "loss": 4.7265,
      "mean_token_accuracy": 0.24087228626012802,
      "num_tokens": 64284501.0,
      "step": 28035
    },
    {
      "entropy": 5.081136417388916,
      "epoch": 2.6935638808837656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004278415332208584,
      "loss": 4.7553,
      "mean_token_accuracy": 0.24015955328941346,
      "num_tokens": 64295627.0,
      "step": 28040
    },
    {
      "entropy": 5.2132330417633055,
      "epoch": 2.6940441882804995,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042781637195670396,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2257955253124237,
      "num_tokens": 64307060.0,
      "step": 28045
    },
    {
      "entropy": 5.179930305480957,
      "epoch": 2.6945244956772334,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004277912071446072,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2363557279109955,
      "num_tokens": 64318329.0,
      "step": 28050
    },
    {
      "entropy": 5.174305248260498,
      "epoch": 2.6950048030739673,
      "grad_norm": 1.0,
      "learning_rate": 0.0004277660387851522,
      "loss": 4.8086,
      "mean_token_accuracy": 0.23666671067476272,
      "num_tokens": 64329871.0,
      "step": 28055
    },
    {
      "entropy": 5.075190782546997,
      "epoch": 2.695485110470701,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004277408668789235,
      "loss": 4.775,
      "mean_token_accuracy": 0.23784157633781433,
      "num_tokens": 64341950.0,
      "step": 28060
    },
    {
      "entropy": 5.146394300460815,
      "epoch": 2.695965417867435,
      "grad_norm": 1.0,
      "learning_rate": 0.0004277156914265054,
      "loss": 4.852,
      "mean_token_accuracy": 0.23663763999938964,
      "num_tokens": 64353841.0,
      "step": 28065
    },
    {
      "entropy": 5.149656963348389,
      "epoch": 2.696445725264169,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004276905124284824,
      "loss": 4.7625,
      "mean_token_accuracy": 0.2391700729727745,
      "num_tokens": 64365914.0,
      "step": 28070
    },
    {
      "entropy": 5.081701850891113,
      "epoch": 2.696926032660903,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004276653298854391,
      "loss": 4.7157,
      "mean_token_accuracy": 0.23959091007709504,
      "num_tokens": 64376321.0,
      "step": 28075
    },
    {
      "entropy": 5.111867141723633,
      "epoch": 2.697406340057637,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004276401437979601,
      "loss": 4.8488,
      "mean_token_accuracy": 0.22764874398708343,
      "num_tokens": 64387344.0,
      "step": 28080
    },
    {
      "entropy": 5.098757362365722,
      "epoch": 2.6978866474543706,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004276149541666303,
      "loss": 4.7477,
      "mean_token_accuracy": 0.2388955146074295,
      "num_tokens": 64398553.0,
      "step": 28085
    },
    {
      "entropy": 5.247788381576538,
      "epoch": 2.698366954851105,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042758976099203444,
      "loss": 4.8904,
      "mean_token_accuracy": 0.23090460002422333,
      "num_tokens": 64409174.0,
      "step": 28090
    },
    {
      "entropy": 5.124149560928345,
      "epoch": 2.6988472622478388,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042756456427475736,
      "loss": 4.7007,
      "mean_token_accuracy": 0.2404904067516327,
      "num_tokens": 64420840.0,
      "step": 28095
    },
    {
      "entropy": 5.097295761108398,
      "epoch": 2.6993275696445727,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004275393640153842,
      "loss": 4.7058,
      "mean_token_accuracy": 0.2397788494825363,
      "num_tokens": 64431796.0,
      "step": 28100
    },
    {
      "entropy": 5.1340985774993895,
      "epoch": 2.6998078770413065,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042751416021449986,
      "loss": 4.7952,
      "mean_token_accuracy": 0.23072203993797302,
      "num_tokens": 64443675.0,
      "step": 28105
    },
    {
      "entropy": 5.14805006980896,
      "epoch": 2.7002881844380404,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004274889528726896,
      "loss": 4.7771,
      "mean_token_accuracy": 0.24536802917718886,
      "num_tokens": 64455556.0,
      "step": 28110
    },
    {
      "entropy": 5.116073799133301,
      "epoch": 2.7007684918347743,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004274637419905388,
      "loss": 4.7761,
      "mean_token_accuracy": 0.23974452912807465,
      "num_tokens": 64466873.0,
      "step": 28115
    },
    {
      "entropy": 5.151073312759399,
      "epoch": 2.701248799231508,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042743852756863253,
      "loss": 4.8151,
      "mean_token_accuracy": 0.2377607211470604,
      "num_tokens": 64478530.0,
      "step": 28120
    },
    {
      "entropy": 5.147063827514648,
      "epoch": 2.701729106628242,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004274133096075563,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23387056589126587,
      "num_tokens": 64488941.0,
      "step": 28125
    },
    {
      "entropy": 5.070276880264283,
      "epoch": 2.702209414024976,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004273880881078956,
      "loss": 4.7281,
      "mean_token_accuracy": 0.23475055247545243,
      "num_tokens": 64500727.0,
      "step": 28130
    },
    {
      "entropy": 5.0977226257324215,
      "epoch": 2.70268972142171,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004273628630702359,
      "loss": 4.7872,
      "mean_token_accuracy": 0.237339848279953,
      "num_tokens": 64513381.0,
      "step": 28135
    },
    {
      "entropy": 5.15371561050415,
      "epoch": 2.7031700288184437,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042733763449516313,
      "loss": 4.8101,
      "mean_token_accuracy": 0.23506758660078048,
      "num_tokens": 64525442.0,
      "step": 28140
    },
    {
      "entropy": 5.196766948699951,
      "epoch": 2.7036503362151776,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042731240238326273,
      "loss": 4.7457,
      "mean_token_accuracy": 0.23184428811073304,
      "num_tokens": 64536008.0,
      "step": 28145
    },
    {
      "entropy": 5.166210699081421,
      "epoch": 2.7041306436119115,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042728716673512065,
      "loss": 4.8894,
      "mean_token_accuracy": 0.224807970225811,
      "num_tokens": 64547161.0,
      "step": 28150
    },
    {
      "entropy": 5.142791700363159,
      "epoch": 2.704610951008646,
      "grad_norm": 1.125,
      "learning_rate": 0.00042726192755132276,
      "loss": 4.8014,
      "mean_token_accuracy": 0.23260540068149566,
      "num_tokens": 64558152.0,
      "step": 28155
    },
    {
      "entropy": 5.187363910675049,
      "epoch": 2.7050912584053792,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042723668483245496,
      "loss": 4.9708,
      "mean_token_accuracy": 0.22354070097208023,
      "num_tokens": 64569150.0,
      "step": 28160
    },
    {
      "entropy": 5.239336252212524,
      "epoch": 2.7055715658021136,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004272114385791035,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23974950164556502,
      "num_tokens": 64580417.0,
      "step": 28165
    },
    {
      "entropy": 5.1511060237884525,
      "epoch": 2.706051873198847,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042718618879185435,
      "loss": 4.7588,
      "mean_token_accuracy": 0.2344699814915657,
      "num_tokens": 64591628.0,
      "step": 28170
    },
    {
      "entropy": 5.1081644058227536,
      "epoch": 2.7065321805955813,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271609354712938,
      "loss": 4.7767,
      "mean_token_accuracy": 0.23696964979171753,
      "num_tokens": 64602801.0,
      "step": 28175
    },
    {
      "entropy": 5.16285457611084,
      "epoch": 2.707012487992315,
      "grad_norm": 1.0,
      "learning_rate": 0.0004271356786180082,
      "loss": 4.8139,
      "mean_token_accuracy": 0.23519274592399597,
      "num_tokens": 64614327.0,
      "step": 28180
    },
    {
      "entropy": 5.165515422821045,
      "epoch": 2.707492795389049,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004271104182325838,
      "loss": 4.8319,
      "mean_token_accuracy": 0.2342700645327568,
      "num_tokens": 64625508.0,
      "step": 28185
    },
    {
      "entropy": 5.1290205955505375,
      "epoch": 2.707973102785783,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042708515431560723,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23102996051311492,
      "num_tokens": 64637428.0,
      "step": 28190
    },
    {
      "entropy": 5.067831230163574,
      "epoch": 2.708453410182517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000427059886867665,
      "loss": 4.7966,
      "mean_token_accuracy": 0.23323487788438796,
      "num_tokens": 64648189.0,
      "step": 28195
    },
    {
      "entropy": 5.103651666641236,
      "epoch": 2.7089337175792507,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004270346158893436,
      "loss": 4.7536,
      "mean_token_accuracy": 0.23623257428407668,
      "num_tokens": 64659250.0,
      "step": 28200
    },
    {
      "entropy": 5.099956703186035,
      "epoch": 2.7094140249759846,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042700934138123004,
      "loss": 4.7392,
      "mean_token_accuracy": 0.2481663316488266,
      "num_tokens": 64669746.0,
      "step": 28205
    },
    {
      "entropy": 5.131474876403809,
      "epoch": 2.7098943323727185,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042698406334391084,
      "loss": 4.7643,
      "mean_token_accuracy": 0.2361418291926384,
      "num_tokens": 64681758.0,
      "step": 28210
    },
    {
      "entropy": 5.165578842163086,
      "epoch": 2.7103746397694524,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000426958781777973,
      "loss": 4.8325,
      "mean_token_accuracy": 0.23686257600784302,
      "num_tokens": 64692944.0,
      "step": 28215
    },
    {
      "entropy": 5.253017854690552,
      "epoch": 2.7108549471661862,
      "grad_norm": 1.0,
      "learning_rate": 0.0004269334966840035,
      "loss": 4.9552,
      "mean_token_accuracy": 0.22338095754384996,
      "num_tokens": 64705436.0,
      "step": 28220
    },
    {
      "entropy": 5.198072242736816,
      "epoch": 2.71133525456292,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042690820806258933,
      "loss": 4.8251,
      "mean_token_accuracy": 0.22887222617864608,
      "num_tokens": 64716542.0,
      "step": 28225
    },
    {
      "entropy": 5.068455648422241,
      "epoch": 2.7118155619596545,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004268829159143176,
      "loss": 4.7303,
      "mean_token_accuracy": 0.23746655732393265,
      "num_tokens": 64727800.0,
      "step": 28230
    },
    {
      "entropy": 5.25019793510437,
      "epoch": 2.712295869356388,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004268576202397757,
      "loss": 4.9343,
      "mean_token_accuracy": 0.22383637428283693,
      "num_tokens": 64739790.0,
      "step": 28235
    },
    {
      "entropy": 5.155651044845581,
      "epoch": 2.712776176753122,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004268323210395506,
      "loss": 4.7699,
      "mean_token_accuracy": 0.23505313098430633,
      "num_tokens": 64751031.0,
      "step": 28240
    },
    {
      "entropy": 5.1130640506744385,
      "epoch": 2.7132564841498557,
      "grad_norm": 1.0,
      "learning_rate": 0.00042680701831423004,
      "loss": 4.7942,
      "mean_token_accuracy": 0.23344760537147521,
      "num_tokens": 64761751.0,
      "step": 28245
    },
    {
      "entropy": 5.24121732711792,
      "epoch": 2.71373679154659,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004267817120644012,
      "loss": 4.8346,
      "mean_token_accuracy": 0.23607376962900162,
      "num_tokens": 64773257.0,
      "step": 28250
    },
    {
      "entropy": 5.2042700290679935,
      "epoch": 2.714217098943324,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042675640229065167,
      "loss": 4.7994,
      "mean_token_accuracy": 0.22694966048002244,
      "num_tokens": 64783409.0,
      "step": 28255
    },
    {
      "entropy": 5.157818269729614,
      "epoch": 2.7146974063400577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042673108899356915,
      "loss": 4.8528,
      "mean_token_accuracy": 0.2280607208609581,
      "num_tokens": 64794392.0,
      "step": 28260
    },
    {
      "entropy": 5.162093257904052,
      "epoch": 2.7151777137367916,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004267057721737413,
      "loss": 4.8412,
      "mean_token_accuracy": 0.23220451921224594,
      "num_tokens": 64806289.0,
      "step": 28265
    },
    {
      "entropy": 5.143112754821777,
      "epoch": 2.7156580211335255,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004266804518317559,
      "loss": 4.8118,
      "mean_token_accuracy": 0.23223440945148469,
      "num_tokens": 64817278.0,
      "step": 28270
    },
    {
      "entropy": 5.138401317596435,
      "epoch": 2.7161383285302594,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004266551279682008,
      "loss": 4.7559,
      "mean_token_accuracy": 0.22929703146219255,
      "num_tokens": 64828889.0,
      "step": 28275
    },
    {
      "entropy": 5.06592288017273,
      "epoch": 2.7166186359269933,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004266298005836639,
      "loss": 4.7665,
      "mean_token_accuracy": 0.2376318246126175,
      "num_tokens": 64840260.0,
      "step": 28280
    },
    {
      "entropy": 5.190047359466552,
      "epoch": 2.717098943323727,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042660446967873327,
      "loss": 4.8548,
      "mean_token_accuracy": 0.2270521730184555,
      "num_tokens": 64853099.0,
      "step": 28285
    },
    {
      "entropy": 5.200354051589966,
      "epoch": 2.717579250720461,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042657913525399703,
      "loss": 4.7766,
      "mean_token_accuracy": 0.2501361832022667,
      "num_tokens": 64863501.0,
      "step": 28290
    },
    {
      "entropy": 5.090906238555908,
      "epoch": 2.718059558117195,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004265537973100435,
      "loss": 4.7885,
      "mean_token_accuracy": 0.23902547508478164,
      "num_tokens": 64874071.0,
      "step": 28295
    },
    {
      "entropy": 5.215436410903931,
      "epoch": 2.718539865513929,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004265284558474607,
      "loss": 4.828,
      "mean_token_accuracy": 0.22860852032899856,
      "num_tokens": 64885238.0,
      "step": 28300
    },
    {
      "entropy": 5.1881101608276365,
      "epoch": 2.7190201729106627,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042650311086683715,
      "loss": 4.8938,
      "mean_token_accuracy": 0.2272648498415947,
      "num_tokens": 64896384.0,
      "step": 28305
    },
    {
      "entropy": 5.196577882766723,
      "epoch": 2.7195004803073966,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004264777623687612,
      "loss": 4.8057,
      "mean_token_accuracy": 0.23298740088939668,
      "num_tokens": 64907219.0,
      "step": 28310
    },
    {
      "entropy": 5.127232694625855,
      "epoch": 2.719980787704131,
      "grad_norm": 1.0,
      "learning_rate": 0.0004264524103538214,
      "loss": 4.7514,
      "mean_token_accuracy": 0.23539942651987075,
      "num_tokens": 64918637.0,
      "step": 28315
    },
    {
      "entropy": 5.130091571807862,
      "epoch": 2.7204610951008643,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004264270548226064,
      "loss": 4.8626,
      "mean_token_accuracy": 0.23187865018844606,
      "num_tokens": 64931532.0,
      "step": 28320
    },
    {
      "entropy": 5.183023309707641,
      "epoch": 2.7209414024975986,
      "grad_norm": 1.0,
      "learning_rate": 0.0004264016957757048,
      "loss": 4.8046,
      "mean_token_accuracy": 0.23655593246221543,
      "num_tokens": 64943214.0,
      "step": 28325
    },
    {
      "entropy": 5.172761011123657,
      "epoch": 2.7214217098943325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042637633321370545,
      "loss": 4.857,
      "mean_token_accuracy": 0.2302336573600769,
      "num_tokens": 64954726.0,
      "step": 28330
    },
    {
      "entropy": 5.177755737304688,
      "epoch": 2.7219020172910664,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004263509671371971,
      "loss": 4.85,
      "mean_token_accuracy": 0.22753977477550508,
      "num_tokens": 64967759.0,
      "step": 28335
    },
    {
      "entropy": 5.23459415435791,
      "epoch": 2.7223823246878003,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042632559754676865,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21986271291971207,
      "num_tokens": 64979431.0,
      "step": 28340
    },
    {
      "entropy": 5.112111377716064,
      "epoch": 2.722862632084534,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004263002244430092,
      "loss": 4.7708,
      "mean_token_accuracy": 0.24269004613161088,
      "num_tokens": 64992148.0,
      "step": 28345
    },
    {
      "entropy": 5.113949728012085,
      "epoch": 2.723342939481268,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004262748478265078,
      "loss": 4.7187,
      "mean_token_accuracy": 0.24013330340385436,
      "num_tokens": 65003750.0,
      "step": 28350
    },
    {
      "entropy": 5.070817661285401,
      "epoch": 2.723823246878002,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004262494676978537,
      "loss": 4.7611,
      "mean_token_accuracy": 0.23984224498271942,
      "num_tokens": 65013921.0,
      "step": 28355
    },
    {
      "entropy": 5.071116399765015,
      "epoch": 2.724303554274736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042622408405763607,
      "loss": 4.7557,
      "mean_token_accuracy": 0.23744438141584395,
      "num_tokens": 65025043.0,
      "step": 28360
    },
    {
      "entropy": 5.134008407592773,
      "epoch": 2.7247838616714697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004261986969064442,
      "loss": 4.7051,
      "mean_token_accuracy": 0.24125912487506868,
      "num_tokens": 65037155.0,
      "step": 28365
    },
    {
      "entropy": 5.123603487014771,
      "epoch": 2.7252641690682036,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042617330624486753,
      "loss": 4.7799,
      "mean_token_accuracy": 0.23062962144613267,
      "num_tokens": 65048876.0,
      "step": 28370
    },
    {
      "entropy": 5.202807140350342,
      "epoch": 2.7257444764649374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004261479120734956,
      "loss": 4.8316,
      "mean_token_accuracy": 0.2335604801774025,
      "num_tokens": 65060321.0,
      "step": 28375
    },
    {
      "entropy": 5.197618770599365,
      "epoch": 2.7262247838616713,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000426122514392918,
      "loss": 4.8428,
      "mean_token_accuracy": 0.23159638047218323,
      "num_tokens": 65073019.0,
      "step": 28380
    },
    {
      "entropy": 5.129364919662476,
      "epoch": 2.726705091258405,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042609711320372435,
      "loss": 4.7953,
      "mean_token_accuracy": 0.23424584567546844,
      "num_tokens": 65084597.0,
      "step": 28385
    },
    {
      "entropy": 5.142024803161621,
      "epoch": 2.7271853986551395,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004260717085065045,
      "loss": 4.7901,
      "mean_token_accuracy": 0.2326941415667534,
      "num_tokens": 65097137.0,
      "step": 28390
    },
    {
      "entropy": 5.18972806930542,
      "epoch": 2.727665706051873,
      "grad_norm": 1.0,
      "learning_rate": 0.00042604630030184797,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21986113488674164,
      "num_tokens": 65110229.0,
      "step": 28395
    },
    {
      "entropy": 5.2334047794342045,
      "epoch": 2.7281460134486073,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000426020888590345,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22631108313798903,
      "num_tokens": 65121767.0,
      "step": 28400
    },
    {
      "entropy": 5.104373264312744,
      "epoch": 2.728626320845341,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042599547337258536,
      "loss": 4.6955,
      "mean_token_accuracy": 0.24855609238147736,
      "num_tokens": 65132918.0,
      "step": 28405
    },
    {
      "entropy": 5.075769758224487,
      "epoch": 2.729106628242075,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042597005464915924,
      "loss": 4.7732,
      "mean_token_accuracy": 0.24030704647302628,
      "num_tokens": 65144384.0,
      "step": 28410
    },
    {
      "entropy": 5.232395935058594,
      "epoch": 2.729586935638809,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042594463242065674,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22168801426887513,
      "num_tokens": 65157495.0,
      "step": 28415
    },
    {
      "entropy": 5.142092275619507,
      "epoch": 2.730067243035543,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004259192066876681,
      "loss": 4.7149,
      "mean_token_accuracy": 0.24552863836288452,
      "num_tokens": 65168583.0,
      "step": 28420
    },
    {
      "entropy": 5.138854312896728,
      "epoch": 2.7305475504322767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042589377745078354,
      "loss": 4.8134,
      "mean_token_accuracy": 0.22812467962503433,
      "num_tokens": 65179909.0,
      "step": 28425
    },
    {
      "entropy": 5.15519437789917,
      "epoch": 2.7310278578290106,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042586834471059366,
      "loss": 4.8702,
      "mean_token_accuracy": 0.22832651883363725,
      "num_tokens": 65192239.0,
      "step": 28430
    },
    {
      "entropy": 5.097844743728638,
      "epoch": 2.7315081652257445,
      "grad_norm": 1.0,
      "learning_rate": 0.00042584290846768867,
      "loss": 4.763,
      "mean_token_accuracy": 0.23508933931589127,
      "num_tokens": 65203454.0,
      "step": 28435
    },
    {
      "entropy": 5.083636093139648,
      "epoch": 2.7319884726224783,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004258174687226593,
      "loss": 4.7317,
      "mean_token_accuracy": 0.2349179744720459,
      "num_tokens": 65216265.0,
      "step": 28440
    },
    {
      "entropy": 5.176862478256226,
      "epoch": 2.7324687800192122,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004257920254760962,
      "loss": 4.7627,
      "mean_token_accuracy": 0.23783342987298967,
      "num_tokens": 65227742.0,
      "step": 28445
    },
    {
      "entropy": 5.0982483386993405,
      "epoch": 2.732949087415946,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004257665787285899,
      "loss": 4.8265,
      "mean_token_accuracy": 0.24092403650283814,
      "num_tokens": 65239485.0,
      "step": 28450
    },
    {
      "entropy": 5.14745306968689,
      "epoch": 2.73342939481268,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042574112848073147,
      "loss": 4.7767,
      "mean_token_accuracy": 0.23167737275362016,
      "num_tokens": 65252818.0,
      "step": 28455
    },
    {
      "entropy": 5.146694707870483,
      "epoch": 2.733909702209414,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042571567473311157,
      "loss": 4.8234,
      "mean_token_accuracy": 0.23402685970067977,
      "num_tokens": 65264636.0,
      "step": 28460
    },
    {
      "entropy": 5.121426391601562,
      "epoch": 2.734390009606148,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004256902174863213,
      "loss": 4.7953,
      "mean_token_accuracy": 0.23269643485546113,
      "num_tokens": 65275426.0,
      "step": 28465
    },
    {
      "entropy": 5.152850914001465,
      "epoch": 2.7348703170028816,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042566475674095155,
      "loss": 4.826,
      "mean_token_accuracy": 0.23427633047103882,
      "num_tokens": 65286241.0,
      "step": 28470
    },
    {
      "entropy": 5.1214271068573,
      "epoch": 2.735350624399616,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004256392924975936,
      "loss": 4.7806,
      "mean_token_accuracy": 0.2370062693953514,
      "num_tokens": 65298307.0,
      "step": 28475
    },
    {
      "entropy": 5.214930725097656,
      "epoch": 2.7358309317963494,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042561382475683854,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22265502214431762,
      "num_tokens": 65310529.0,
      "step": 28480
    },
    {
      "entropy": 5.161530303955078,
      "epoch": 2.7363112391930837,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004255883535192777,
      "loss": 4.7845,
      "mean_token_accuracy": 0.23841722756624223,
      "num_tokens": 65321061.0,
      "step": 28485
    },
    {
      "entropy": 5.1181927680969235,
      "epoch": 2.7367915465898176,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004255628787855025,
      "loss": 4.7667,
      "mean_token_accuracy": 0.23622124046087264,
      "num_tokens": 65332068.0,
      "step": 28490
    },
    {
      "entropy": 5.257646226882935,
      "epoch": 2.7372718539865515,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004255374005561043,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2225349724292755,
      "num_tokens": 65344484.0,
      "step": 28495
    },
    {
      "entropy": 5.159114503860474,
      "epoch": 2.7377521613832854,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042551191883167464,
      "loss": 4.8268,
      "mean_token_accuracy": 0.23834110647439957,
      "num_tokens": 65355021.0,
      "step": 28500
    },
    {
      "entropy": 5.09635968208313,
      "epoch": 2.7382324687800192,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004254864336128052,
      "loss": 4.715,
      "mean_token_accuracy": 0.23463438898324968,
      "num_tokens": 65367024.0,
      "step": 28505
    },
    {
      "entropy": 5.117451572418213,
      "epoch": 2.738712776176753,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042546094490008765,
      "loss": 4.7523,
      "mean_token_accuracy": 0.23971293717622758,
      "num_tokens": 65377894.0,
      "step": 28510
    },
    {
      "entropy": 5.182496738433838,
      "epoch": 2.739193083573487,
      "grad_norm": 1.0,
      "learning_rate": 0.0004254354526941136,
      "loss": 4.7836,
      "mean_token_accuracy": 0.23180389553308486,
      "num_tokens": 65389627.0,
      "step": 28515
    },
    {
      "entropy": 5.128242111206054,
      "epoch": 2.739673390970221,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004254099569954751,
      "loss": 4.7068,
      "mean_token_accuracy": 0.24367718547582626,
      "num_tokens": 65399715.0,
      "step": 28520
    },
    {
      "entropy": 5.095369052886963,
      "epoch": 2.7401536983669548,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004253844578047641,
      "loss": 4.8298,
      "mean_token_accuracy": 0.23552560657262803,
      "num_tokens": 65410459.0,
      "step": 28525
    },
    {
      "entropy": 5.077534341812134,
      "epoch": 2.7406340057636887,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004253589551225725,
      "loss": 4.7419,
      "mean_token_accuracy": 0.23885937929153442,
      "num_tokens": 65422412.0,
      "step": 28530
    },
    {
      "entropy": 5.129709768295288,
      "epoch": 2.7411143131604225,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042533344894949245,
      "loss": 4.8142,
      "mean_token_accuracy": 0.2301225930452347,
      "num_tokens": 65433982.0,
      "step": 28535
    },
    {
      "entropy": 5.122131586074829,
      "epoch": 2.7415946205571564,
      "grad_norm": 1.0,
      "learning_rate": 0.00042530793928611605,
      "loss": 4.766,
      "mean_token_accuracy": 0.23602638095617295,
      "num_tokens": 65445828.0,
      "step": 28540
    },
    {
      "entropy": 5.155824184417725,
      "epoch": 2.7420749279538903,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004252824261330357,
      "loss": 4.7786,
      "mean_token_accuracy": 0.234107705950737,
      "num_tokens": 65457271.0,
      "step": 28545
    },
    {
      "entropy": 5.209269952774048,
      "epoch": 2.7425552353506246,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042525690949084364,
      "loss": 4.877,
      "mean_token_accuracy": 0.23849795162677764,
      "num_tokens": 65468414.0,
      "step": 28550
    },
    {
      "entropy": 5.137510299682617,
      "epoch": 2.743035542747358,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042523138936013233,
      "loss": 4.7388,
      "mean_token_accuracy": 0.23561877608299256,
      "num_tokens": 65479454.0,
      "step": 28555
    },
    {
      "entropy": 5.125989484786987,
      "epoch": 2.7435158501440924,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042520586574149423,
      "loss": 4.8825,
      "mean_token_accuracy": 0.2337900161743164,
      "num_tokens": 65490028.0,
      "step": 28560
    },
    {
      "entropy": 5.084506607055664,
      "epoch": 2.7439961575408263,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042518033863552185,
      "loss": 4.7481,
      "mean_token_accuracy": 0.24215775430202485,
      "num_tokens": 65501929.0,
      "step": 28565
    },
    {
      "entropy": 5.16605772972107,
      "epoch": 2.74447646493756,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004251548080428081,
      "loss": 4.7394,
      "mean_token_accuracy": 0.24046671390533447,
      "num_tokens": 65513654.0,
      "step": 28570
    },
    {
      "entropy": 5.065210580825806,
      "epoch": 2.744956772334294,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004251292739639455,
      "loss": 4.7326,
      "mean_token_accuracy": 0.24638991206884384,
      "num_tokens": 65525219.0,
      "step": 28575
    },
    {
      "entropy": 5.066080856323242,
      "epoch": 2.745437079731028,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042510373639952694,
      "loss": 4.7215,
      "mean_token_accuracy": 0.23759246319532396,
      "num_tokens": 65536255.0,
      "step": 28580
    },
    {
      "entropy": 5.215390634536743,
      "epoch": 2.745917387127762,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042507819535014547,
      "loss": 4.8867,
      "mean_token_accuracy": 0.22847847938537597,
      "num_tokens": 65547489.0,
      "step": 28585
    },
    {
      "entropy": 5.129744291305542,
      "epoch": 2.7463976945244957,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042505265081639376,
      "loss": 4.772,
      "mean_token_accuracy": 0.23865943551063537,
      "num_tokens": 65559892.0,
      "step": 28590
    },
    {
      "entropy": 5.263288354873657,
      "epoch": 2.7468780019212296,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004250271027988652,
      "loss": 4.9128,
      "mean_token_accuracy": 0.22625427097082138,
      "num_tokens": 65570742.0,
      "step": 28595
    },
    {
      "entropy": 5.181989049911499,
      "epoch": 2.7473583093179634,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042500155129815274,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23144059628248215,
      "num_tokens": 65584202.0,
      "step": 28600
    },
    {
      "entropy": 5.084861183166504,
      "epoch": 2.7478386167146973,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042497599631484965,
      "loss": 4.7299,
      "mean_token_accuracy": 0.23914059102535248,
      "num_tokens": 65595744.0,
      "step": 28605
    },
    {
      "entropy": 5.070834112167359,
      "epoch": 2.748318924111431,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042495043784954926,
      "loss": 4.7432,
      "mean_token_accuracy": 0.23738683462142945,
      "num_tokens": 65607914.0,
      "step": 28610
    },
    {
      "entropy": 5.167619752883911,
      "epoch": 2.748799231508165,
      "grad_norm": 0.9375,
      "learning_rate": 0.000424924875902845,
      "loss": 4.7614,
      "mean_token_accuracy": 0.2372438997030258,
      "num_tokens": 65620134.0,
      "step": 28615
    },
    {
      "entropy": 5.142332315444946,
      "epoch": 2.749279538904899,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004248993104753303,
      "loss": 4.7497,
      "mean_token_accuracy": 0.2400053009390831,
      "num_tokens": 65631804.0,
      "step": 28620
    },
    {
      "entropy": 5.164214754104615,
      "epoch": 2.7497598463016333,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004248737415675987,
      "loss": 4.8549,
      "mean_token_accuracy": 0.23483745753765106,
      "num_tokens": 65644067.0,
      "step": 28625
    },
    {
      "entropy": 5.152899169921875,
      "epoch": 2.7502401536983667,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004248481691802439,
      "loss": 4.8392,
      "mean_token_accuracy": 0.2348172828555107,
      "num_tokens": 65655036.0,
      "step": 28630
    },
    {
      "entropy": 5.1658727645874025,
      "epoch": 2.750720461095101,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004248225933138595,
      "loss": 4.7736,
      "mean_token_accuracy": 0.23387196958065032,
      "num_tokens": 65666827.0,
      "step": 28635
    },
    {
      "entropy": 5.181745815277099,
      "epoch": 2.751200768491835,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042479701396903945,
      "loss": 4.8705,
      "mean_token_accuracy": 0.2287372589111328,
      "num_tokens": 65677635.0,
      "step": 28640
    },
    {
      "entropy": 5.089689016342163,
      "epoch": 2.751681075888569,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004247714311463775,
      "loss": 4.7945,
      "mean_token_accuracy": 0.23245285004377364,
      "num_tokens": 65689695.0,
      "step": 28645
    },
    {
      "entropy": 5.129916143417359,
      "epoch": 2.7521613832853027,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042474584484646766,
      "loss": 4.7259,
      "mean_token_accuracy": 0.2369358316063881,
      "num_tokens": 65700969.0,
      "step": 28650
    },
    {
      "entropy": 5.199172592163086,
      "epoch": 2.7526416906820366,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004247202550699039,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23069519698619842,
      "num_tokens": 65712743.0,
      "step": 28655
    },
    {
      "entropy": 4.996461963653564,
      "epoch": 2.7531219980787704,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004246946618172805,
      "loss": 4.6298,
      "mean_token_accuracy": 0.2481934040784836,
      "num_tokens": 65724912.0,
      "step": 28660
    },
    {
      "entropy": 5.115214490890503,
      "epoch": 2.7536023054755043,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004246690650891915,
      "loss": 4.7483,
      "mean_token_accuracy": 0.23474295288324357,
      "num_tokens": 65736295.0,
      "step": 28665
    },
    {
      "entropy": 5.1805602550506595,
      "epoch": 2.754082612872238,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004246434648862312,
      "loss": 4.7658,
      "mean_token_accuracy": 0.24073395878076553,
      "num_tokens": 65747032.0,
      "step": 28670
    },
    {
      "entropy": 5.149608421325683,
      "epoch": 2.754562920268972,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004246178612089941,
      "loss": 4.8594,
      "mean_token_accuracy": 0.2332184538245201,
      "num_tokens": 65759001.0,
      "step": 28675
    },
    {
      "entropy": 5.115271043777466,
      "epoch": 2.755043227665706,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004245922540580744,
      "loss": 4.7805,
      "mean_token_accuracy": 0.23087127208709718,
      "num_tokens": 65769781.0,
      "step": 28680
    },
    {
      "entropy": 5.136278676986694,
      "epoch": 2.75552353506244,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004245666434340668,
      "loss": 4.8024,
      "mean_token_accuracy": 0.23268161416053773,
      "num_tokens": 65781281.0,
      "step": 28685
    },
    {
      "entropy": 5.201958465576172,
      "epoch": 2.7560038424591737,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004245410293375659,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22200808078050613,
      "num_tokens": 65792884.0,
      "step": 28690
    },
    {
      "entropy": 5.130381917953491,
      "epoch": 2.7564841498559076,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004245154117691664,
      "loss": 4.7491,
      "mean_token_accuracy": 0.24411555528640747,
      "num_tokens": 65804616.0,
      "step": 28695
    },
    {
      "entropy": 5.09432373046875,
      "epoch": 2.756964457252642,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004244897907294628,
      "loss": 4.7397,
      "mean_token_accuracy": 0.2448887661099434,
      "num_tokens": 65814859.0,
      "step": 28700
    },
    {
      "entropy": 5.119327783584595,
      "epoch": 2.7574447646493754,
      "grad_norm": 1.0,
      "learning_rate": 0.0004244641662190504,
      "loss": 4.7523,
      "mean_token_accuracy": 0.23810336738824844,
      "num_tokens": 65824416.0,
      "step": 28705
    },
    {
      "entropy": 5.161362266540527,
      "epoch": 2.7579250720461097,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042443853823852376,
      "loss": 4.836,
      "mean_token_accuracy": 0.23275505602359772,
      "num_tokens": 65835428.0,
      "step": 28710
    },
    {
      "entropy": 5.147485446929932,
      "epoch": 2.7584053794428436,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000424412906788478,
      "loss": 4.8135,
      "mean_token_accuracy": 0.22862967997789382,
      "num_tokens": 65846683.0,
      "step": 28715
    },
    {
      "entropy": 5.1948991298675535,
      "epoch": 2.7588856868395775,
      "grad_norm": 1.0,
      "learning_rate": 0.0004243872718695082,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23046147972345352,
      "num_tokens": 65857850.0,
      "step": 28720
    },
    {
      "entropy": 5.056407356262207,
      "epoch": 2.7593659942363113,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042436163348220956,
      "loss": 4.6934,
      "mean_token_accuracy": 0.25126656591892244,
      "num_tokens": 65869688.0,
      "step": 28725
    },
    {
      "entropy": 5.1470818519592285,
      "epoch": 2.7598463016330452,
      "grad_norm": 1.0,
      "learning_rate": 0.0004243359916271773,
      "loss": 4.7628,
      "mean_token_accuracy": 0.24040691405534745,
      "num_tokens": 65880705.0,
      "step": 28730
    },
    {
      "entropy": 5.142822408676148,
      "epoch": 2.760326609029779,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004243103463050067,
      "loss": 4.7419,
      "mean_token_accuracy": 0.23737859725952148,
      "num_tokens": 65894070.0,
      "step": 28735
    },
    {
      "entropy": 5.155004596710205,
      "epoch": 2.760806916426513,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004242846975162933,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22752270698547364,
      "num_tokens": 65905899.0,
      "step": 28740
    },
    {
      "entropy": 5.08448896408081,
      "epoch": 2.761287223823247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042425904526163246,
      "loss": 4.7975,
      "mean_token_accuracy": 0.23270562440156936,
      "num_tokens": 65917173.0,
      "step": 28745
    },
    {
      "entropy": 5.189679718017578,
      "epoch": 2.7617675312199808,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004242333895416198,
      "loss": 4.876,
      "mean_token_accuracy": 0.2256169557571411,
      "num_tokens": 65930125.0,
      "step": 28750
    },
    {
      "entropy": 5.1556010246276855,
      "epoch": 2.7622478386167146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000424207730356851,
      "loss": 4.7822,
      "mean_token_accuracy": 0.22892038077116011,
      "num_tokens": 65941569.0,
      "step": 28755
    },
    {
      "entropy": 5.112465143203735,
      "epoch": 2.7627281460134485,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004241820677079218,
      "loss": 4.7504,
      "mean_token_accuracy": 0.23392239212989807,
      "num_tokens": 65953970.0,
      "step": 28760
    },
    {
      "entropy": 5.086101627349853,
      "epoch": 2.7632084534101824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042415640159542783,
      "loss": 4.7614,
      "mean_token_accuracy": 0.23781574070453643,
      "num_tokens": 65965740.0,
      "step": 28765
    },
    {
      "entropy": 5.149328327178955,
      "epoch": 2.7636887608069163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004241307320199652,
      "loss": 4.775,
      "mean_token_accuracy": 0.23488989621400833,
      "num_tokens": 65975986.0,
      "step": 28770
    },
    {
      "entropy": 5.166319274902344,
      "epoch": 2.7641690682036506,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004241050589821298,
      "loss": 4.8738,
      "mean_token_accuracy": 0.22968099862337113,
      "num_tokens": 65988356.0,
      "step": 28775
    },
    {
      "entropy": 5.143064546585083,
      "epoch": 2.764649375600384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004240793824825177,
      "loss": 4.7532,
      "mean_token_accuracy": 0.23732197880744935,
      "num_tokens": 65998526.0,
      "step": 28780
    },
    {
      "entropy": 5.093892526626587,
      "epoch": 2.7651296829971184,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042405370252172496,
      "loss": 4.683,
      "mean_token_accuracy": 0.24713898301124573,
      "num_tokens": 66009605.0,
      "step": 28785
    },
    {
      "entropy": 5.081457901000976,
      "epoch": 2.765609990393852,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004240280191003479,
      "loss": 4.694,
      "mean_token_accuracy": 0.24243203699588775,
      "num_tokens": 66020794.0,
      "step": 28790
    },
    {
      "entropy": 5.130601692199707,
      "epoch": 2.766090297790586,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004240023322189828,
      "loss": 4.8287,
      "mean_token_accuracy": 0.2382814809679985,
      "num_tokens": 66032174.0,
      "step": 28795
    },
    {
      "entropy": 5.161336755752563,
      "epoch": 2.76657060518732,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004239766418782258,
      "loss": 4.7631,
      "mean_token_accuracy": 0.23821392953395842,
      "num_tokens": 66042552.0,
      "step": 28800
    },
    {
      "entropy": 5.075876760482788,
      "epoch": 2.767050912584054,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004239509480786737,
      "loss": 4.7559,
      "mean_token_accuracy": 0.23907833695411682,
      "num_tokens": 66053935.0,
      "step": 28805
    },
    {
      "entropy": 5.036463975906372,
      "epoch": 2.7675312199807878,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042392525082092286,
      "loss": 4.6135,
      "mean_token_accuracy": 0.2439704567193985,
      "num_tokens": 66064957.0,
      "step": 28810
    },
    {
      "entropy": 5.133957958221435,
      "epoch": 2.7680115273775217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004238995501055699,
      "loss": 4.8369,
      "mean_token_accuracy": 0.23587024062871934,
      "num_tokens": 66076270.0,
      "step": 28815
    },
    {
      "entropy": 5.229872512817383,
      "epoch": 2.7684918347742555,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004238738459332115,
      "loss": 4.8653,
      "mean_token_accuracy": 0.2299604520201683,
      "num_tokens": 66088458.0,
      "step": 28820
    },
    {
      "entropy": 5.156051158905029,
      "epoch": 2.7689721421709894,
      "grad_norm": 1.0,
      "learning_rate": 0.0004238481383044445,
      "loss": 4.7462,
      "mean_token_accuracy": 0.23166140764951706,
      "num_tokens": 66100114.0,
      "step": 28825
    },
    {
      "entropy": 5.13808240890503,
      "epoch": 2.7694524495677233,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042382242721986573,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23299262374639512,
      "num_tokens": 66112448.0,
      "step": 28830
    },
    {
      "entropy": 5.136671543121338,
      "epoch": 2.769932756964457,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042379671268007207,
      "loss": 4.7986,
      "mean_token_accuracy": 0.2405425176024437,
      "num_tokens": 66124122.0,
      "step": 28835
    },
    {
      "entropy": 5.101956033706665,
      "epoch": 2.770413064361191,
      "grad_norm": 1.0,
      "learning_rate": 0.0004237709946856607,
      "loss": 4.7583,
      "mean_token_accuracy": 0.23196348547935486,
      "num_tokens": 66135375.0,
      "step": 28840
    },
    {
      "entropy": 5.1831090450286865,
      "epoch": 2.770893371757925,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042374527323722836,
      "loss": 4.8586,
      "mean_token_accuracy": 0.2250829592347145,
      "num_tokens": 66148850.0,
      "step": 28845
    },
    {
      "entropy": 5.067659950256347,
      "epoch": 2.771373679154659,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042371954833537263,
      "loss": 4.73,
      "mean_token_accuracy": 0.24156603068113328,
      "num_tokens": 66160040.0,
      "step": 28850
    },
    {
      "entropy": 5.146295785903931,
      "epoch": 2.7718539865513927,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042369381998069055,
      "loss": 4.7605,
      "mean_token_accuracy": 0.2368394613265991,
      "num_tokens": 66172857.0,
      "step": 28855
    },
    {
      "entropy": 5.075040102005005,
      "epoch": 2.772334293948127,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004236680881737795,
      "loss": 4.7448,
      "mean_token_accuracy": 0.23833366185426713,
      "num_tokens": 66184546.0,
      "step": 28860
    },
    {
      "entropy": 5.126919460296631,
      "epoch": 2.7728146013448605,
      "grad_norm": 1.0,
      "learning_rate": 0.0004236423529152369,
      "loss": 4.7626,
      "mean_token_accuracy": 0.23462526649236679,
      "num_tokens": 66195270.0,
      "step": 28865
    },
    {
      "entropy": 5.118718957901001,
      "epoch": 2.773294908741595,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004236166142056602,
      "loss": 4.7844,
      "mean_token_accuracy": 0.2338502898812294,
      "num_tokens": 66206998.0,
      "step": 28870
    },
    {
      "entropy": 5.203345727920532,
      "epoch": 2.7737752161383287,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004235908720456471,
      "loss": 4.8293,
      "mean_token_accuracy": 0.23124637305736542,
      "num_tokens": 66217645.0,
      "step": 28875
    },
    {
      "entropy": 5.084938287734985,
      "epoch": 2.7742555235350626,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004235651264357951,
      "loss": 4.7498,
      "mean_token_accuracy": 0.2444412738084793,
      "num_tokens": 66228740.0,
      "step": 28880
    },
    {
      "entropy": 5.00184006690979,
      "epoch": 2.7747358309317964,
      "grad_norm": 1.0,
      "learning_rate": 0.00042353937737670206,
      "loss": 4.643,
      "mean_token_accuracy": 0.2453790232539177,
      "num_tokens": 66239459.0,
      "step": 28885
    },
    {
      "entropy": 5.15271635055542,
      "epoch": 2.7752161383285303,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004235136248689658,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23579445332288743,
      "num_tokens": 66251470.0,
      "step": 28890
    },
    {
      "entropy": 5.183565139770508,
      "epoch": 2.775696445725264,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004234878689131841,
      "loss": 4.799,
      "mean_token_accuracy": 0.23223887085914613,
      "num_tokens": 66262045.0,
      "step": 28895
    },
    {
      "entropy": 5.0519270420074465,
      "epoch": 2.776176753121998,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000423462109509955,
      "loss": 4.7242,
      "mean_token_accuracy": 0.24032147228717804,
      "num_tokens": 66272077.0,
      "step": 28900
    },
    {
      "entropy": 5.08918023109436,
      "epoch": 2.776657060518732,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004234363466598765,
      "loss": 4.7435,
      "mean_token_accuracy": 0.24188703298568726,
      "num_tokens": 66283091.0,
      "step": 28905
    },
    {
      "entropy": 5.151421642303466,
      "epoch": 2.777137367915466,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042341058036354687,
      "loss": 4.8309,
      "mean_token_accuracy": 0.23639825582504273,
      "num_tokens": 66293770.0,
      "step": 28910
    },
    {
      "entropy": 5.104672050476074,
      "epoch": 2.7776176753121997,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042338481062156424,
      "loss": 4.7603,
      "mean_token_accuracy": 0.24107868671417237,
      "num_tokens": 66305288.0,
      "step": 28915
    },
    {
      "entropy": 5.182474136352539,
      "epoch": 2.7780979827089336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042335903743452694,
      "loss": 4.8395,
      "mean_token_accuracy": 0.23299387246370315,
      "num_tokens": 66316414.0,
      "step": 28920
    },
    {
      "entropy": 5.233222961425781,
      "epoch": 2.7785782901056675,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004233332608030333,
      "loss": 4.8661,
      "mean_token_accuracy": 0.227619905769825,
      "num_tokens": 66327573.0,
      "step": 28925
    },
    {
      "entropy": 5.14265456199646,
      "epoch": 2.7790585975024014,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042330748072768183,
      "loss": 4.8309,
      "mean_token_accuracy": 0.2361629545688629,
      "num_tokens": 66338950.0,
      "step": 28930
    },
    {
      "entropy": 5.048491525650024,
      "epoch": 2.7795389048991357,
      "grad_norm": 1.0625,
      "learning_rate": 0.000423281697209071,
      "loss": 4.7025,
      "mean_token_accuracy": 0.24305184185504913,
      "num_tokens": 66348814.0,
      "step": 28935
    },
    {
      "entropy": 5.1058351516723635,
      "epoch": 2.780019212295869,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004232559102477995,
      "loss": 4.7049,
      "mean_token_accuracy": 0.24277715831995011,
      "num_tokens": 66360119.0,
      "step": 28940
    },
    {
      "entropy": 5.1534992218017575,
      "epoch": 2.7804995196926034,
      "grad_norm": 1.0,
      "learning_rate": 0.000423230119844466,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23287494033575057,
      "num_tokens": 66370843.0,
      "step": 28945
    },
    {
      "entropy": 5.157514905929565,
      "epoch": 2.7809798270893373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004232043259996692,
      "loss": 4.7818,
      "mean_token_accuracy": 0.232870352268219,
      "num_tokens": 66382420.0,
      "step": 28950
    },
    {
      "entropy": 5.099154090881347,
      "epoch": 2.781460134486071,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004231785287140081,
      "loss": 4.7591,
      "mean_token_accuracy": 0.23883823603391646,
      "num_tokens": 66394327.0,
      "step": 28955
    },
    {
      "entropy": 5.111931800842285,
      "epoch": 2.781940441882805,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004231527279880816,
      "loss": 4.8525,
      "mean_token_accuracy": 0.22483856976032257,
      "num_tokens": 66407927.0,
      "step": 28960
    },
    {
      "entropy": 5.114114809036255,
      "epoch": 2.782420749279539,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004231269238224885,
      "loss": 4.6718,
      "mean_token_accuracy": 0.24329811334609985,
      "num_tokens": 66419248.0,
      "step": 28965
    },
    {
      "entropy": 5.1095788955688475,
      "epoch": 2.782901056676273,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004231011162178282,
      "loss": 4.708,
      "mean_token_accuracy": 0.23800562620162963,
      "num_tokens": 66430750.0,
      "step": 28970
    },
    {
      "entropy": 5.09884090423584,
      "epoch": 2.7833813640730067,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004230753051746998,
      "loss": 4.8017,
      "mean_token_accuracy": 0.23570553511381148,
      "num_tokens": 66443792.0,
      "step": 28975
    },
    {
      "entropy": 5.11474723815918,
      "epoch": 2.7838616714697406,
      "grad_norm": 0.90625,
      "learning_rate": 0.00042304949069370246,
      "loss": 4.6986,
      "mean_token_accuracy": 0.2421714261174202,
      "num_tokens": 66455091.0,
      "step": 28980
    },
    {
      "entropy": 5.156706094741821,
      "epoch": 2.7843419788664745,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042302367277543553,
      "loss": 4.8167,
      "mean_token_accuracy": 0.2359660431742668,
      "num_tokens": 66466269.0,
      "step": 28985
    },
    {
      "entropy": 5.124939203262329,
      "epoch": 2.7848222862632084,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042299785142049855,
      "loss": 4.7592,
      "mean_token_accuracy": 0.2382224515080452,
      "num_tokens": 66476032.0,
      "step": 28990
    },
    {
      "entropy": 5.083884572982788,
      "epoch": 2.7853025936599423,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004229720266294908,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23624165952205659,
      "num_tokens": 66487265.0,
      "step": 28995
    },
    {
      "entropy": 5.230425930023193,
      "epoch": 2.785782901056676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000422946198403012,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2302054464817047,
      "num_tokens": 66498762.0,
      "step": 29000
    },
    {
      "entropy": 5.229783582687378,
      "epoch": 2.78626320845341,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004229203667416619,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22490211427211762,
      "num_tokens": 66509455.0,
      "step": 29005
    },
    {
      "entropy": 5.238659763336182,
      "epoch": 2.7867435158501443,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042289453164604,
      "loss": 4.9383,
      "mean_token_accuracy": 0.22423601895570755,
      "num_tokens": 66520858.0,
      "step": 29010
    },
    {
      "entropy": 5.13766303062439,
      "epoch": 2.787223823246878,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004228686931167463,
      "loss": 4.7915,
      "mean_token_accuracy": 0.23527013808488845,
      "num_tokens": 66532680.0,
      "step": 29015
    },
    {
      "entropy": 5.14015302658081,
      "epoch": 2.787704130643612,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004228428511543806,
      "loss": 4.8087,
      "mean_token_accuracy": 0.23736958503723143,
      "num_tokens": 66543554.0,
      "step": 29020
    },
    {
      "entropy": 5.08807635307312,
      "epoch": 2.7881844380403455,
      "grad_norm": 1.0,
      "learning_rate": 0.00042281700575954283,
      "loss": 4.705,
      "mean_token_accuracy": 0.23993852734565735,
      "num_tokens": 66554041.0,
      "step": 29025
    },
    {
      "entropy": 5.096477365493774,
      "epoch": 2.78866474543708,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004227911569328332,
      "loss": 4.7485,
      "mean_token_accuracy": 0.2382219284772873,
      "num_tokens": 66566216.0,
      "step": 29030
    },
    {
      "entropy": 5.119823980331421,
      "epoch": 2.7891450528338138,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004227653046748517,
      "loss": 4.7955,
      "mean_token_accuracy": 0.23521044999361038,
      "num_tokens": 66577964.0,
      "step": 29035
    },
    {
      "entropy": 5.140038251876831,
      "epoch": 2.7896253602305476,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042273944898619864,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23976943790912628,
      "num_tokens": 66590172.0,
      "step": 29040
    },
    {
      "entropy": 5.071757888793945,
      "epoch": 2.7901056676272815,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042271358986747427,
      "loss": 4.7273,
      "mean_token_accuracy": 0.24189767837524415,
      "num_tokens": 66601277.0,
      "step": 29045
    },
    {
      "entropy": 5.081712913513184,
      "epoch": 2.7905859750240154,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042268772731927895,
      "loss": 4.7154,
      "mean_token_accuracy": 0.2429250955581665,
      "num_tokens": 66612080.0,
      "step": 29050
    },
    {
      "entropy": 5.099741697311401,
      "epoch": 2.7910662824207493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042266186134221317,
      "loss": 4.7807,
      "mean_token_accuracy": 0.23548422753810883,
      "num_tokens": 66623080.0,
      "step": 29055
    },
    {
      "entropy": 5.120627689361572,
      "epoch": 2.791546589817483,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004226359919368774,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23628033250570296,
      "num_tokens": 66635486.0,
      "step": 29060
    },
    {
      "entropy": 5.1084432125091555,
      "epoch": 2.792026897214217,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042261011910387224,
      "loss": 4.7509,
      "mean_token_accuracy": 0.24187008589506148,
      "num_tokens": 66647137.0,
      "step": 29065
    },
    {
      "entropy": 5.21622862815857,
      "epoch": 2.792507204610951,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004225842428437985,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22482393085956573,
      "num_tokens": 66658890.0,
      "step": 29070
    },
    {
      "entropy": 5.066889810562134,
      "epoch": 2.792987512007685,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042255836315725694,
      "loss": 4.7291,
      "mean_token_accuracy": 0.24433477371931075,
      "num_tokens": 66670706.0,
      "step": 29075
    },
    {
      "entropy": 5.143387413024902,
      "epoch": 2.7934678194044187,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004225324800448483,
      "loss": 4.8003,
      "mean_token_accuracy": 0.23103850185871125,
      "num_tokens": 66681930.0,
      "step": 29080
    },
    {
      "entropy": 5.114485216140747,
      "epoch": 2.793948126801153,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042250659350717343,
      "loss": 4.7265,
      "mean_token_accuracy": 0.24363391250371932,
      "num_tokens": 66694224.0,
      "step": 29085
    },
    {
      "entropy": 5.194471025466919,
      "epoch": 2.7944284341978864,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042248070354483354,
      "loss": 4.8138,
      "mean_token_accuracy": 0.2380561575293541,
      "num_tokens": 66705905.0,
      "step": 29090
    },
    {
      "entropy": 5.127234554290771,
      "epoch": 2.7949087415946208,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004224548101584297,
      "loss": 4.8014,
      "mean_token_accuracy": 0.23707718700170516,
      "num_tokens": 66717189.0,
      "step": 29095
    },
    {
      "entropy": 5.10033802986145,
      "epoch": 2.795389048991354,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000422428913348563,
      "loss": 4.7122,
      "mean_token_accuracy": 0.23692432492971421,
      "num_tokens": 66728066.0,
      "step": 29100
    },
    {
      "entropy": 5.131421422958374,
      "epoch": 2.7958693563880885,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004224030131158346,
      "loss": 4.8087,
      "mean_token_accuracy": 0.24445258229970931,
      "num_tokens": 66738530.0,
      "step": 29105
    },
    {
      "entropy": 5.142169380187989,
      "epoch": 2.7963496637848224,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004223771094608461,
      "loss": 4.8478,
      "mean_token_accuracy": 0.23236024379730225,
      "num_tokens": 66749563.0,
      "step": 29110
    },
    {
      "entropy": 5.153646612167359,
      "epoch": 2.7968299711815563,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004223512023841986,
      "loss": 4.8425,
      "mean_token_accuracy": 0.22953287661075591,
      "num_tokens": 66761376.0,
      "step": 29115
    },
    {
      "entropy": 5.143674516677857,
      "epoch": 2.79731027857829,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00042232529188649374,
      "loss": 4.7016,
      "mean_token_accuracy": 0.2417916998267174,
      "num_tokens": 66774193.0,
      "step": 29120
    },
    {
      "entropy": 5.121995687484741,
      "epoch": 2.797790585975024,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004222993779683331,
      "loss": 4.7617,
      "mean_token_accuracy": 0.23543039858341216,
      "num_tokens": 66787007.0,
      "step": 29125
    },
    {
      "entropy": 5.082212162017822,
      "epoch": 2.798270893371758,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042227346063031837,
      "loss": 4.7039,
      "mean_token_accuracy": 0.2425306335091591,
      "num_tokens": 66798426.0,
      "step": 29130
    },
    {
      "entropy": 5.128195095062256,
      "epoch": 2.798751200768492,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004222475398730511,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23607843071222306,
      "num_tokens": 66810843.0,
      "step": 29135
    },
    {
      "entropy": 5.061735153198242,
      "epoch": 2.7992315081652257,
      "grad_norm": 1.0,
      "learning_rate": 0.0004222216156971332,
      "loss": 4.718,
      "mean_token_accuracy": 0.2427769497036934,
      "num_tokens": 66823350.0,
      "step": 29140
    },
    {
      "entropy": 5.109006071090699,
      "epoch": 2.7997118155619596,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042219568810316656,
      "loss": 4.8246,
      "mean_token_accuracy": 0.24057965278625487,
      "num_tokens": 66836830.0,
      "step": 29145
    },
    {
      "entropy": 5.222476434707642,
      "epoch": 2.8001921229586935,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004221697570917531,
      "loss": 4.881,
      "mean_token_accuracy": 0.2228244885802269,
      "num_tokens": 66848843.0,
      "step": 29150
    },
    {
      "entropy": 5.166522884368897,
      "epoch": 2.8006724303554273,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000422143822663495,
      "loss": 4.804,
      "mean_token_accuracy": 0.23656369298696517,
      "num_tokens": 66858598.0,
      "step": 29155
    },
    {
      "entropy": 5.040959739685059,
      "epoch": 2.8011527377521612,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004221178848189941,
      "loss": 4.7349,
      "mean_token_accuracy": 0.23993058055639266,
      "num_tokens": 66869633.0,
      "step": 29160
    },
    {
      "entropy": 5.104132032394409,
      "epoch": 2.801633045148895,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042209194355885283,
      "loss": 4.7932,
      "mean_token_accuracy": 0.2401985704898834,
      "num_tokens": 66880501.0,
      "step": 29165
    },
    {
      "entropy": 5.197134780883789,
      "epoch": 2.8021133525456294,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004220659988836734,
      "loss": 4.8473,
      "mean_token_accuracy": 0.22625454217195512,
      "num_tokens": 66892488.0,
      "step": 29170
    },
    {
      "entropy": 5.0835198879241945,
      "epoch": 2.802593659942363,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004220400507940582,
      "loss": 4.6433,
      "mean_token_accuracy": 0.25394022315740583,
      "num_tokens": 66903861.0,
      "step": 29175
    },
    {
      "entropy": 5.126417636871338,
      "epoch": 2.803073967339097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042201409929060955,
      "loss": 4.7525,
      "mean_token_accuracy": 0.23355970531702042,
      "num_tokens": 66915526.0,
      "step": 29180
    },
    {
      "entropy": 5.069150495529175,
      "epoch": 2.803554274735831,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004219881443739301,
      "loss": 4.7304,
      "mean_token_accuracy": 0.24006086438894272,
      "num_tokens": 66926798.0,
      "step": 29185
    },
    {
      "entropy": 5.209089469909668,
      "epoch": 2.804034582132565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004219621860446225,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2241172432899475,
      "num_tokens": 66938185.0,
      "step": 29190
    },
    {
      "entropy": 5.18147644996643,
      "epoch": 2.804514889529299,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004219362243032892,
      "loss": 4.8044,
      "mean_token_accuracy": 0.23820945620536804,
      "num_tokens": 66950423.0,
      "step": 29195
    },
    {
      "entropy": 5.129042148590088,
      "epoch": 2.8049951969260327,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042191025915053323,
      "loss": 4.7467,
      "mean_token_accuracy": 0.23096424639225005,
      "num_tokens": 66961545.0,
      "step": 29200
    },
    {
      "entropy": 5.178604793548584,
      "epoch": 2.8054755043227666,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042188429058695714,
      "loss": 4.7927,
      "mean_token_accuracy": 0.2362284004688263,
      "num_tokens": 66971932.0,
      "step": 29205
    },
    {
      "entropy": 5.067948961257935,
      "epoch": 2.8059558117195005,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042185831861316406,
      "loss": 4.7483,
      "mean_token_accuracy": 0.2397117033600807,
      "num_tokens": 66983830.0,
      "step": 29210
    },
    {
      "entropy": 5.113395738601684,
      "epoch": 2.8064361191162344,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004218323432297568,
      "loss": 4.7914,
      "mean_token_accuracy": 0.2367238864302635,
      "num_tokens": 66995587.0,
      "step": 29215
    },
    {
      "entropy": 5.207929039001465,
      "epoch": 2.8069164265129682,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042180636443733864,
      "loss": 4.8213,
      "mean_token_accuracy": 0.2339792713522911,
      "num_tokens": 67007947.0,
      "step": 29220
    },
    {
      "entropy": 5.061005735397339,
      "epoch": 2.807396733909702,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042178038223651253,
      "loss": 4.739,
      "mean_token_accuracy": 0.23675398528575897,
      "num_tokens": 67019708.0,
      "step": 29225
    },
    {
      "entropy": 5.095407152175904,
      "epoch": 2.807877041306436,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042175439662788195,
      "loss": 4.7212,
      "mean_token_accuracy": 0.24195690006017684,
      "num_tokens": 67030372.0,
      "step": 29230
    },
    {
      "entropy": 5.249728107452393,
      "epoch": 2.80835734870317,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042172840761204986,
      "loss": 4.9132,
      "mean_token_accuracy": 0.22821317315101625,
      "num_tokens": 67040614.0,
      "step": 29235
    },
    {
      "entropy": 5.1661293506622314,
      "epoch": 2.8088376560999038,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004217024151896199,
      "loss": 4.8273,
      "mean_token_accuracy": 0.23810011893510818,
      "num_tokens": 67053409.0,
      "step": 29240
    },
    {
      "entropy": 5.185674619674683,
      "epoch": 2.809317963496638,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042167641936119557,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23151748031377792,
      "num_tokens": 67065649.0,
      "step": 29245
    },
    {
      "entropy": 5.170576000213623,
      "epoch": 2.8097982708933715,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004216504201273802,
      "loss": 4.8664,
      "mean_token_accuracy": 0.2236305996775627,
      "num_tokens": 67077714.0,
      "step": 29250
    },
    {
      "entropy": 5.192189502716064,
      "epoch": 2.810278578290106,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004216244174887776,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2336450397968292,
      "num_tokens": 67089674.0,
      "step": 29255
    },
    {
      "entropy": 5.060421943664551,
      "epoch": 2.8107588856868397,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042159841144599145,
      "loss": 4.7418,
      "mean_token_accuracy": 0.23831888288259506,
      "num_tokens": 67101843.0,
      "step": 29260
    },
    {
      "entropy": 5.1833305835723875,
      "epoch": 2.8112391930835736,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042157240199962537,
      "loss": 4.8123,
      "mean_token_accuracy": 0.2337646469473839,
      "num_tokens": 67113529.0,
      "step": 29265
    },
    {
      "entropy": 5.137193918228149,
      "epoch": 2.8117195004803075,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004215463891502834,
      "loss": 4.7519,
      "mean_token_accuracy": 0.23867789804935455,
      "num_tokens": 67125067.0,
      "step": 29270
    },
    {
      "entropy": 5.048962259292603,
      "epoch": 2.8121998078770414,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042152037289856954,
      "loss": 4.7498,
      "mean_token_accuracy": 0.23911434561014175,
      "num_tokens": 67137550.0,
      "step": 29275
    },
    {
      "entropy": 5.123716497421265,
      "epoch": 2.8126801152737753,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042149435324508755,
      "loss": 4.6807,
      "mean_token_accuracy": 0.24376419186592102,
      "num_tokens": 67149730.0,
      "step": 29280
    },
    {
      "entropy": 5.1606498718261715,
      "epoch": 2.813160422670509,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004214683301904417,
      "loss": 4.8128,
      "mean_token_accuracy": 0.2357403516769409,
      "num_tokens": 67161050.0,
      "step": 29285
    },
    {
      "entropy": 5.07976508140564,
      "epoch": 2.813640730067243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042144230373523624,
      "loss": 4.7718,
      "mean_token_accuracy": 0.23318505734205247,
      "num_tokens": 67173131.0,
      "step": 29290
    },
    {
      "entropy": 5.108788108825683,
      "epoch": 2.814121037463977,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004214162738800753,
      "loss": 4.7834,
      "mean_token_accuracy": 0.2329147607088089,
      "num_tokens": 67183723.0,
      "step": 29295
    },
    {
      "entropy": 5.168136310577393,
      "epoch": 2.814601344860711,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004213902406255632,
      "loss": 4.8202,
      "mean_token_accuracy": 0.23489454984664918,
      "num_tokens": 67196622.0,
      "step": 29300
    },
    {
      "entropy": 5.139545869827271,
      "epoch": 2.8150816522574447,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004213642039723044,
      "loss": 4.7737,
      "mean_token_accuracy": 0.2368985027074814,
      "num_tokens": 67208581.0,
      "step": 29305
    },
    {
      "entropy": 5.180890369415283,
      "epoch": 2.8155619596541785,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042133816392090343,
      "loss": 4.8191,
      "mean_token_accuracy": 0.23220699429512023,
      "num_tokens": 67220451.0,
      "step": 29310
    },
    {
      "entropy": 5.1144105911254885,
      "epoch": 2.8160422670509124,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042131212047196484,
      "loss": 4.7543,
      "mean_token_accuracy": 0.23407179415225982,
      "num_tokens": 67232191.0,
      "step": 29315
    },
    {
      "entropy": 5.0587705135345455,
      "epoch": 2.8165225744476468,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042128607362609317,
      "loss": 4.6543,
      "mean_token_accuracy": 0.24638958871364594,
      "num_tokens": 67242518.0,
      "step": 29320
    },
    {
      "entropy": 5.149604368209839,
      "epoch": 2.81700288184438,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042126002338389336,
      "loss": 4.8586,
      "mean_token_accuracy": 0.2301258772611618,
      "num_tokens": 67254598.0,
      "step": 29325
    },
    {
      "entropy": 5.093988370895386,
      "epoch": 2.8174831892411145,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042123396974597007,
      "loss": 4.7372,
      "mean_token_accuracy": 0.24300947785377502,
      "num_tokens": 67266054.0,
      "step": 29330
    },
    {
      "entropy": 5.188849210739136,
      "epoch": 2.817963496637848,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042120791271292823,
      "loss": 4.8882,
      "mean_token_accuracy": 0.22702818363904953,
      "num_tokens": 67277703.0,
      "step": 29335
    },
    {
      "entropy": 5.118722915649414,
      "epoch": 2.8184438040345823,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042118185228537283,
      "loss": 4.7384,
      "mean_token_accuracy": 0.2402300000190735,
      "num_tokens": 67288992.0,
      "step": 29340
    },
    {
      "entropy": 5.101164865493774,
      "epoch": 2.818924111431316,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042115578846390884,
      "loss": 4.7346,
      "mean_token_accuracy": 0.24083875566720964,
      "num_tokens": 67299728.0,
      "step": 29345
    },
    {
      "entropy": 5.224266862869262,
      "epoch": 2.81940441882805,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004211297212491414,
      "loss": 4.881,
      "mean_token_accuracy": 0.22259285151958466,
      "num_tokens": 67310836.0,
      "step": 29350
    },
    {
      "entropy": 5.149200344085694,
      "epoch": 2.819884726224784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004211036506416759,
      "loss": 4.8215,
      "mean_token_accuracy": 0.23451528698205948,
      "num_tokens": 67323426.0,
      "step": 29355
    },
    {
      "entropy": 5.133793306350708,
      "epoch": 2.820365033621518,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004210775766421173,
      "loss": 4.7948,
      "mean_token_accuracy": 0.23362186849117278,
      "num_tokens": 67334909.0,
      "step": 29360
    },
    {
      "entropy": 5.107807350158692,
      "epoch": 2.8208453410182517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004210514992510713,
      "loss": 4.7251,
      "mean_token_accuracy": 0.2389359265565872,
      "num_tokens": 67347315.0,
      "step": 29365
    },
    {
      "entropy": 5.086011362075806,
      "epoch": 2.8213256484149856,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004210254184691431,
      "loss": 4.7068,
      "mean_token_accuracy": 0.24709315299987794,
      "num_tokens": 67358092.0,
      "step": 29370
    },
    {
      "entropy": 5.122100734710694,
      "epoch": 2.8218059558117194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042099933429693814,
      "loss": 4.8616,
      "mean_token_accuracy": 0.23476725369691848,
      "num_tokens": 67369318.0,
      "step": 29375
    },
    {
      "entropy": 5.150946998596192,
      "epoch": 2.8222862632084533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004209732467350624,
      "loss": 4.7993,
      "mean_token_accuracy": 0.23458718806505202,
      "num_tokens": 67379604.0,
      "step": 29380
    },
    {
      "entropy": 5.141421556472778,
      "epoch": 2.822766570605187,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004209471557841212,
      "loss": 4.7225,
      "mean_token_accuracy": 0.24724196046590804,
      "num_tokens": 67391134.0,
      "step": 29385
    },
    {
      "entropy": 5.1156408309936525,
      "epoch": 2.823246878001921,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004209210614447204,
      "loss": 4.7019,
      "mean_token_accuracy": 0.24700187742710114,
      "num_tokens": 67400868.0,
      "step": 29390
    },
    {
      "entropy": 5.129614543914795,
      "epoch": 2.8237271853986554,
      "grad_norm": 1.140625,
      "learning_rate": 0.000420894963717466,
      "loss": 4.7753,
      "mean_token_accuracy": 0.22774496972560881,
      "num_tokens": 67412172.0,
      "step": 29395
    },
    {
      "entropy": 5.075762462615967,
      "epoch": 2.824207492795389,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004208688626029636,
      "loss": 4.7335,
      "mean_token_accuracy": 0.23525745123624803,
      "num_tokens": 67423700.0,
      "step": 29400
    },
    {
      "entropy": 5.190551996231079,
      "epoch": 2.824687800192123,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004208427581018194,
      "loss": 4.8702,
      "mean_token_accuracy": 0.22871236950159074,
      "num_tokens": 67436471.0,
      "step": 29405
    },
    {
      "entropy": 5.23363037109375,
      "epoch": 2.8251681075888566,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004208166502146394,
      "loss": 4.892,
      "mean_token_accuracy": 0.2258172556757927,
      "num_tokens": 67447832.0,
      "step": 29410
    },
    {
      "entropy": 5.15479416847229,
      "epoch": 2.825648414985591,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042079053894202977,
      "loss": 4.8393,
      "mean_token_accuracy": 0.23405761867761612,
      "num_tokens": 67460467.0,
      "step": 29415
    },
    {
      "entropy": 5.160784673690796,
      "epoch": 2.826128722382325,
      "grad_norm": 1.0,
      "learning_rate": 0.0004207644242845968,
      "loss": 4.7753,
      "mean_token_accuracy": 0.2372881919145584,
      "num_tokens": 67471979.0,
      "step": 29420
    },
    {
      "entropy": 5.118109321594238,
      "epoch": 2.8266090297790587,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004207383062429467,
      "loss": 4.7973,
      "mean_token_accuracy": 0.22974623441696168,
      "num_tokens": 67483441.0,
      "step": 29425
    },
    {
      "entropy": 5.207797718048096,
      "epoch": 2.8270893371757926,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004207121848176858,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22874169498682023,
      "num_tokens": 67496691.0,
      "step": 29430
    },
    {
      "entropy": 5.204637479782105,
      "epoch": 2.8275696445725265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042068606000942075,
      "loss": 4.837,
      "mean_token_accuracy": 0.23120469450950623,
      "num_tokens": 67508120.0,
      "step": 29435
    },
    {
      "entropy": 5.152711296081543,
      "epoch": 2.8280499519692603,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042065993181875794,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22387812435626983,
      "num_tokens": 67519699.0,
      "step": 29440
    },
    {
      "entropy": 5.104093742370606,
      "epoch": 2.8285302593659942,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000420633800246304,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22917501181364058,
      "num_tokens": 67530842.0,
      "step": 29445
    },
    {
      "entropy": 5.193478918075561,
      "epoch": 2.829010566762728,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042060766529266577,
      "loss": 4.8408,
      "mean_token_accuracy": 0.23078052699565887,
      "num_tokens": 67543447.0,
      "step": 29450
    },
    {
      "entropy": 5.196764087677002,
      "epoch": 2.829490874159462,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042058152695844986,
      "loss": 4.8034,
      "mean_token_accuracy": 0.23855597376823426,
      "num_tokens": 67555339.0,
      "step": 29455
    },
    {
      "entropy": 5.097189712524414,
      "epoch": 2.829971181556196,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042055538524426317,
      "loss": 4.7575,
      "mean_token_accuracy": 0.2399198442697525,
      "num_tokens": 67566747.0,
      "step": 29460
    },
    {
      "entropy": 5.052466011047363,
      "epoch": 2.8304514889529298,
      "grad_norm": 1.0,
      "learning_rate": 0.0004205292401507127,
      "loss": 4.7111,
      "mean_token_accuracy": 0.2422829180955887,
      "num_tokens": 67577690.0,
      "step": 29465
    },
    {
      "entropy": 5.221011924743652,
      "epoch": 2.8309317963496636,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004205030916784053,
      "loss": 4.9299,
      "mean_token_accuracy": 0.22454874515533446,
      "num_tokens": 67589287.0,
      "step": 29470
    },
    {
      "entropy": 5.176647138595581,
      "epoch": 2.8314121037463975,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042047693982794824,
      "loss": 4.7799,
      "mean_token_accuracy": 0.23140384554862975,
      "num_tokens": 67602233.0,
      "step": 29475
    },
    {
      "entropy": 5.09964509010315,
      "epoch": 2.831892411143132,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042045078459994854,
      "loss": 4.7392,
      "mean_token_accuracy": 0.23739102333784104,
      "num_tokens": 67613076.0,
      "step": 29480
    },
    {
      "entropy": 5.103259086608887,
      "epoch": 2.8323727185398653,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004204246259950136,
      "loss": 4.769,
      "mean_token_accuracy": 0.23806957602500917,
      "num_tokens": 67623517.0,
      "step": 29485
    },
    {
      "entropy": 5.1753967761993405,
      "epoch": 2.8328530259365996,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042039846401375065,
      "loss": 4.7895,
      "mean_token_accuracy": 0.23485387116670609,
      "num_tokens": 67635982.0,
      "step": 29490
    },
    {
      "entropy": 5.111962747573853,
      "epoch": 2.8333333333333335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042037229865676714,
      "loss": 4.6927,
      "mean_token_accuracy": 0.2400985285639763,
      "num_tokens": 67646436.0,
      "step": 29495
    },
    {
      "entropy": 5.164253854751587,
      "epoch": 2.8338136407300674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042034612992467046,
      "loss": 4.7824,
      "mean_token_accuracy": 0.2425023064017296,
      "num_tokens": 67657436.0,
      "step": 29500
    },
    {
      "entropy": 5.122543859481811,
      "epoch": 2.8342939481268012,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004203199578180683,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23240161538124085,
      "num_tokens": 67669542.0,
      "step": 29505
    },
    {
      "entropy": 5.093420505523682,
      "epoch": 2.834774255523535,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004202937823375682,
      "loss": 4.6702,
      "mean_token_accuracy": 0.2497760236263275,
      "num_tokens": 67680512.0,
      "step": 29510
    },
    {
      "entropy": 5.165628147125244,
      "epoch": 2.835254562920269,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004202676034837779,
      "loss": 4.745,
      "mean_token_accuracy": 0.23669492602348327,
      "num_tokens": 67691392.0,
      "step": 29515
    },
    {
      "entropy": 5.120699834823609,
      "epoch": 2.835734870317003,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004202414212573052,
      "loss": 4.7696,
      "mean_token_accuracy": 0.24279378950595856,
      "num_tokens": 67701674.0,
      "step": 29520
    },
    {
      "entropy": 5.1163524150848385,
      "epoch": 2.8362151777137368,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042021523565875796,
      "loss": 4.8316,
      "mean_token_accuracy": 0.23193923830986024,
      "num_tokens": 67713960.0,
      "step": 29525
    },
    {
      "entropy": 5.149273014068603,
      "epoch": 2.8366954851104706,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004201890466887442,
      "loss": 4.7352,
      "mean_token_accuracy": 0.24509025514125823,
      "num_tokens": 67725644.0,
      "step": 29530
    },
    {
      "entropy": 5.123471975326538,
      "epoch": 2.8371757925072045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004201628543478718,
      "loss": 4.7465,
      "mean_token_accuracy": 0.24366096705198287,
      "num_tokens": 67737997.0,
      "step": 29535
    },
    {
      "entropy": 5.106903505325318,
      "epoch": 2.8376560999039384,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000420136658636749,
      "loss": 4.809,
      "mean_token_accuracy": 0.23146942555904387,
      "num_tokens": 67750379.0,
      "step": 29540
    },
    {
      "entropy": 5.166104078292847,
      "epoch": 2.8381364073006723,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004201104595559841,
      "loss": 4.8633,
      "mean_token_accuracy": 0.2277505323290825,
      "num_tokens": 67762567.0,
      "step": 29545
    },
    {
      "entropy": 5.149979829788208,
      "epoch": 2.838616714697406,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042008425710618507,
      "loss": 4.7825,
      "mean_token_accuracy": 0.23784510046243668,
      "num_tokens": 67774660.0,
      "step": 29550
    },
    {
      "entropy": 5.172838306427002,
      "epoch": 2.8390970220941405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042005805128796043,
      "loss": 4.752,
      "mean_token_accuracy": 0.2398751899600029,
      "num_tokens": 67784757.0,
      "step": 29555
    },
    {
      "entropy": 5.118153142929077,
      "epoch": 2.839577329490874,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004200318421019186,
      "loss": 4.74,
      "mean_token_accuracy": 0.23778224140405654,
      "num_tokens": 67795723.0,
      "step": 29560
    },
    {
      "entropy": 5.258893775939941,
      "epoch": 2.8400576368876083,
      "grad_norm": 0.96875,
      "learning_rate": 0.000420005629548668,
      "loss": 4.9365,
      "mean_token_accuracy": 0.22778963595628737,
      "num_tokens": 67807685.0,
      "step": 29565
    },
    {
      "entropy": 5.116344928741455,
      "epoch": 2.840537944284342,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041997941362881735,
      "loss": 4.6985,
      "mean_token_accuracy": 0.24673137813806534,
      "num_tokens": 67818887.0,
      "step": 29570
    },
    {
      "entropy": 5.130417966842652,
      "epoch": 2.841018251681076,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004199531943429752,
      "loss": 4.7988,
      "mean_token_accuracy": 0.24115533530712127,
      "num_tokens": 67830210.0,
      "step": 29575
    },
    {
      "entropy": 5.082629442214966,
      "epoch": 2.84149855907781,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004199269716917502,
      "loss": 4.7891,
      "mean_token_accuracy": 0.23390978276729585,
      "num_tokens": 67841948.0,
      "step": 29580
    },
    {
      "entropy": 5.168309736251831,
      "epoch": 2.841978866474544,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004199007456757513,
      "loss": 4.7673,
      "mean_token_accuracy": 0.23433402180671692,
      "num_tokens": 67854117.0,
      "step": 29585
    },
    {
      "entropy": 5.17713041305542,
      "epoch": 2.8424591738712777,
      "grad_norm": 0.875,
      "learning_rate": 0.00041987451629558743,
      "loss": 4.82,
      "mean_token_accuracy": 0.2324904039502144,
      "num_tokens": 67866004.0,
      "step": 29590
    },
    {
      "entropy": 5.123828744888305,
      "epoch": 2.8429394812680115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004198482835518674,
      "loss": 4.7535,
      "mean_token_accuracy": 0.24386327862739562,
      "num_tokens": 67877178.0,
      "step": 29595
    },
    {
      "entropy": 5.068818283081055,
      "epoch": 2.8434197886647454,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004198220474452004,
      "loss": 4.726,
      "mean_token_accuracy": 0.24055779576301575,
      "num_tokens": 67888183.0,
      "step": 29600
    },
    {
      "entropy": 5.22925181388855,
      "epoch": 2.8439000960614793,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004197958079761954,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22748487442731857,
      "num_tokens": 67899300.0,
      "step": 29605
    },
    {
      "entropy": 5.131304454803467,
      "epoch": 2.844380403458213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041976956514546185,
      "loss": 4.7476,
      "mean_token_accuracy": 0.24263266026973723,
      "num_tokens": 67909494.0,
      "step": 29610
    },
    {
      "entropy": 5.157851457595825,
      "epoch": 2.844860710854947,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041974331895360873,
      "loss": 4.8408,
      "mean_token_accuracy": 0.23151797652244568,
      "num_tokens": 67921241.0,
      "step": 29615
    },
    {
      "entropy": 5.08039927482605,
      "epoch": 2.845341018251681,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004197170694012456,
      "loss": 4.7061,
      "mean_token_accuracy": 0.24066463112831116,
      "num_tokens": 67933571.0,
      "step": 29620
    },
    {
      "entropy": 5.182117557525634,
      "epoch": 2.845821325648415,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004196908164889818,
      "loss": 4.8292,
      "mean_token_accuracy": 0.23444078266620635,
      "num_tokens": 67944100.0,
      "step": 29625
    },
    {
      "entropy": 5.13575758934021,
      "epoch": 2.846301633045149,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004196645602174269,
      "loss": 4.7452,
      "mean_token_accuracy": 0.23667082041502,
      "num_tokens": 67955908.0,
      "step": 29630
    },
    {
      "entropy": 5.1823218822479244,
      "epoch": 2.8467819404418826,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041963830058719046,
      "loss": 4.8056,
      "mean_token_accuracy": 0.23114986419677735,
      "num_tokens": 67966763.0,
      "step": 29635
    },
    {
      "entropy": 5.09346342086792,
      "epoch": 2.847262247838617,
      "grad_norm": 1.0,
      "learning_rate": 0.0004196120375988822,
      "loss": 4.7328,
      "mean_token_accuracy": 0.2357368677854538,
      "num_tokens": 67977668.0,
      "step": 29640
    },
    {
      "entropy": 5.087451457977295,
      "epoch": 2.8477425552353504,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004195857712531119,
      "loss": 4.7735,
      "mean_token_accuracy": 0.24173774272203447,
      "num_tokens": 67989429.0,
      "step": 29645
    },
    {
      "entropy": 5.219403648376465,
      "epoch": 2.8482228626320847,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004195595015504892,
      "loss": 4.8586,
      "mean_token_accuracy": 0.23345735520124436,
      "num_tokens": 68000911.0,
      "step": 29650
    },
    {
      "entropy": 5.1388298034667965,
      "epoch": 2.8487031700288186,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041953322849162415,
      "loss": 4.7429,
      "mean_token_accuracy": 0.24055960923433303,
      "num_tokens": 68012804.0,
      "step": 29655
    },
    {
      "entropy": 5.185705184936523,
      "epoch": 2.8491834774255524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004195069520771268,
      "loss": 4.829,
      "mean_token_accuracy": 0.23207500129938125,
      "num_tokens": 68023508.0,
      "step": 29660
    },
    {
      "entropy": 5.149076509475708,
      "epoch": 2.8496637848222863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041948067230760706,
      "loss": 4.8169,
      "mean_token_accuracy": 0.23391374498605727,
      "num_tokens": 68035383.0,
      "step": 29665
    },
    {
      "entropy": 5.195984315872193,
      "epoch": 2.85014409221902,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041945438918367513,
      "loss": 4.8023,
      "mean_token_accuracy": 0.22656171321868895,
      "num_tokens": 68045922.0,
      "step": 29670
    },
    {
      "entropy": 5.005151653289795,
      "epoch": 2.850624399615754,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041942810270594115,
      "loss": 4.7221,
      "mean_token_accuracy": 0.23693075329065322,
      "num_tokens": 68057291.0,
      "step": 29675
    },
    {
      "entropy": 5.152499103546143,
      "epoch": 2.851104707012488,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004194018128750157,
      "loss": 4.8591,
      "mean_token_accuracy": 0.23153621554374695,
      "num_tokens": 68069693.0,
      "step": 29680
    },
    {
      "entropy": 5.1717610359191895,
      "epoch": 2.851585014409222,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041937551969150873,
      "loss": 4.8108,
      "mean_token_accuracy": 0.23444279730319978,
      "num_tokens": 68079696.0,
      "step": 29685
    },
    {
      "entropy": 5.113613748550415,
      "epoch": 2.8520653218059557,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000419349223156031,
      "loss": 4.7563,
      "mean_token_accuracy": 0.23810358643531798,
      "num_tokens": 68091330.0,
      "step": 29690
    },
    {
      "entropy": 5.113778495788575,
      "epoch": 2.8525456292026896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004193229232691929,
      "loss": 4.7091,
      "mean_token_accuracy": 0.24023185819387435,
      "num_tokens": 68102499.0,
      "step": 29695
    },
    {
      "entropy": 5.132311582565308,
      "epoch": 2.8530259365994235,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041929662003160504,
      "loss": 4.7795,
      "mean_token_accuracy": 0.2315684199333191,
      "num_tokens": 68114563.0,
      "step": 29700
    },
    {
      "entropy": 5.1274505138397215,
      "epoch": 2.8535062439961574,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041927031344387824,
      "loss": 4.8085,
      "mean_token_accuracy": 0.235743248462677,
      "num_tokens": 68124798.0,
      "step": 29705
    },
    {
      "entropy": 5.195479822158814,
      "epoch": 2.8539865513928913,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041924400350662304,
      "loss": 4.872,
      "mean_token_accuracy": 0.22560183256864547,
      "num_tokens": 68136190.0,
      "step": 29710
    },
    {
      "entropy": 5.150496101379394,
      "epoch": 2.8544668587896256,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041921769022045045,
      "loss": 4.7731,
      "mean_token_accuracy": 0.241357484459877,
      "num_tokens": 68146368.0,
      "step": 29715
    },
    {
      "entropy": 5.168339824676513,
      "epoch": 2.854947166186359,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041919137358597137,
      "loss": 4.8526,
      "mean_token_accuracy": 0.23076463490724564,
      "num_tokens": 68159338.0,
      "step": 29720
    },
    {
      "entropy": 5.227389669418335,
      "epoch": 2.8554274735830933,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004191650536037967,
      "loss": 4.7958,
      "mean_token_accuracy": 0.23148199617862703,
      "num_tokens": 68170906.0,
      "step": 29725
    },
    {
      "entropy": 5.194037199020386,
      "epoch": 2.8559077809798272,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041913873027453756,
      "loss": 4.7801,
      "mean_token_accuracy": 0.2337573140859604,
      "num_tokens": 68181963.0,
      "step": 29730
    },
    {
      "entropy": 5.066288948059082,
      "epoch": 2.856388088376561,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041911240359880517,
      "loss": 4.6919,
      "mean_token_accuracy": 0.24236624538898469,
      "num_tokens": 68193184.0,
      "step": 29735
    },
    {
      "entropy": 5.14004545211792,
      "epoch": 2.856868395773295,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041908607357721067,
      "loss": 4.8154,
      "mean_token_accuracy": 0.24409887343645095,
      "num_tokens": 68203984.0,
      "step": 29740
    },
    {
      "entropy": 5.144165897369385,
      "epoch": 2.857348703170029,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041905974021036533,
      "loss": 4.7196,
      "mean_token_accuracy": 0.2419520676136017,
      "num_tokens": 68215294.0,
      "step": 29745
    },
    {
      "entropy": 5.105319499969482,
      "epoch": 2.8578290105667628,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041903340349888065,
      "loss": 4.8309,
      "mean_token_accuracy": 0.24032192379236222,
      "num_tokens": 68226800.0,
      "step": 29750
    },
    {
      "entropy": 5.109892511367798,
      "epoch": 2.8583093179634966,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000419007063443368,
      "loss": 4.767,
      "mean_token_accuracy": 0.24264518171548843,
      "num_tokens": 68238868.0,
      "step": 29755
    },
    {
      "entropy": 5.148657035827637,
      "epoch": 2.8587896253602305,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041898072004443906,
      "loss": 4.798,
      "mean_token_accuracy": 0.2445417031645775,
      "num_tokens": 68250133.0,
      "step": 29760
    },
    {
      "entropy": 5.182205390930176,
      "epoch": 2.8592699327569644,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004189543733027052,
      "loss": 4.8164,
      "mean_token_accuracy": 0.22813091725111007,
      "num_tokens": 68261338.0,
      "step": 29765
    },
    {
      "entropy": 5.195447635650635,
      "epoch": 2.8597502401536983,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004189280232187783,
      "loss": 4.7806,
      "mean_token_accuracy": 0.23717291802167892,
      "num_tokens": 68272369.0,
      "step": 29770
    },
    {
      "entropy": 5.114936256408692,
      "epoch": 2.860230547550432,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004189016697932701,
      "loss": 4.8094,
      "mean_token_accuracy": 0.2418594017624855,
      "num_tokens": 68283997.0,
      "step": 29775
    },
    {
      "entropy": 5.069814014434814,
      "epoch": 2.860710854947166,
      "grad_norm": 1.0,
      "learning_rate": 0.0004188753130267924,
      "loss": 4.7331,
      "mean_token_accuracy": 0.23284811675548553,
      "num_tokens": 68295719.0,
      "step": 29780
    },
    {
      "entropy": 5.110177803039551,
      "epoch": 2.8611911623439,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004188489529199572,
      "loss": 4.8169,
      "mean_token_accuracy": 0.23893224447965622,
      "num_tokens": 68307352.0,
      "step": 29785
    },
    {
      "entropy": 5.102125644683838,
      "epoch": 2.8616714697406342,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041882258947337637,
      "loss": 4.6906,
      "mean_token_accuracy": 0.23819111436605453,
      "num_tokens": 68318889.0,
      "step": 29790
    },
    {
      "entropy": 5.059566354751587,
      "epoch": 2.8621517771373677,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041879622268766207,
      "loss": 4.6658,
      "mean_token_accuracy": 0.2432067036628723,
      "num_tokens": 68329681.0,
      "step": 29795
    },
    {
      "entropy": 5.128244686126709,
      "epoch": 2.862632084534102,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004187698525634266,
      "loss": 4.7992,
      "mean_token_accuracy": 0.2409598261117935,
      "num_tokens": 68340323.0,
      "step": 29800
    },
    {
      "entropy": 5.0467276096344,
      "epoch": 2.863112391930836,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041874347910128193,
      "loss": 4.6366,
      "mean_token_accuracy": 0.24680333137512206,
      "num_tokens": 68351732.0,
      "step": 29805
    },
    {
      "entropy": 5.15931544303894,
      "epoch": 2.8635926993275698,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004187171023018406,
      "loss": 4.7997,
      "mean_token_accuracy": 0.23631879687309265,
      "num_tokens": 68362847.0,
      "step": 29810
    },
    {
      "entropy": 5.117188119888306,
      "epoch": 2.8640730067243036,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041869072216571486,
      "loss": 4.7397,
      "mean_token_accuracy": 0.23919386714696883,
      "num_tokens": 68374368.0,
      "step": 29815
    },
    {
      "entropy": 5.060667657852173,
      "epoch": 2.8645533141210375,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041866433869351715,
      "loss": 4.738,
      "mean_token_accuracy": 0.23731100410223008,
      "num_tokens": 68386463.0,
      "step": 29820
    },
    {
      "entropy": 5.0868151664733885,
      "epoch": 2.8650336215177714,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004186379518858602,
      "loss": 4.7203,
      "mean_token_accuracy": 0.24482073932886123,
      "num_tokens": 68397873.0,
      "step": 29825
    },
    {
      "entropy": 5.172638463973999,
      "epoch": 2.8655139289145053,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004186115617433565,
      "loss": 4.8082,
      "mean_token_accuracy": 0.22521547079086304,
      "num_tokens": 68408561.0,
      "step": 29830
    },
    {
      "entropy": 5.11163444519043,
      "epoch": 2.865994236311239,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041858516826661876,
      "loss": 4.699,
      "mean_token_accuracy": 0.24324326515197753,
      "num_tokens": 68419878.0,
      "step": 29835
    },
    {
      "entropy": 5.143398094177246,
      "epoch": 2.866474543707973,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041855877145625974,
      "loss": 4.8145,
      "mean_token_accuracy": 0.24528425484895705,
      "num_tokens": 68429913.0,
      "step": 29840
    },
    {
      "entropy": 5.1443346500396725,
      "epoch": 2.866954851104707,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004185323713128924,
      "loss": 4.8194,
      "mean_token_accuracy": 0.2373996287584305,
      "num_tokens": 68441881.0,
      "step": 29845
    },
    {
      "entropy": 5.175118017196655,
      "epoch": 2.867435158501441,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041850596783712956,
      "loss": 4.7584,
      "mean_token_accuracy": 0.23672414422035218,
      "num_tokens": 68453191.0,
      "step": 29850
    },
    {
      "entropy": 5.1678468704223635,
      "epoch": 2.8679154658981747,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004184795610295843,
      "loss": 4.7793,
      "mean_token_accuracy": 0.24150069504976274,
      "num_tokens": 68464373.0,
      "step": 29855
    },
    {
      "entropy": 5.151861572265625,
      "epoch": 2.8683957732949086,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004184531508908697,
      "loss": 4.8052,
      "mean_token_accuracy": 0.23484041690826415,
      "num_tokens": 68474970.0,
      "step": 29860
    },
    {
      "entropy": 5.070051145553589,
      "epoch": 2.868876080691643,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004184267374215989,
      "loss": 4.7658,
      "mean_token_accuracy": 0.23883662968873978,
      "num_tokens": 68486587.0,
      "step": 29865
    },
    {
      "entropy": 5.1345032215118405,
      "epoch": 2.8693563880883763,
      "grad_norm": 0.96875,
      "learning_rate": 0.000418400320622385,
      "loss": 4.8095,
      "mean_token_accuracy": 0.23656003326177596,
      "num_tokens": 68498091.0,
      "step": 29870
    },
    {
      "entropy": 5.161547613143921,
      "epoch": 2.8698366954851107,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004183739004938416,
      "loss": 4.7809,
      "mean_token_accuracy": 0.23408834487199784,
      "num_tokens": 68510431.0,
      "step": 29875
    },
    {
      "entropy": 5.0592268943786625,
      "epoch": 2.870317002881844,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000418347477036582,
      "loss": 4.68,
      "mean_token_accuracy": 0.24768368899822235,
      "num_tokens": 68522085.0,
      "step": 29880
    },
    {
      "entropy": 5.137062883377075,
      "epoch": 2.8707973102785784,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041832105025121956,
      "loss": 4.776,
      "mean_token_accuracy": 0.23071133494377136,
      "num_tokens": 68533251.0,
      "step": 29885
    },
    {
      "entropy": 5.155922794342041,
      "epoch": 2.8712776176753123,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004182946201383679,
      "loss": 4.7111,
      "mean_token_accuracy": 0.23883095383644104,
      "num_tokens": 68545131.0,
      "step": 29890
    },
    {
      "entropy": 5.093001127243042,
      "epoch": 2.871757925072046,
      "grad_norm": 1.0,
      "learning_rate": 0.0004182681866986407,
      "loss": 4.7452,
      "mean_token_accuracy": 0.24040748476982116,
      "num_tokens": 68556838.0,
      "step": 29895
    },
    {
      "entropy": 5.140456581115723,
      "epoch": 2.87223823246878,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041824174993265165,
      "loss": 4.7854,
      "mean_token_accuracy": 0.2370212569832802,
      "num_tokens": 68570506.0,
      "step": 29900
    },
    {
      "entropy": 5.136846113204956,
      "epoch": 2.872718539865514,
      "grad_norm": 1.0,
      "learning_rate": 0.00041821530984101444,
      "loss": 4.7254,
      "mean_token_accuracy": 0.24010846614837647,
      "num_tokens": 68581364.0,
      "step": 29905
    },
    {
      "entropy": 5.116699504852295,
      "epoch": 2.873198847262248,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000418188866424343,
      "loss": 4.7463,
      "mean_token_accuracy": 0.2350742369890213,
      "num_tokens": 68593047.0,
      "step": 29910
    },
    {
      "entropy": 5.148551416397095,
      "epoch": 2.8736791546589817,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004181624196832513,
      "loss": 4.7834,
      "mean_token_accuracy": 0.2346457913517952,
      "num_tokens": 68604592.0,
      "step": 29915
    },
    {
      "entropy": 5.159334659576416,
      "epoch": 2.8741594620557156,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00041813596961835336,
      "loss": 4.7757,
      "mean_token_accuracy": 0.23639463186264037,
      "num_tokens": 68616566.0,
      "step": 29920
    },
    {
      "entropy": 5.020766687393189,
      "epoch": 2.8746397694524495,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041810951623026313,
      "loss": 4.6426,
      "mean_token_accuracy": 0.24727847278118134,
      "num_tokens": 68626507.0,
      "step": 29925
    },
    {
      "entropy": 5.0200214862823485,
      "epoch": 2.8751200768491834,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041808305951959496,
      "loss": 4.632,
      "mean_token_accuracy": 0.24927257746458054,
      "num_tokens": 68637748.0,
      "step": 29930
    },
    {
      "entropy": 5.122966766357422,
      "epoch": 2.8756003842459172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000418056599486963,
      "loss": 4.777,
      "mean_token_accuracy": 0.23944731503725053,
      "num_tokens": 68649379.0,
      "step": 29935
    },
    {
      "entropy": 5.15406379699707,
      "epoch": 2.8760806916426516,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004180301361329816,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23179761469364166,
      "num_tokens": 68660834.0,
      "step": 29940
    },
    {
      "entropy": 5.082159805297851,
      "epoch": 2.876560999039385,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004180036694582651,
      "loss": 4.7941,
      "mean_token_accuracy": 0.23212667405605317,
      "num_tokens": 68673748.0,
      "step": 29945
    },
    {
      "entropy": 5.12637996673584,
      "epoch": 2.8770413064361193,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041797719946342813,
      "loss": 4.768,
      "mean_token_accuracy": 0.23559096306562424,
      "num_tokens": 68684594.0,
      "step": 29950
    },
    {
      "entropy": 5.218717813491821,
      "epoch": 2.8775216138328528,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041795072614908503,
      "loss": 4.8872,
      "mean_token_accuracy": 0.2234296977519989,
      "num_tokens": 68697954.0,
      "step": 29955
    },
    {
      "entropy": 5.235721111297607,
      "epoch": 2.878001921229587,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041792424951585055,
      "loss": 4.9038,
      "mean_token_accuracy": 0.2353292390704155,
      "num_tokens": 68708616.0,
      "step": 29960
    },
    {
      "entropy": 4.9854833602905275,
      "epoch": 2.878482228626321,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041789776956433947,
      "loss": 4.6393,
      "mean_token_accuracy": 0.25177151709795,
      "num_tokens": 68720226.0,
      "step": 29965
    },
    {
      "entropy": 5.143875932693481,
      "epoch": 2.878962536023055,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041787128629516645,
      "loss": 4.7693,
      "mean_token_accuracy": 0.2361595258116722,
      "num_tokens": 68731134.0,
      "step": 29970
    },
    {
      "entropy": 5.170514059066773,
      "epoch": 2.8794428434197887,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004178447997089464,
      "loss": 4.8318,
      "mean_token_accuracy": 0.2381017029285431,
      "num_tokens": 68743158.0,
      "step": 29975
    },
    {
      "entropy": 5.126296949386597,
      "epoch": 2.8799231508165226,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004178183098062943,
      "loss": 4.8027,
      "mean_token_accuracy": 0.23677198439836503,
      "num_tokens": 68755411.0,
      "step": 29980
    },
    {
      "entropy": 5.13998498916626,
      "epoch": 2.8804034582132565,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004177918165878251,
      "loss": 4.8302,
      "mean_token_accuracy": 0.23692332357168197,
      "num_tokens": 68767621.0,
      "step": 29985
    },
    {
      "entropy": 5.083087539672851,
      "epoch": 2.8808837656099904,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004177653200541539,
      "loss": 4.7195,
      "mean_token_accuracy": 0.24462345838546753,
      "num_tokens": 68778861.0,
      "step": 29990
    },
    {
      "entropy": 5.1405065059661865,
      "epoch": 2.8813640730067243,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000417738820205896,
      "loss": 4.7545,
      "mean_token_accuracy": 0.23947452008724213,
      "num_tokens": 68790101.0,
      "step": 29995
    },
    {
      "entropy": 5.020632696151734,
      "epoch": 2.881844380403458,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004177123170436665,
      "loss": 4.653,
      "mean_token_accuracy": 0.24817303866147994,
      "num_tokens": 68801612.0,
      "step": 30000
    },
    {
      "epoch": 2.881844380403458,
      "eval_entropy": 4.925132572098746,
      "eval_loss": 4.862937927246094,
      "eval_mean_token_accuracy": 0.24089105648312378,
      "eval_num_tokens": 68801612.0,
      "eval_runtime": 26.6205,
      "eval_samples_per_second": 1232.698,
      "eval_steps_per_second": 154.092,
      "step": 30000
    },
    {
      "entropy": 5.153954935073853,
      "epoch": 2.882324687800192,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004176858105680807,
      "loss": 4.8758,
      "mean_token_accuracy": 0.2293311059474945,
      "num_tokens": 68814322.0,
      "step": 30005
    },
    {
      "entropy": 5.120109701156617,
      "epoch": 2.882804995196926,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041765930077975415,
      "loss": 4.6821,
      "mean_token_accuracy": 0.2494473174214363,
      "num_tokens": 68825024.0,
      "step": 30010
    },
    {
      "entropy": 5.136367893218994,
      "epoch": 2.88328530259366,
      "grad_norm": 1.0,
      "learning_rate": 0.00041763278767930213,
      "loss": 4.7332,
      "mean_token_accuracy": 0.24265310168266296,
      "num_tokens": 68834751.0,
      "step": 30015
    },
    {
      "entropy": 4.9642963886260985,
      "epoch": 2.8837656099903937,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004176062712673404,
      "loss": 4.6158,
      "mean_token_accuracy": 0.2531613454222679,
      "num_tokens": 68845613.0,
      "step": 30020
    },
    {
      "entropy": 5.094738721847534,
      "epoch": 2.884245917387128,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004175797515444845,
      "loss": 4.7023,
      "mean_token_accuracy": 0.24137797355651855,
      "num_tokens": 68856902.0,
      "step": 30025
    },
    {
      "entropy": 5.156707239151001,
      "epoch": 2.8847262247838614,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004175532285113501,
      "loss": 4.8094,
      "mean_token_accuracy": 0.23547806441783906,
      "num_tokens": 68868612.0,
      "step": 30030
    },
    {
      "entropy": 5.064518690109253,
      "epoch": 2.8852065321805958,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004175267021685531,
      "loss": 4.7044,
      "mean_token_accuracy": 0.24593541920185089,
      "num_tokens": 68880213.0,
      "step": 30035
    },
    {
      "entropy": 5.149133682250977,
      "epoch": 2.8856868395773296,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041750017251670926,
      "loss": 4.7835,
      "mean_token_accuracy": 0.2372957706451416,
      "num_tokens": 68891844.0,
      "step": 30040
    },
    {
      "entropy": 5.069781255722046,
      "epoch": 2.8861671469740635,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004174736395564345,
      "loss": 4.6336,
      "mean_token_accuracy": 0.2431752935051918,
      "num_tokens": 68904566.0,
      "step": 30045
    },
    {
      "entropy": 5.08394865989685,
      "epoch": 2.8866474543707974,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041744710328834493,
      "loss": 4.795,
      "mean_token_accuracy": 0.2346426710486412,
      "num_tokens": 68917233.0,
      "step": 30050
    },
    {
      "entropy": 5.133976221084595,
      "epoch": 2.8871277617675313,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041742056371305665,
      "loss": 4.7615,
      "mean_token_accuracy": 0.23726814538240432,
      "num_tokens": 68928065.0,
      "step": 30055
    },
    {
      "entropy": 5.2154566764831545,
      "epoch": 2.887608069164265,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041739402083118576,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22491383403539658,
      "num_tokens": 68940130.0,
      "step": 30060
    },
    {
      "entropy": 5.077569007873535,
      "epoch": 2.888088376560999,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004173674746433485,
      "loss": 4.6542,
      "mean_token_accuracy": 0.24077947437763214,
      "num_tokens": 68951486.0,
      "step": 30065
    },
    {
      "entropy": 5.08476619720459,
      "epoch": 2.888568683957733,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041734092515016127,
      "loss": 4.6471,
      "mean_token_accuracy": 0.24742942452430725,
      "num_tokens": 68961988.0,
      "step": 30070
    },
    {
      "entropy": 5.04672532081604,
      "epoch": 2.889048991354467,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041731437235224036,
      "loss": 4.7516,
      "mean_token_accuracy": 0.24119323939085008,
      "num_tokens": 68973922.0,
      "step": 30075
    },
    {
      "entropy": 5.149067211151123,
      "epoch": 2.8895292987512007,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004172878162502023,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23259582072496415,
      "num_tokens": 68985609.0,
      "step": 30080
    },
    {
      "entropy": 5.242415571212769,
      "epoch": 2.8900096061479346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041726125684466374,
      "loss": 4.8188,
      "mean_token_accuracy": 0.23169975876808166,
      "num_tokens": 68997087.0,
      "step": 30085
    },
    {
      "entropy": 5.145707511901856,
      "epoch": 2.8904899135446684,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004172346941362412,
      "loss": 4.7702,
      "mean_token_accuracy": 0.23157819658517836,
      "num_tokens": 69008668.0,
      "step": 30090
    },
    {
      "entropy": 5.110038137435913,
      "epoch": 2.8909702209414023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041720812812555137,
      "loss": 4.7537,
      "mean_token_accuracy": 0.24072204381227494,
      "num_tokens": 69019834.0,
      "step": 30095
    },
    {
      "entropy": 5.128544044494629,
      "epoch": 2.8914505283381366,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004171815588132111,
      "loss": 4.7466,
      "mean_token_accuracy": 0.23948175758123397,
      "num_tokens": 69031072.0,
      "step": 30100
    },
    {
      "entropy": 5.174934244155883,
      "epoch": 2.89193083573487,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004171549861998372,
      "loss": 4.8467,
      "mean_token_accuracy": 0.2310475304722786,
      "num_tokens": 69043670.0,
      "step": 30105
    },
    {
      "entropy": 5.112704706192017,
      "epoch": 2.8924111431316044,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004171284102860467,
      "loss": 4.7478,
      "mean_token_accuracy": 0.2402876153588295,
      "num_tokens": 69056756.0,
      "step": 30110
    },
    {
      "entropy": 5.183635711669922,
      "epoch": 2.8928914505283383,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004171018310724565,
      "loss": 4.7727,
      "mean_token_accuracy": 0.23673189133405687,
      "num_tokens": 69069406.0,
      "step": 30115
    },
    {
      "entropy": 5.129577255249023,
      "epoch": 2.893371757925072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004170752485596838,
      "loss": 4.7775,
      "mean_token_accuracy": 0.23556047379970552,
      "num_tokens": 69081887.0,
      "step": 30120
    },
    {
      "entropy": 5.070524549484253,
      "epoch": 2.893852065321806,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041704866274834557,
      "loss": 4.7047,
      "mean_token_accuracy": 0.24107740819454193,
      "num_tokens": 69092272.0,
      "step": 30125
    },
    {
      "entropy": 5.132317972183228,
      "epoch": 2.89433237271854,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041702207363905933,
      "loss": 4.7955,
      "mean_token_accuracy": 0.23600437641143798,
      "num_tokens": 69103677.0,
      "step": 30130
    },
    {
      "entropy": 5.095260763168335,
      "epoch": 2.894812680115274,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041699548123244216,
      "loss": 4.7456,
      "mean_token_accuracy": 0.24086078405380248,
      "num_tokens": 69114863.0,
      "step": 30135
    },
    {
      "entropy": 5.1940850734710695,
      "epoch": 2.8952929875120077,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004169688855291116,
      "loss": 4.8277,
      "mean_token_accuracy": 0.23100880533456802,
      "num_tokens": 69125998.0,
      "step": 30140
    },
    {
      "entropy": 5.281360054016114,
      "epoch": 2.8957732949087416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004169422865296851,
      "loss": 4.9337,
      "mean_token_accuracy": 0.21916140168905257,
      "num_tokens": 69136706.0,
      "step": 30145
    },
    {
      "entropy": 5.20279541015625,
      "epoch": 2.8962536023054755,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004169156842347802,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2299065113067627,
      "num_tokens": 69147360.0,
      "step": 30150
    },
    {
      "entropy": 5.195073413848877,
      "epoch": 2.8967339097022093,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004168890786450144,
      "loss": 4.8863,
      "mean_token_accuracy": 0.22870680689811707,
      "num_tokens": 69159230.0,
      "step": 30155
    },
    {
      "entropy": 5.1583507537841795,
      "epoch": 2.8972142170989432,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004168624697610056,
      "loss": 4.8368,
      "mean_token_accuracy": 0.23350329995155333,
      "num_tokens": 69170469.0,
      "step": 30160
    },
    {
      "entropy": 5.244571161270142,
      "epoch": 2.897694524495677,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041683585758337156,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23220717608928682,
      "num_tokens": 69182576.0,
      "step": 30165
    },
    {
      "entropy": 5.113494348526001,
      "epoch": 2.898174831892411,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004168092421127299,
      "loss": 4.6928,
      "mean_token_accuracy": 0.24348915368318558,
      "num_tokens": 69193471.0,
      "step": 30170
    },
    {
      "entropy": 5.085099077224731,
      "epoch": 2.8986551392891453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004167826233496989,
      "loss": 4.749,
      "mean_token_accuracy": 0.232527756690979,
      "num_tokens": 69204416.0,
      "step": 30175
    },
    {
      "entropy": 5.11197075843811,
      "epoch": 2.8991354466858787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004167560012948963,
      "loss": 4.7787,
      "mean_token_accuracy": 0.2393120273947716,
      "num_tokens": 69215198.0,
      "step": 30180
    },
    {
      "entropy": 5.220945692062378,
      "epoch": 2.899615754082613,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041672937594894034,
      "loss": 4.918,
      "mean_token_accuracy": 0.22498827278614045,
      "num_tokens": 69229015.0,
      "step": 30185
    },
    {
      "entropy": 5.2682819843292235,
      "epoch": 2.9000960614793465,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041670274731244903,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22444083392620087,
      "num_tokens": 69240559.0,
      "step": 30190
    },
    {
      "entropy": 5.070811128616333,
      "epoch": 2.900576368876081,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004166761153860408,
      "loss": 4.7256,
      "mean_token_accuracy": 0.24445248544216155,
      "num_tokens": 69251741.0,
      "step": 30195
    },
    {
      "entropy": 5.095681095123291,
      "epoch": 2.9010566762728147,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041664948017033383,
      "loss": 4.7519,
      "mean_token_accuracy": 0.23481339663267137,
      "num_tokens": 69264086.0,
      "step": 30200
    },
    {
      "entropy": 5.217259263992309,
      "epoch": 2.9015369836695486,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004166228416659465,
      "loss": 4.8287,
      "mean_token_accuracy": 0.2352574124932289,
      "num_tokens": 69275886.0,
      "step": 30205
    },
    {
      "entropy": 5.123374319076538,
      "epoch": 2.9020172910662825,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041659619987349734,
      "loss": 4.7967,
      "mean_token_accuracy": 0.23295451402664186,
      "num_tokens": 69286649.0,
      "step": 30210
    },
    {
      "entropy": 5.0952033519744875,
      "epoch": 2.9024975984630164,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041656955479360487,
      "loss": 4.6808,
      "mean_token_accuracy": 0.24449730515480042,
      "num_tokens": 69296945.0,
      "step": 30215
    },
    {
      "entropy": 5.127448892593383,
      "epoch": 2.9029779058597502,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004165429064268877,
      "loss": 4.7907,
      "mean_token_accuracy": 0.23770235627889633,
      "num_tokens": 69307338.0,
      "step": 30220
    },
    {
      "entropy": 5.074193954467773,
      "epoch": 2.903458213256484,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004165162547739646,
      "loss": 4.7047,
      "mean_token_accuracy": 0.24482691884040833,
      "num_tokens": 69318406.0,
      "step": 30225
    },
    {
      "entropy": 5.0991894721984865,
      "epoch": 2.903938520653218,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004164895998354542,
      "loss": 4.6789,
      "mean_token_accuracy": 0.24519446939229966,
      "num_tokens": 69328702.0,
      "step": 30230
    },
    {
      "entropy": 5.156052494049073,
      "epoch": 2.904418828049952,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004164629416119755,
      "loss": 4.8126,
      "mean_token_accuracy": 0.23813844472169876,
      "num_tokens": 69340953.0,
      "step": 30235
    },
    {
      "entropy": 5.094689798355103,
      "epoch": 2.9048991354466858,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041643628010414735,
      "loss": 4.7149,
      "mean_token_accuracy": 0.2431272268295288,
      "num_tokens": 69350924.0,
      "step": 30240
    },
    {
      "entropy": 5.181777906417847,
      "epoch": 2.9053794428434196,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041640961531258877,
      "loss": 4.8587,
      "mean_token_accuracy": 0.23435833901166916,
      "num_tokens": 69361100.0,
      "step": 30245
    },
    {
      "entropy": 5.1123366355896,
      "epoch": 2.905859750240154,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004163829472379187,
      "loss": 4.668,
      "mean_token_accuracy": 0.25041979402303693,
      "num_tokens": 69372032.0,
      "step": 30250
    },
    {
      "entropy": 5.017369794845581,
      "epoch": 2.9063400576368874,
      "grad_norm": 0.890625,
      "learning_rate": 0.00041635627588075655,
      "loss": 4.6769,
      "mean_token_accuracy": 0.24664778560400008,
      "num_tokens": 69382828.0,
      "step": 30255
    },
    {
      "entropy": 5.189776515960693,
      "epoch": 2.9068203650336217,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004163296012417213,
      "loss": 4.8022,
      "mean_token_accuracy": 0.2410820558667183,
      "num_tokens": 69393508.0,
      "step": 30260
    },
    {
      "entropy": 5.195446872711182,
      "epoch": 2.907300672430355,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041630292332143245,
      "loss": 4.771,
      "mean_token_accuracy": 0.23789985626935958,
      "num_tokens": 69405278.0,
      "step": 30265
    },
    {
      "entropy": 5.125273323059082,
      "epoch": 2.9077809798270895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004162762421205093,
      "loss": 4.8119,
      "mean_token_accuracy": 0.23395038694143294,
      "num_tokens": 69416617.0,
      "step": 30270
    },
    {
      "entropy": 5.140470170974732,
      "epoch": 2.9082612872238234,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041624955763957134,
      "loss": 4.733,
      "mean_token_accuracy": 0.24274429976940154,
      "num_tokens": 69427667.0,
      "step": 30275
    },
    {
      "entropy": 5.233157777786255,
      "epoch": 2.9087415946205573,
      "grad_norm": 1.0,
      "learning_rate": 0.0004162228698792381,
      "loss": 4.9715,
      "mean_token_accuracy": 0.22310574650764464,
      "num_tokens": 69438280.0,
      "step": 30280
    },
    {
      "entropy": 5.10569748878479,
      "epoch": 2.909221902017291,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041619617884012904,
      "loss": 4.6736,
      "mean_token_accuracy": 0.23609825670719148,
      "num_tokens": 69449042.0,
      "step": 30285
    },
    {
      "entropy": 5.114455890655518,
      "epoch": 2.909702209414025,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004161694845228641,
      "loss": 4.7261,
      "mean_token_accuracy": 0.24518774449825287,
      "num_tokens": 69459710.0,
      "step": 30290
    },
    {
      "entropy": 5.1488258838653564,
      "epoch": 2.910182516810759,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004161427869280628,
      "loss": 4.7455,
      "mean_token_accuracy": 0.24012666195631027,
      "num_tokens": 69472543.0,
      "step": 30295
    },
    {
      "entropy": 5.100553226470947,
      "epoch": 2.910662824207493,
      "grad_norm": 0.890625,
      "learning_rate": 0.00041611608605634517,
      "loss": 4.745,
      "mean_token_accuracy": 0.23889791071414948,
      "num_tokens": 69485197.0,
      "step": 30300
    },
    {
      "entropy": 5.129918384552002,
      "epoch": 2.9111431316042267,
      "grad_norm": 0.953125,
      "learning_rate": 0.000416089381908331,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23055357187986375,
      "num_tokens": 69496287.0,
      "step": 30305
    },
    {
      "entropy": 5.065196561813354,
      "epoch": 2.9116234390009605,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004160626744846404,
      "loss": 4.7257,
      "mean_token_accuracy": 0.23816379755735398,
      "num_tokens": 69508142.0,
      "step": 30310
    },
    {
      "entropy": 5.150148344039917,
      "epoch": 2.9121037463976944,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004160359637858933,
      "loss": 4.7759,
      "mean_token_accuracy": 0.23660601526498795,
      "num_tokens": 69520070.0,
      "step": 30315
    },
    {
      "entropy": 5.164110803604126,
      "epoch": 2.9125840537944283,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041600924981270997,
      "loss": 4.8513,
      "mean_token_accuracy": 0.23883511871099472,
      "num_tokens": 69532448.0,
      "step": 30320
    },
    {
      "entropy": 5.239139890670776,
      "epoch": 2.913064361191162,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041598253256571057,
      "loss": 4.8998,
      "mean_token_accuracy": 0.22251630276441575,
      "num_tokens": 69543074.0,
      "step": 30325
    },
    {
      "entropy": 5.198301553726196,
      "epoch": 2.913544668587896,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004159558120455154,
      "loss": 4.7955,
      "mean_token_accuracy": 0.23338040858507156,
      "num_tokens": 69555078.0,
      "step": 30330
    },
    {
      "entropy": 5.175036954879761,
      "epoch": 2.9140249759846304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004159290882527448,
      "loss": 4.8038,
      "mean_token_accuracy": 0.2378573626279831,
      "num_tokens": 69567654.0,
      "step": 30335
    },
    {
      "entropy": 5.118332815170288,
      "epoch": 2.914505283381364,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004159023611880192,
      "loss": 4.8237,
      "mean_token_accuracy": 0.2358691945672035,
      "num_tokens": 69579024.0,
      "step": 30340
    },
    {
      "entropy": 5.191421318054199,
      "epoch": 2.914985590778098,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004158756308519592,
      "loss": 4.8151,
      "mean_token_accuracy": 0.23813893795013427,
      "num_tokens": 69590793.0,
      "step": 30345
    },
    {
      "entropy": 5.156089210510254,
      "epoch": 2.915465898174832,
      "grad_norm": 1.0,
      "learning_rate": 0.00041584889724518545,
      "loss": 4.7682,
      "mean_token_accuracy": 0.23393660485744477,
      "num_tokens": 69602605.0,
      "step": 30350
    },
    {
      "entropy": 5.1067346096038815,
      "epoch": 2.915946205571566,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041582216036831844,
      "loss": 4.7904,
      "mean_token_accuracy": 0.23654446750879288,
      "num_tokens": 69614523.0,
      "step": 30355
    },
    {
      "entropy": 5.079884958267212,
      "epoch": 2.9164265129683,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000415795420221979,
      "loss": 4.7092,
      "mean_token_accuracy": 0.24600469321012497,
      "num_tokens": 69624551.0,
      "step": 30360
    },
    {
      "entropy": 5.057892465591431,
      "epoch": 2.9169068203650337,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041576867680678803,
      "loss": 4.7032,
      "mean_token_accuracy": 0.24684316664934158,
      "num_tokens": 69636363.0,
      "step": 30365
    },
    {
      "entropy": 5.247738170623779,
      "epoch": 2.9173871277617676,
      "grad_norm": 1.0,
      "learning_rate": 0.0004157419301233664,
      "loss": 4.8571,
      "mean_token_accuracy": 0.23150922507047653,
      "num_tokens": 69647841.0,
      "step": 30370
    },
    {
      "entropy": 5.132164096832275,
      "epoch": 2.9178674351585014,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041571518017233505,
      "loss": 4.7037,
      "mean_token_accuracy": 0.246920146048069,
      "num_tokens": 69657649.0,
      "step": 30375
    },
    {
      "entropy": 5.0434645652771,
      "epoch": 2.9183477425552353,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004156884269543151,
      "loss": 4.6783,
      "mean_token_accuracy": 0.253414548933506,
      "num_tokens": 69668580.0,
      "step": 30380
    },
    {
      "entropy": 5.082406616210937,
      "epoch": 2.918828049951969,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004156616704699275,
      "loss": 4.745,
      "mean_token_accuracy": 0.24331954270601272,
      "num_tokens": 69679258.0,
      "step": 30385
    },
    {
      "entropy": 5.142877388000488,
      "epoch": 2.919308357348703,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041563491071979375,
      "loss": 4.8191,
      "mean_token_accuracy": 0.23628956973552703,
      "num_tokens": 69691119.0,
      "step": 30390
    },
    {
      "entropy": 5.1823649406433105,
      "epoch": 2.919788664745437,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041560814770453495,
      "loss": 4.7776,
      "mean_token_accuracy": 0.24088479727506637,
      "num_tokens": 69702827.0,
      "step": 30395
    },
    {
      "entropy": 5.123574304580688,
      "epoch": 2.920268972142171,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041558138142477235,
      "loss": 4.6891,
      "mean_token_accuracy": 0.23904197961091994,
      "num_tokens": 69715008.0,
      "step": 30400
    },
    {
      "entropy": 5.1079487800598145,
      "epoch": 2.9207492795389047,
      "grad_norm": 1.0,
      "learning_rate": 0.00041555461188112763,
      "loss": 4.74,
      "mean_token_accuracy": 0.2400098517537117,
      "num_tokens": 69725603.0,
      "step": 30405
    },
    {
      "entropy": 5.043749713897705,
      "epoch": 2.921229586935639,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041552783907422217,
      "loss": 4.6998,
      "mean_token_accuracy": 0.24408645033836365,
      "num_tokens": 69735733.0,
      "step": 30410
    },
    {
      "entropy": 5.016913890838623,
      "epoch": 2.9217098943323725,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004155010630046775,
      "loss": 4.7133,
      "mean_token_accuracy": 0.2423143520951271,
      "num_tokens": 69748078.0,
      "step": 30415
    },
    {
      "entropy": 5.060306167602539,
      "epoch": 2.922190201729107,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004154742836731155,
      "loss": 4.73,
      "mean_token_accuracy": 0.24316249042749405,
      "num_tokens": 69759461.0,
      "step": 30420
    },
    {
      "entropy": 5.121638441085816,
      "epoch": 2.9226705091258407,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004154475010801576,
      "loss": 4.7495,
      "mean_token_accuracy": 0.2421607106924057,
      "num_tokens": 69770921.0,
      "step": 30425
    },
    {
      "entropy": 5.141956377029419,
      "epoch": 2.9231508165225746,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041542071522642583,
      "loss": 4.7926,
      "mean_token_accuracy": 0.2359586924314499,
      "num_tokens": 69782399.0,
      "step": 30430
    },
    {
      "entropy": 5.151498460769654,
      "epoch": 2.9236311239193085,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004153939261125421,
      "loss": 4.7542,
      "mean_token_accuracy": 0.24192732572555542,
      "num_tokens": 69792584.0,
      "step": 30435
    },
    {
      "entropy": 5.118769407272339,
      "epoch": 2.9241114313160423,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004153671337391281,
      "loss": 4.7088,
      "mean_token_accuracy": 0.24824296534061432,
      "num_tokens": 69803582.0,
      "step": 30440
    },
    {
      "entropy": 5.172267913818359,
      "epoch": 2.9245917387127762,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004153403381068062,
      "loss": 4.7671,
      "mean_token_accuracy": 0.23994677513837814,
      "num_tokens": 69815635.0,
      "step": 30445
    },
    {
      "entropy": 5.14039454460144,
      "epoch": 2.92507204610951,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041531353921619833,
      "loss": 4.7708,
      "mean_token_accuracy": 0.2352833107113838,
      "num_tokens": 69827260.0,
      "step": 30450
    },
    {
      "entropy": 5.013454151153565,
      "epoch": 2.925552353506244,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004152867370679267,
      "loss": 4.6582,
      "mean_token_accuracy": 0.24340566843748093,
      "num_tokens": 69839050.0,
      "step": 30455
    },
    {
      "entropy": 5.296353101730347,
      "epoch": 2.926032660902978,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041525993166261366,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22365088313817977,
      "num_tokens": 69849543.0,
      "step": 30460
    },
    {
      "entropy": 5.206067943572998,
      "epoch": 2.9265129682997117,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004152331230008814,
      "loss": 4.7833,
      "mean_token_accuracy": 0.23461216688156128,
      "num_tokens": 69861203.0,
      "step": 30465
    },
    {
      "entropy": 5.1721302509307865,
      "epoch": 2.9269932756964456,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041520631108335254,
      "loss": 4.7956,
      "mean_token_accuracy": 0.2305859610438347,
      "num_tokens": 69873357.0,
      "step": 30470
    },
    {
      "entropy": 5.08233528137207,
      "epoch": 2.9274735830931795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004151794959106494,
      "loss": 4.7376,
      "mean_token_accuracy": 0.24346067011356354,
      "num_tokens": 69885078.0,
      "step": 30475
    },
    {
      "entropy": 5.105818128585815,
      "epoch": 2.9279538904899134,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004151526774833947,
      "loss": 4.765,
      "mean_token_accuracy": 0.2341263547539711,
      "num_tokens": 69897076.0,
      "step": 30480
    },
    {
      "entropy": 5.075425004959106,
      "epoch": 2.9284341978866477,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041512585580221086,
      "loss": 4.6405,
      "mean_token_accuracy": 0.25478704571723937,
      "num_tokens": 69907958.0,
      "step": 30485
    },
    {
      "entropy": 5.117607927322387,
      "epoch": 2.928914505283381,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004150990308677208,
      "loss": 4.7697,
      "mean_token_accuracy": 0.2355414628982544,
      "num_tokens": 69918684.0,
      "step": 30490
    },
    {
      "entropy": 5.214997291564941,
      "epoch": 2.9293948126801155,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00041507220268054737,
      "loss": 4.8057,
      "mean_token_accuracy": 0.23425210416316986,
      "num_tokens": 69931458.0,
      "step": 30495
    },
    {
      "entropy": 5.181309843063355,
      "epoch": 2.929875120076849,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004150453712413131,
      "loss": 4.7289,
      "mean_token_accuracy": 0.2369502604007721,
      "num_tokens": 69941837.0,
      "step": 30500
    },
    {
      "entropy": 5.167145919799805,
      "epoch": 2.9303554274735832,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041501853655064134,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23402242511510848,
      "num_tokens": 69954932.0,
      "step": 30505
    },
    {
      "entropy": 5.1008641719818115,
      "epoch": 2.930835734870317,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004149916986091548,
      "loss": 4.7708,
      "mean_token_accuracy": 0.24890220165252686,
      "num_tokens": 69965990.0,
      "step": 30510
    },
    {
      "entropy": 5.1180895328521725,
      "epoch": 2.931316042267051,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004149648574174768,
      "loss": 4.6836,
      "mean_token_accuracy": 0.2433813214302063,
      "num_tokens": 69977603.0,
      "step": 30515
    },
    {
      "entropy": 5.064050149917603,
      "epoch": 2.931796349663785,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004149380129762304,
      "loss": 4.7082,
      "mean_token_accuracy": 0.24364856034517288,
      "num_tokens": 69988043.0,
      "step": 30520
    },
    {
      "entropy": 5.056798315048217,
      "epoch": 2.9322766570605188,
      "grad_norm": 1.0,
      "learning_rate": 0.0004149111652860389,
      "loss": 4.6883,
      "mean_token_accuracy": 0.24188051223754883,
      "num_tokens": 69999397.0,
      "step": 30525
    },
    {
      "entropy": 5.141499090194702,
      "epoch": 2.9327569644572526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004148843143475255,
      "loss": 4.8172,
      "mean_token_accuracy": 0.23510959148406982,
      "num_tokens": 70010673.0,
      "step": 30530
    },
    {
      "entropy": 5.099827527999878,
      "epoch": 2.9332372718539865,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004148574601613137,
      "loss": 4.7181,
      "mean_token_accuracy": 0.24334189742803575,
      "num_tokens": 70022288.0,
      "step": 30535
    },
    {
      "entropy": 5.072341251373291,
      "epoch": 2.9337175792507204,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004148306027280271,
      "loss": 4.7005,
      "mean_token_accuracy": 0.24283528625965117,
      "num_tokens": 70033918.0,
      "step": 30540
    },
    {
      "entropy": 5.032701444625855,
      "epoch": 2.9341978866474543,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041480374204828896,
      "loss": 4.6868,
      "mean_token_accuracy": 0.2474943831562996,
      "num_tokens": 70046072.0,
      "step": 30545
    },
    {
      "entropy": 5.170220136642456,
      "epoch": 2.934678194044188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041477687812272314,
      "loss": 4.7892,
      "mean_token_accuracy": 0.2394439846277237,
      "num_tokens": 70057977.0,
      "step": 30550
    },
    {
      "entropy": 5.210332345962525,
      "epoch": 2.935158501440922,
      "grad_norm": 1.0,
      "learning_rate": 0.00041475001095195324,
      "loss": 4.7607,
      "mean_token_accuracy": 0.2390919953584671,
      "num_tokens": 70069000.0,
      "step": 30555
    },
    {
      "entropy": 5.133387756347656,
      "epoch": 2.9356388088376564,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004147231405366031,
      "loss": 4.7498,
      "mean_token_accuracy": 0.24131525456905364,
      "num_tokens": 70080648.0,
      "step": 30560
    },
    {
      "entropy": 5.06474461555481,
      "epoch": 2.93611911623439,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004146962668772965,
      "loss": 4.7151,
      "mean_token_accuracy": 0.24358271956443786,
      "num_tokens": 70092097.0,
      "step": 30565
    },
    {
      "entropy": 5.0910454273223875,
      "epoch": 2.936599423631124,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041466938997465744,
      "loss": 4.6682,
      "mean_token_accuracy": 0.24275936633348466,
      "num_tokens": 70103665.0,
      "step": 30570
    },
    {
      "entropy": 5.080952835083008,
      "epoch": 2.9370797310278576,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041464250982930974,
      "loss": 4.7036,
      "mean_token_accuracy": 0.24506312757730483,
      "num_tokens": 70115587.0,
      "step": 30575
    },
    {
      "entropy": 5.083723831176758,
      "epoch": 2.937560038424592,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00041461562644187777,
      "loss": 4.7831,
      "mean_token_accuracy": 0.2364642947912216,
      "num_tokens": 70126312.0,
      "step": 30580
    },
    {
      "entropy": 5.050444555282593,
      "epoch": 2.938040345821326,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041458873981298547,
      "loss": 4.6343,
      "mean_token_accuracy": 0.2588354185223579,
      "num_tokens": 70136688.0,
      "step": 30585
    },
    {
      "entropy": 5.108046340942383,
      "epoch": 2.9385206532180597,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041456184994325714,
      "loss": 4.7449,
      "mean_token_accuracy": 0.2375996842980385,
      "num_tokens": 70149169.0,
      "step": 30590
    },
    {
      "entropy": 5.23122353553772,
      "epoch": 2.9390009606147935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041453495683331694,
      "loss": 4.8563,
      "mean_token_accuracy": 0.2290556773543358,
      "num_tokens": 70160246.0,
      "step": 30595
    },
    {
      "entropy": 5.1436155319213865,
      "epoch": 2.9394812680115274,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041450806048378954,
      "loss": 4.754,
      "mean_token_accuracy": 0.23686522096395493,
      "num_tokens": 70171553.0,
      "step": 30600
    },
    {
      "entropy": 5.114828968048096,
      "epoch": 2.9399615754082613,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004144811608952992,
      "loss": 4.7683,
      "mean_token_accuracy": 0.23783933073282243,
      "num_tokens": 70182675.0,
      "step": 30605
    },
    {
      "entropy": 5.139086055755615,
      "epoch": 2.940441882804995,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041445425806847043,
      "loss": 4.7718,
      "mean_token_accuracy": 0.23809271156787873,
      "num_tokens": 70194222.0,
      "step": 30610
    },
    {
      "entropy": 5.095599508285522,
      "epoch": 2.940922190201729,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041442735200392783,
      "loss": 4.656,
      "mean_token_accuracy": 0.2460920736193657,
      "num_tokens": 70204735.0,
      "step": 30615
    },
    {
      "entropy": 5.191938877105713,
      "epoch": 2.941402497598463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004144004427022962,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2300799936056137,
      "num_tokens": 70216368.0,
      "step": 30620
    },
    {
      "entropy": 5.138504409790039,
      "epoch": 2.941882804995197,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041437353016420025,
      "loss": 4.8115,
      "mean_token_accuracy": 0.229746413230896,
      "num_tokens": 70229497.0,
      "step": 30625
    },
    {
      "entropy": 5.129399633407592,
      "epoch": 2.9423631123919307,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004143466143902648,
      "loss": 4.7325,
      "mean_token_accuracy": 0.2351190835237503,
      "num_tokens": 70240826.0,
      "step": 30630
    },
    {
      "entropy": 5.150893259048462,
      "epoch": 2.9428434197886646,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00041431969538111463,
      "loss": 4.8143,
      "mean_token_accuracy": 0.23581030815839768,
      "num_tokens": 70253337.0,
      "step": 30635
    },
    {
      "entropy": 5.125476455688476,
      "epoch": 2.9433237271853985,
      "grad_norm": 1.0,
      "learning_rate": 0.0004142927731373749,
      "loss": 4.6953,
      "mean_token_accuracy": 0.24709575325250627,
      "num_tokens": 70263881.0,
      "step": 30640
    },
    {
      "entropy": 5.109433317184449,
      "epoch": 2.943804034582133,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004142658476596706,
      "loss": 4.7577,
      "mean_token_accuracy": 0.24407728761434555,
      "num_tokens": 70276321.0,
      "step": 30645
    },
    {
      "entropy": 5.061673927307129,
      "epoch": 2.9442843419788662,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041423891894862687,
      "loss": 4.7462,
      "mean_token_accuracy": 0.24099866598844527,
      "num_tokens": 70288778.0,
      "step": 30650
    },
    {
      "entropy": 5.246423673629761,
      "epoch": 2.9447646493756006,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004142119870048689,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2251395657658577,
      "num_tokens": 70300338.0,
      "step": 30655
    },
    {
      "entropy": 5.263963413238526,
      "epoch": 2.9452449567723344,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004141850518290219,
      "loss": 4.8011,
      "mean_token_accuracy": 0.2344187706708908,
      "num_tokens": 70312682.0,
      "step": 30660
    },
    {
      "entropy": 5.168008852005005,
      "epoch": 2.9457252641690683,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041415811342171134,
      "loss": 4.8258,
      "mean_token_accuracy": 0.2391446977853775,
      "num_tokens": 70324408.0,
      "step": 30665
    },
    {
      "entropy": 5.059015607833862,
      "epoch": 2.946205571565802,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004141311717835625,
      "loss": 4.692,
      "mean_token_accuracy": 0.24441739320755004,
      "num_tokens": 70335501.0,
      "step": 30670
    },
    {
      "entropy": 5.090706253051758,
      "epoch": 2.946685878962536,
      "grad_norm": 0.90625,
      "learning_rate": 0.00041410422691520114,
      "loss": 4.6807,
      "mean_token_accuracy": 0.2471085473895073,
      "num_tokens": 70346848.0,
      "step": 30675
    },
    {
      "entropy": 5.16997389793396,
      "epoch": 2.94716618635927,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041407727881725265,
      "loss": 4.7984,
      "mean_token_accuracy": 0.2283072367310524,
      "num_tokens": 70356707.0,
      "step": 30680
    },
    {
      "entropy": 5.16409912109375,
      "epoch": 2.947646493756004,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004140503274903426,
      "loss": 4.8357,
      "mean_token_accuracy": 0.22950833886861802,
      "num_tokens": 70368115.0,
      "step": 30685
    },
    {
      "entropy": 5.168568706512451,
      "epoch": 2.9481268011527377,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000414023372935097,
      "loss": 4.7251,
      "mean_token_accuracy": 0.24447382092475892,
      "num_tokens": 70379883.0,
      "step": 30690
    },
    {
      "entropy": 5.168913507461548,
      "epoch": 2.9486071085494716,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00041399641515214137,
      "loss": 4.7673,
      "mean_token_accuracy": 0.23434825390577316,
      "num_tokens": 70390833.0,
      "step": 30695
    },
    {
      "entropy": 5.1006187915802,
      "epoch": 2.9490874159462055,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004139694541421018,
      "loss": 4.8049,
      "mean_token_accuracy": 0.23353264182806016,
      "num_tokens": 70402679.0,
      "step": 30700
    },
    {
      "entropy": 5.158392667770386,
      "epoch": 2.9495677233429394,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004139424899056042,
      "loss": 4.8276,
      "mean_token_accuracy": 0.22481898814439774,
      "num_tokens": 70415623.0,
      "step": 30705
    },
    {
      "entropy": 5.211047887802124,
      "epoch": 2.9500480307396733,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00041391552244327446,
      "loss": 4.8603,
      "mean_token_accuracy": 0.23337038606405258,
      "num_tokens": 70427192.0,
      "step": 30710
    },
    {
      "entropy": 5.135737800598145,
      "epoch": 2.950528338136407,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004138885517557387,
      "loss": 4.7652,
      "mean_token_accuracy": 0.242301145195961,
      "num_tokens": 70439002.0,
      "step": 30715
    },
    {
      "entropy": 5.109500360488892,
      "epoch": 2.9510086455331415,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004138615778436234,
      "loss": 4.7199,
      "mean_token_accuracy": 0.24121250063180924,
      "num_tokens": 70449790.0,
      "step": 30720
    },
    {
      "entropy": 5.176414585113525,
      "epoch": 2.951488952929875,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041383460070755447,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23183022141456605,
      "num_tokens": 70460687.0,
      "step": 30725
    },
    {
      "entropy": 5.100621318817138,
      "epoch": 2.9519692603266092,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041380762034815834,
      "loss": 4.666,
      "mean_token_accuracy": 0.2436969041824341,
      "num_tokens": 70470851.0,
      "step": 30730
    },
    {
      "entropy": 5.097142457962036,
      "epoch": 2.952449567723343,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041378063676606147,
      "loss": 4.7551,
      "mean_token_accuracy": 0.24399888068437575,
      "num_tokens": 70481979.0,
      "step": 30735
    },
    {
      "entropy": 5.128300523757934,
      "epoch": 2.952929875120077,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041375364996189035,
      "loss": 4.705,
      "mean_token_accuracy": 0.23814308345317842,
      "num_tokens": 70493107.0,
      "step": 30740
    },
    {
      "entropy": 5.026430702209472,
      "epoch": 2.953410182516811,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041372665993627143,
      "loss": 4.6851,
      "mean_token_accuracy": 0.2405782699584961,
      "num_tokens": 70504385.0,
      "step": 30745
    },
    {
      "entropy": 5.074408102035522,
      "epoch": 2.9538904899135447,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041369966668983144,
      "loss": 4.7074,
      "mean_token_accuracy": 0.24341859966516494,
      "num_tokens": 70515906.0,
      "step": 30750
    },
    {
      "entropy": 5.217910623550415,
      "epoch": 2.9543707973102786,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041367267022319706,
      "loss": 4.8435,
      "mean_token_accuracy": 0.23252399265766144,
      "num_tokens": 70526927.0,
      "step": 30755
    },
    {
      "entropy": 5.202370357513428,
      "epoch": 2.9548511047070125,
      "grad_norm": 0.953125,
      "learning_rate": 0.000413645670536995,
      "loss": 4.7996,
      "mean_token_accuracy": 0.23483735471963882,
      "num_tokens": 70537521.0,
      "step": 30760
    },
    {
      "entropy": 5.2233587265014645,
      "epoch": 2.9553314121037464,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004136186676318522,
      "loss": 4.7979,
      "mean_token_accuracy": 0.237154957652092,
      "num_tokens": 70547851.0,
      "step": 30765
    },
    {
      "entropy": 5.077259111404419,
      "epoch": 2.9558117195004803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004135916615083956,
      "loss": 4.7064,
      "mean_token_accuracy": 0.2428145721554756,
      "num_tokens": 70559725.0,
      "step": 30770
    },
    {
      "entropy": 5.086357641220093,
      "epoch": 2.956292026897214,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041356465216725195,
      "loss": 4.7214,
      "mean_token_accuracy": 0.246979658305645,
      "num_tokens": 70571410.0,
      "step": 30775
    },
    {
      "entropy": 5.157089805603027,
      "epoch": 2.956772334293948,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041353763960904873,
      "loss": 4.7997,
      "mean_token_accuracy": 0.23580764532089232,
      "num_tokens": 70582665.0,
      "step": 30780
    },
    {
      "entropy": 5.107215929031372,
      "epoch": 2.957252641690682,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041351062383441286,
      "loss": 4.765,
      "mean_token_accuracy": 0.2469639837741852,
      "num_tokens": 70593653.0,
      "step": 30785
    },
    {
      "entropy": 5.1765697479248045,
      "epoch": 2.957732949087416,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004134836048439716,
      "loss": 4.8274,
      "mean_token_accuracy": 0.23656340539455414,
      "num_tokens": 70603706.0,
      "step": 30790
    },
    {
      "entropy": 5.129392862319946,
      "epoch": 2.95821325648415,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041345658263835215,
      "loss": 4.7116,
      "mean_token_accuracy": 0.23703265339136123,
      "num_tokens": 70614940.0,
      "step": 30795
    },
    {
      "entropy": 5.2329682350158695,
      "epoch": 2.9586935638808836,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041342955721818207,
      "loss": 4.8456,
      "mean_token_accuracy": 0.2344607576727867,
      "num_tokens": 70626503.0,
      "step": 30800
    },
    {
      "entropy": 5.090293836593628,
      "epoch": 2.959173871277618,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041340252858408866,
      "loss": 4.7324,
      "mean_token_accuracy": 0.24376551806926727,
      "num_tokens": 70638483.0,
      "step": 30805
    },
    {
      "entropy": 5.064789152145385,
      "epoch": 2.9596541786743513,
      "grad_norm": 0.890625,
      "learning_rate": 0.00041337549673669963,
      "loss": 4.7378,
      "mean_token_accuracy": 0.24274600446224212,
      "num_tokens": 70650609.0,
      "step": 30810
    },
    {
      "entropy": 5.108625173568726,
      "epoch": 2.9601344860710856,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004133484616766423,
      "loss": 4.7521,
      "mean_token_accuracy": 0.2436141088604927,
      "num_tokens": 70661368.0,
      "step": 30815
    },
    {
      "entropy": 5.2180260181427,
      "epoch": 2.9606147934678195,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041332142340454463,
      "loss": 4.7653,
      "mean_token_accuracy": 0.23971326351165773,
      "num_tokens": 70673324.0,
      "step": 30820
    },
    {
      "entropy": 5.116898155212402,
      "epoch": 2.9610951008645534,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004132943819210342,
      "loss": 4.7381,
      "mean_token_accuracy": 0.24599200934171678,
      "num_tokens": 70684639.0,
      "step": 30825
    },
    {
      "entropy": 5.034547901153564,
      "epoch": 2.9615754082612873,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041326733722673876,
      "loss": 4.6471,
      "mean_token_accuracy": 0.24587527215480803,
      "num_tokens": 70695742.0,
      "step": 30830
    },
    {
      "entropy": 5.156430196762085,
      "epoch": 2.962055715658021,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041324028932228645,
      "loss": 4.7857,
      "mean_token_accuracy": 0.23438030034303664,
      "num_tokens": 70706711.0,
      "step": 30835
    },
    {
      "entropy": 5.120493698120117,
      "epoch": 2.962536023054755,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004132132382083051,
      "loss": 4.7537,
      "mean_token_accuracy": 0.24160445779561995,
      "num_tokens": 70719090.0,
      "step": 30840
    },
    {
      "entropy": 5.100801038742065,
      "epoch": 2.963016330451489,
      "grad_norm": 1.0,
      "learning_rate": 0.00041318618388542274,
      "loss": 4.7956,
      "mean_token_accuracy": 0.23661700040102004,
      "num_tokens": 70730356.0,
      "step": 30845
    },
    {
      "entropy": 5.215507078170776,
      "epoch": 2.963496637848223,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004131591263542675,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2310863956809044,
      "num_tokens": 70744864.0,
      "step": 30850
    },
    {
      "entropy": 5.222301864624024,
      "epoch": 2.9639769452449567,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004131320656154676,
      "loss": 4.8648,
      "mean_token_accuracy": 0.23220660388469697,
      "num_tokens": 70756882.0,
      "step": 30855
    },
    {
      "entropy": 5.123324346542359,
      "epoch": 2.9644572526416906,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004131050016696514,
      "loss": 4.7258,
      "mean_token_accuracy": 0.2395365133881569,
      "num_tokens": 70768501.0,
      "step": 30860
    },
    {
      "entropy": 5.109246301651001,
      "epoch": 2.9649375600384245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000413077934517447,
      "loss": 4.7152,
      "mean_token_accuracy": 0.24292102009057998,
      "num_tokens": 70779545.0,
      "step": 30865
    },
    {
      "entropy": 5.064622259140014,
      "epoch": 2.9654178674351583,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004130508641594831,
      "loss": 4.7617,
      "mean_token_accuracy": 0.23982039839029312,
      "num_tokens": 70791423.0,
      "step": 30870
    },
    {
      "entropy": 5.0940018653869625,
      "epoch": 2.965898174831892,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041302379059638794,
      "loss": 4.7533,
      "mean_token_accuracy": 0.2399999365210533,
      "num_tokens": 70802938.0,
      "step": 30875
    },
    {
      "entropy": 5.196796321868897,
      "epoch": 2.9663784822286265,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041299671382879024,
      "loss": 4.7931,
      "mean_token_accuracy": 0.23076225072145462,
      "num_tokens": 70813285.0,
      "step": 30880
    },
    {
      "entropy": 5.080271053314209,
      "epoch": 2.96685878962536,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004129696338573187,
      "loss": 4.7502,
      "mean_token_accuracy": 0.23470867276191712,
      "num_tokens": 70826509.0,
      "step": 30885
    },
    {
      "entropy": 5.073721837997437,
      "epoch": 2.9673390970220943,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004129425506826019,
      "loss": 4.7318,
      "mean_token_accuracy": 0.24063251912593842,
      "num_tokens": 70837648.0,
      "step": 30890
    },
    {
      "entropy": 5.101913547515869,
      "epoch": 2.967819404418828,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041291546430526863,
      "loss": 4.7338,
      "mean_token_accuracy": 0.24371240437030792,
      "num_tokens": 70849702.0,
      "step": 30895
    },
    {
      "entropy": 5.16077880859375,
      "epoch": 2.968299711815562,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004128883747259478,
      "loss": 4.819,
      "mean_token_accuracy": 0.2392261803150177,
      "num_tokens": 70861445.0,
      "step": 30900
    },
    {
      "entropy": 5.180429172515869,
      "epoch": 2.968780019212296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004128612819452684,
      "loss": 4.7467,
      "mean_token_accuracy": 0.240490160882473,
      "num_tokens": 70872448.0,
      "step": 30905
    },
    {
      "entropy": 5.122031593322754,
      "epoch": 2.96926032660903,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041283418596385944,
      "loss": 4.652,
      "mean_token_accuracy": 0.24544265866279602,
      "num_tokens": 70883777.0,
      "step": 30910
    },
    {
      "entropy": 5.054535913467407,
      "epoch": 2.9697406340057637,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004128070867823499,
      "loss": 4.7317,
      "mean_token_accuracy": 0.2429581344127655,
      "num_tokens": 70895084.0,
      "step": 30915
    },
    {
      "entropy": 5.1057116985321045,
      "epoch": 2.9702209414024976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000412779984401369,
      "loss": 4.7487,
      "mean_token_accuracy": 0.24346766024827957,
      "num_tokens": 70906304.0,
      "step": 30920
    },
    {
      "entropy": 5.155746030807495,
      "epoch": 2.9707012487992315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000412752878821546,
      "loss": 4.6832,
      "mean_token_accuracy": 0.243827947974205,
      "num_tokens": 70917448.0,
      "step": 30925
    },
    {
      "entropy": 5.0430872440338135,
      "epoch": 2.9711815561959654,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041272577004351026,
      "loss": 4.6791,
      "mean_token_accuracy": 0.2479093000292778,
      "num_tokens": 70927991.0,
      "step": 30930
    },
    {
      "entropy": 5.0626856803894045,
      "epoch": 2.9716618635926992,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041269865806789095,
      "loss": 4.7048,
      "mean_token_accuracy": 0.24032177329063414,
      "num_tokens": 70939490.0,
      "step": 30935
    },
    {
      "entropy": 5.145792627334595,
      "epoch": 2.972142170989433,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004126715428953178,
      "loss": 4.7436,
      "mean_token_accuracy": 0.23982749730348588,
      "num_tokens": 70951611.0,
      "step": 30940
    },
    {
      "entropy": 5.143175649642944,
      "epoch": 2.972622478386167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004126444245264202,
      "loss": 4.7867,
      "mean_token_accuracy": 0.23149303793907167,
      "num_tokens": 70962003.0,
      "step": 30945
    },
    {
      "entropy": 5.0007306098937985,
      "epoch": 2.973102785782901,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004126173029618278,
      "loss": 4.5844,
      "mean_token_accuracy": 0.25529240518808366,
      "num_tokens": 70972022.0,
      "step": 30950
    },
    {
      "entropy": 5.105014801025391,
      "epoch": 2.973583093179635,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004125901782021702,
      "loss": 4.7972,
      "mean_token_accuracy": 0.23774835765361785,
      "num_tokens": 70984086.0,
      "step": 30955
    },
    {
      "entropy": 5.226615953445434,
      "epoch": 2.9740634005763686,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004125630502480773,
      "loss": 4.8543,
      "mean_token_accuracy": 0.22880603671073912,
      "num_tokens": 70995705.0,
      "step": 30960
    },
    {
      "entropy": 5.193907976150513,
      "epoch": 2.974543707973103,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004125359191001788,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23378911465406418,
      "num_tokens": 71007812.0,
      "step": 30965
    },
    {
      "entropy": 5.1122174739837645,
      "epoch": 2.975024015369837,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004125087847591047,
      "loss": 4.7276,
      "mean_token_accuracy": 0.24071373492479325,
      "num_tokens": 71018950.0,
      "step": 30970
    },
    {
      "entropy": 5.09380316734314,
      "epoch": 2.9755043227665707,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041248164722548493,
      "loss": 4.7187,
      "mean_token_accuracy": 0.23919163793325424,
      "num_tokens": 71030391.0,
      "step": 30975
    },
    {
      "entropy": 5.139894485473633,
      "epoch": 2.9759846301633046,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004124545064999495,
      "loss": 4.7407,
      "mean_token_accuracy": 0.23884033858776094,
      "num_tokens": 71041741.0,
      "step": 30980
    },
    {
      "entropy": 5.147083330154419,
      "epoch": 2.9764649375600385,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041242736258312866,
      "loss": 4.7431,
      "mean_token_accuracy": 0.24201656132936478,
      "num_tokens": 71052062.0,
      "step": 30985
    },
    {
      "entropy": 5.112147092819214,
      "epoch": 2.9769452449567724,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004124002154756525,
      "loss": 4.7505,
      "mean_token_accuracy": 0.2399471491575241,
      "num_tokens": 71063608.0,
      "step": 30990
    },
    {
      "entropy": 5.16088171005249,
      "epoch": 2.9774255523535063,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041237306517815124,
      "loss": 4.7849,
      "mean_token_accuracy": 0.2397480070590973,
      "num_tokens": 71073526.0,
      "step": 30995
    },
    {
      "entropy": 5.015842723846435,
      "epoch": 2.97790585975024,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004123459116912554,
      "loss": 4.6514,
      "mean_token_accuracy": 0.2504919707775116,
      "num_tokens": 71085311.0,
      "step": 31000
    },
    {
      "entropy": 5.142609262466431,
      "epoch": 2.978386167146974,
      "grad_norm": 0.890625,
      "learning_rate": 0.00041231875501559535,
      "loss": 4.7963,
      "mean_token_accuracy": 0.23674438744783402,
      "num_tokens": 71096438.0,
      "step": 31005
    },
    {
      "entropy": 5.1144371509552,
      "epoch": 2.978866474543708,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041229159515180155,
      "loss": 4.7176,
      "mean_token_accuracy": 0.2395267218351364,
      "num_tokens": 71108262.0,
      "step": 31010
    },
    {
      "entropy": 5.061624670028687,
      "epoch": 2.979346781940442,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004122644321005046,
      "loss": 4.748,
      "mean_token_accuracy": 0.23395977765321732,
      "num_tokens": 71120905.0,
      "step": 31015
    },
    {
      "entropy": 5.096264886856079,
      "epoch": 2.9798270893371757,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041223726586233505,
      "loss": 4.7843,
      "mean_token_accuracy": 0.24699696600437165,
      "num_tokens": 71132522.0,
      "step": 31020
    },
    {
      "entropy": 5.113017272949219,
      "epoch": 2.9803073967339095,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041221009643792377,
      "loss": 4.7154,
      "mean_token_accuracy": 0.25052625983953475,
      "num_tokens": 71143406.0,
      "step": 31025
    },
    {
      "entropy": 5.157391119003296,
      "epoch": 2.980787704130644,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004121829238279014,
      "loss": 4.7881,
      "mean_token_accuracy": 0.23625537902116775,
      "num_tokens": 71153525.0,
      "step": 31030
    },
    {
      "entropy": 5.1379858493804935,
      "epoch": 2.9812680115273773,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041215574803289896,
      "loss": 4.746,
      "mean_token_accuracy": 0.23221674859523772,
      "num_tokens": 71165672.0,
      "step": 31035
    },
    {
      "entropy": 5.09129147529602,
      "epoch": 2.9817483189241116,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004121285690535473,
      "loss": 4.7236,
      "mean_token_accuracy": 0.24247112572193147,
      "num_tokens": 71176411.0,
      "step": 31040
    },
    {
      "entropy": 5.152729463577271,
      "epoch": 2.982228626320845,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041210138689047745,
      "loss": 4.7538,
      "mean_token_accuracy": 0.23558194190263748,
      "num_tokens": 71189520.0,
      "step": 31045
    },
    {
      "entropy": 5.1336071491241455,
      "epoch": 2.9827089337175794,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004120742015443206,
      "loss": 4.7143,
      "mean_token_accuracy": 0.24685877114534377,
      "num_tokens": 71200199.0,
      "step": 31050
    },
    {
      "entropy": 5.1885498046875,
      "epoch": 2.9831892411143133,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004120470130157077,
      "loss": 4.804,
      "mean_token_accuracy": 0.2288921892642975,
      "num_tokens": 71211909.0,
      "step": 31055
    },
    {
      "entropy": 5.077307271957397,
      "epoch": 2.983669548511047,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041201982130527006,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24276573061943055,
      "num_tokens": 71222202.0,
      "step": 31060
    },
    {
      "entropy": 5.19982500076294,
      "epoch": 2.984149855907781,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041199262641363914,
      "loss": 4.8644,
      "mean_token_accuracy": 0.23000875413417815,
      "num_tokens": 71233681.0,
      "step": 31065
    },
    {
      "entropy": 5.18394513130188,
      "epoch": 2.984630163304515,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041196542834144617,
      "loss": 4.8052,
      "mean_token_accuracy": 0.24224952459335328,
      "num_tokens": 71244566.0,
      "step": 31070
    },
    {
      "entropy": 5.184459400177002,
      "epoch": 2.985110470701249,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041193822708932265,
      "loss": 4.7647,
      "mean_token_accuracy": 0.23872831016778945,
      "num_tokens": 71254892.0,
      "step": 31075
    },
    {
      "entropy": 5.071117544174195,
      "epoch": 2.9855907780979827,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004119110226579002,
      "loss": 4.7077,
      "mean_token_accuracy": 0.24190827012062072,
      "num_tokens": 71266952.0,
      "step": 31080
    },
    {
      "entropy": 5.103740310668945,
      "epoch": 2.9860710854947166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041188381504781026,
      "loss": 4.6776,
      "mean_token_accuracy": 0.23919121026992798,
      "num_tokens": 71278039.0,
      "step": 31085
    },
    {
      "entropy": 5.165815114974976,
      "epoch": 2.9865513928914504,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004118566042596846,
      "loss": 4.7619,
      "mean_token_accuracy": 0.23757737129926682,
      "num_tokens": 71289201.0,
      "step": 31090
    },
    {
      "entropy": 5.114630031585693,
      "epoch": 2.9870317002881843,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000411829390294155,
      "loss": 4.6864,
      "mean_token_accuracy": 0.2538988828659058,
      "num_tokens": 71301360.0,
      "step": 31095
    },
    {
      "entropy": 5.15974702835083,
      "epoch": 2.987512007684918,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041180217315185333,
      "loss": 4.7831,
      "mean_token_accuracy": 0.23813520073890687,
      "num_tokens": 71311846.0,
      "step": 31100
    },
    {
      "entropy": 5.155130910873413,
      "epoch": 2.9879923150816525,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041177495283341124,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2397029623389244,
      "num_tokens": 71324420.0,
      "step": 31105
    },
    {
      "entropy": 5.085882759094238,
      "epoch": 2.988472622478386,
      "grad_norm": 1.0625,
      "learning_rate": 0.000411747729339461,
      "loss": 4.6112,
      "mean_token_accuracy": 0.2497408375144005,
      "num_tokens": 71335641.0,
      "step": 31110
    },
    {
      "entropy": 5.0239049911499025,
      "epoch": 2.9889529298751203,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004117205026706345,
      "loss": 4.6979,
      "mean_token_accuracy": 0.2514405757188797,
      "num_tokens": 71346360.0,
      "step": 31115
    },
    {
      "entropy": 5.112328243255615,
      "epoch": 2.9894332372718537,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041169327282756396,
      "loss": 4.6776,
      "mean_token_accuracy": 0.2522357404232025,
      "num_tokens": 71357363.0,
      "step": 31120
    },
    {
      "entropy": 5.089079999923706,
      "epoch": 2.989913544668588,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004116660398108815,
      "loss": 4.6363,
      "mean_token_accuracy": 0.24447752088308333,
      "num_tokens": 71368731.0,
      "step": 31125
    },
    {
      "entropy": 5.059494829177856,
      "epoch": 2.990393852065322,
      "grad_norm": 1.0,
      "learning_rate": 0.0004116388036212194,
      "loss": 4.7163,
      "mean_token_accuracy": 0.24206260740756988,
      "num_tokens": 71379891.0,
      "step": 31130
    },
    {
      "entropy": 5.09461612701416,
      "epoch": 2.990874159462056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041161156425921004,
      "loss": 4.6441,
      "mean_token_accuracy": 0.24818260669708253,
      "num_tokens": 71391353.0,
      "step": 31135
    },
    {
      "entropy": 5.088548564910889,
      "epoch": 2.9913544668587897,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041158432172548577,
      "loss": 4.6897,
      "mean_token_accuracy": 0.242967589199543,
      "num_tokens": 71401968.0,
      "step": 31140
    },
    {
      "entropy": 5.070551347732544,
      "epoch": 2.9918347742555236,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041155707602067923,
      "loss": 4.7297,
      "mean_token_accuracy": 0.24480826854705812,
      "num_tokens": 71412056.0,
      "step": 31145
    },
    {
      "entropy": 5.091709089279175,
      "epoch": 2.9923150816522575,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004115298271454227,
      "loss": 4.7233,
      "mean_token_accuracy": 0.24261091500520707,
      "num_tokens": 71422182.0,
      "step": 31150
    },
    {
      "entropy": 5.132114028930664,
      "epoch": 2.9927953890489913,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004115025751003491,
      "loss": 4.7812,
      "mean_token_accuracy": 0.23689234405755996,
      "num_tokens": 71433796.0,
      "step": 31155
    },
    {
      "entropy": 5.166507720947266,
      "epoch": 2.993275696445725,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000411475319886091,
      "loss": 4.7644,
      "mean_token_accuracy": 0.24157465547323226,
      "num_tokens": 71445756.0,
      "step": 31160
    },
    {
      "entropy": 5.197844505310059,
      "epoch": 2.993756003842459,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041144806150328117,
      "loss": 4.8007,
      "mean_token_accuracy": 0.2418261557817459,
      "num_tokens": 71457124.0,
      "step": 31165
    },
    {
      "entropy": 5.080309152603149,
      "epoch": 2.994236311239193,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004114207999525527,
      "loss": 4.6967,
      "mean_token_accuracy": 0.24710508137941362,
      "num_tokens": 71469287.0,
      "step": 31170
    },
    {
      "entropy": 5.088234376907349,
      "epoch": 2.994716618635927,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041139353523453814,
      "loss": 4.6753,
      "mean_token_accuracy": 0.24315319657325746,
      "num_tokens": 71480403.0,
      "step": 31175
    },
    {
      "entropy": 5.1567706108093265,
      "epoch": 2.9951969260326607,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004113662673498708,
      "loss": 4.7692,
      "mean_token_accuracy": 0.24663938134908675,
      "num_tokens": 71492431.0,
      "step": 31180
    },
    {
      "entropy": 5.106464099884033,
      "epoch": 2.9956772334293946,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041133899629918364,
      "loss": 4.687,
      "mean_token_accuracy": 0.2437478110194206,
      "num_tokens": 71503588.0,
      "step": 31185
    },
    {
      "entropy": 5.014996719360352,
      "epoch": 2.996157540826129,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041131172208310986,
      "loss": 4.6573,
      "mean_token_accuracy": 0.24780642986297607,
      "num_tokens": 71514813.0,
      "step": 31190
    },
    {
      "entropy": 5.117214584350586,
      "epoch": 2.9966378482228624,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041128444470228253,
      "loss": 4.7197,
      "mean_token_accuracy": 0.2354291409254074,
      "num_tokens": 71526961.0,
      "step": 31195
    },
    {
      "entropy": 5.062591934204102,
      "epoch": 2.9971181556195967,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041125716415733524,
      "loss": 4.6491,
      "mean_token_accuracy": 0.25326710641384126,
      "num_tokens": 71537793.0,
      "step": 31200
    },
    {
      "entropy": 5.082094287872314,
      "epoch": 2.9975984630163306,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004112298804489011,
      "loss": 4.7296,
      "mean_token_accuracy": 0.2395835652947426,
      "num_tokens": 71549959.0,
      "step": 31205
    },
    {
      "entropy": 5.159132289886474,
      "epoch": 2.9980787704130645,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004112025935776137,
      "loss": 4.8495,
      "mean_token_accuracy": 0.2326044112443924,
      "num_tokens": 71560340.0,
      "step": 31210
    },
    {
      "entropy": 5.146960163116455,
      "epoch": 2.9985590778097984,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041117530354410647,
      "loss": 4.7869,
      "mean_token_accuracy": 0.2368721529841423,
      "num_tokens": 71571788.0,
      "step": 31215
    },
    {
      "entropy": 5.179553127288818,
      "epoch": 2.9990393852065322,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004111480103490131,
      "loss": 4.7701,
      "mean_token_accuracy": 0.24075673073530196,
      "num_tokens": 71582660.0,
      "step": 31220
    },
    {
      "entropy": 5.081602144241333,
      "epoch": 2.999519692603266,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041112071399296724,
      "loss": 4.7509,
      "mean_token_accuracy": 0.23954785764217376,
      "num_tokens": 71594672.0,
      "step": 31225
    },
    {
      "entropy": 5.121292018890381,
      "epoch": 3.0,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004110934144766025,
      "loss": 4.7702,
      "mean_token_accuracy": 0.23730227947235108,
      "num_tokens": 71605608.0,
      "step": 31230
    },
    {
      "entropy": 5.117084884643555,
      "epoch": 3.000480307396734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041106611180055284,
      "loss": 4.6843,
      "mean_token_accuracy": 0.2424149125814438,
      "num_tokens": 71616917.0,
      "step": 31235
    },
    {
      "entropy": 5.128552865982056,
      "epoch": 3.0009606147934678,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041103880596545206,
      "loss": 4.6995,
      "mean_token_accuracy": 0.24727243036031724,
      "num_tokens": 71630190.0,
      "step": 31240
    },
    {
      "entropy": 5.112913036346436,
      "epoch": 3.0014409221902016,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004110114969719342,
      "loss": 4.684,
      "mean_token_accuracy": 0.24920916557312012,
      "num_tokens": 71642305.0,
      "step": 31245
    },
    {
      "entropy": 5.087633514404297,
      "epoch": 3.0019212295869355,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004109841848206333,
      "loss": 4.6915,
      "mean_token_accuracy": 0.24150362461805344,
      "num_tokens": 71652735.0,
      "step": 31250
    },
    {
      "entropy": 5.0639690399169925,
      "epoch": 3.0024015369836694,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004109568695121833,
      "loss": 4.5806,
      "mean_token_accuracy": 0.2544976338744164,
      "num_tokens": 71664051.0,
      "step": 31255
    },
    {
      "entropy": 5.160264587402343,
      "epoch": 3.0028818443804033,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004109295510472186,
      "loss": 4.6861,
      "mean_token_accuracy": 0.2436651647090912,
      "num_tokens": 71675785.0,
      "step": 31260
    },
    {
      "entropy": 5.095330476760864,
      "epoch": 3.003362151777137,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041090222942637323,
      "loss": 4.6163,
      "mean_token_accuracy": 0.254238648712635,
      "num_tokens": 71687322.0,
      "step": 31265
    },
    {
      "entropy": 5.097984170913696,
      "epoch": 3.0038424591738715,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041087490465028175,
      "loss": 4.7078,
      "mean_token_accuracy": 0.2398919329047203,
      "num_tokens": 71698269.0,
      "step": 31270
    },
    {
      "entropy": 5.143783235549927,
      "epoch": 3.0043227665706054,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041084757671957844,
      "loss": 4.7296,
      "mean_token_accuracy": 0.23778907507658004,
      "num_tokens": 71710078.0,
      "step": 31275
    },
    {
      "entropy": 5.1690065383911135,
      "epoch": 3.0048030739673393,
      "grad_norm": 1.0,
      "learning_rate": 0.00041082024563489773,
      "loss": 4.7751,
      "mean_token_accuracy": 0.24432914555072785,
      "num_tokens": 71721882.0,
      "step": 31280
    },
    {
      "entropy": 5.1613311767578125,
      "epoch": 3.005283381364073,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004107929113968743,
      "loss": 4.7057,
      "mean_token_accuracy": 0.24474416971206664,
      "num_tokens": 71733710.0,
      "step": 31285
    },
    {
      "entropy": 5.08675594329834,
      "epoch": 3.005763688760807,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004107655740061427,
      "loss": 4.6274,
      "mean_token_accuracy": 0.24499188959598542,
      "num_tokens": 71745294.0,
      "step": 31290
    },
    {
      "entropy": 5.094456768035888,
      "epoch": 3.006243996157541,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004107382334633375,
      "loss": 4.6722,
      "mean_token_accuracy": 0.25274425446987153,
      "num_tokens": 71757184.0,
      "step": 31295
    },
    {
      "entropy": 5.1353381156921385,
      "epoch": 3.006724303554275,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004107108897690936,
      "loss": 4.708,
      "mean_token_accuracy": 0.24032434076070786,
      "num_tokens": 71768618.0,
      "step": 31300
    },
    {
      "entropy": 5.19363112449646,
      "epoch": 3.0072046109510087,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004106835429240458,
      "loss": 4.8016,
      "mean_token_accuracy": 0.24347516149282455,
      "num_tokens": 71779741.0,
      "step": 31305
    },
    {
      "entropy": 5.06469841003418,
      "epoch": 3.0076849183477425,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004106561929288292,
      "loss": 4.5884,
      "mean_token_accuracy": 0.2459713786840439,
      "num_tokens": 71791085.0,
      "step": 31310
    },
    {
      "entropy": 5.12068452835083,
      "epoch": 3.0081652257444764,
      "grad_norm": 1.0,
      "learning_rate": 0.00041062883978407844,
      "loss": 4.6716,
      "mean_token_accuracy": 0.24647442996501923,
      "num_tokens": 71804061.0,
      "step": 31315
    },
    {
      "entropy": 5.21141619682312,
      "epoch": 3.0086455331412103,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041060148349042876,
      "loss": 4.7264,
      "mean_token_accuracy": 0.24136997312307357,
      "num_tokens": 71815303.0,
      "step": 31320
    },
    {
      "entropy": 5.169840145111084,
      "epoch": 3.009125840537944,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041057412404851536,
      "loss": 4.7875,
      "mean_token_accuracy": 0.2338212937116623,
      "num_tokens": 71826492.0,
      "step": 31325
    },
    {
      "entropy": 5.118964385986328,
      "epoch": 3.009606147934678,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004105467614589734,
      "loss": 4.6802,
      "mean_token_accuracy": 0.2455555707216263,
      "num_tokens": 71837245.0,
      "step": 31330
    },
    {
      "entropy": 5.145743465423584,
      "epoch": 3.010086455331412,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004105193957224381,
      "loss": 4.6835,
      "mean_token_accuracy": 0.24131604433059692,
      "num_tokens": 71848684.0,
      "step": 31335
    },
    {
      "entropy": 5.159944725036621,
      "epoch": 3.010566762728146,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041049202683954473,
      "loss": 4.6914,
      "mean_token_accuracy": 0.2517484247684479,
      "num_tokens": 71861472.0,
      "step": 31340
    },
    {
      "entropy": 5.079114723205566,
      "epoch": 3.0110470701248797,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041046465481092893,
      "loss": 4.6579,
      "mean_token_accuracy": 0.24665774703025817,
      "num_tokens": 71873076.0,
      "step": 31345
    },
    {
      "entropy": 5.0809645652771,
      "epoch": 3.011527377521614,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041043727963722607,
      "loss": 4.674,
      "mean_token_accuracy": 0.24236734956502914,
      "num_tokens": 71884416.0,
      "step": 31350
    },
    {
      "entropy": 5.121291446685791,
      "epoch": 3.012007684918348,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004104099013190718,
      "loss": 4.6729,
      "mean_token_accuracy": 0.24722956120967865,
      "num_tokens": 71894113.0,
      "step": 31355
    },
    {
      "entropy": 5.140651416778565,
      "epoch": 3.012487992315082,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041038251985710164,
      "loss": 4.711,
      "mean_token_accuracy": 0.2393583133816719,
      "num_tokens": 71904732.0,
      "step": 31360
    },
    {
      "entropy": 5.084239864349366,
      "epoch": 3.0129682997118157,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004103551352519514,
      "loss": 4.6712,
      "mean_token_accuracy": 0.24825433045625686,
      "num_tokens": 71915112.0,
      "step": 31365
    },
    {
      "entropy": 5.078463220596314,
      "epoch": 3.0134486071085496,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041032774750425683,
      "loss": 4.6429,
      "mean_token_accuracy": 0.24332420229911805,
      "num_tokens": 71927175.0,
      "step": 31370
    },
    {
      "entropy": 5.20628080368042,
      "epoch": 3.0139289145052834,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004103003566146538,
      "loss": 4.7551,
      "mean_token_accuracy": 0.2336386740207672,
      "num_tokens": 71937596.0,
      "step": 31375
    },
    {
      "entropy": 5.06389856338501,
      "epoch": 3.0144092219020173,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004102729625837782,
      "loss": 4.602,
      "mean_token_accuracy": 0.2576372519135475,
      "num_tokens": 71948594.0,
      "step": 31380
    },
    {
      "entropy": 5.165049934387207,
      "epoch": 3.014889529298751,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004102455654122662,
      "loss": 4.7567,
      "mean_token_accuracy": 0.24366158992052078,
      "num_tokens": 71961171.0,
      "step": 31385
    },
    {
      "entropy": 5.150928974151611,
      "epoch": 3.015369836695485,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041021816510075366,
      "loss": 4.6382,
      "mean_token_accuracy": 0.2522772908210754,
      "num_tokens": 71972697.0,
      "step": 31390
    },
    {
      "entropy": 5.101291608810425,
      "epoch": 3.015850144092219,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00041019076164987696,
      "loss": 4.7072,
      "mean_token_accuracy": 0.2372485101222992,
      "num_tokens": 71985431.0,
      "step": 31395
    },
    {
      "entropy": 5.003251457214356,
      "epoch": 3.016330451488953,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004101633550602721,
      "loss": 4.5694,
      "mean_token_accuracy": 0.2610380470752716,
      "num_tokens": 71996935.0,
      "step": 31400
    },
    {
      "entropy": 5.091038942337036,
      "epoch": 3.0168107588856867,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004101359453325755,
      "loss": 4.6663,
      "mean_token_accuracy": 0.24339766055345535,
      "num_tokens": 72008227.0,
      "step": 31405
    },
    {
      "entropy": 5.238728857040405,
      "epoch": 3.0172910662824206,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041010853246742357,
      "loss": 4.7407,
      "mean_token_accuracy": 0.2404123455286026,
      "num_tokens": 72020478.0,
      "step": 31410
    },
    {
      "entropy": 5.18777756690979,
      "epoch": 3.0177713736791545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004100811164654527,
      "loss": 4.7113,
      "mean_token_accuracy": 0.2363765776157379,
      "num_tokens": 72031542.0,
      "step": 31415
    },
    {
      "entropy": 5.167224597930908,
      "epoch": 3.0182516810758884,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004100536973272994,
      "loss": 4.75,
      "mean_token_accuracy": 0.2388278976082802,
      "num_tokens": 72043956.0,
      "step": 31420
    },
    {
      "entropy": 5.138173627853393,
      "epoch": 3.0187319884726227,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004100262750536003,
      "loss": 4.6995,
      "mean_token_accuracy": 0.2472890406847,
      "num_tokens": 72055586.0,
      "step": 31425
    },
    {
      "entropy": 5.026668930053711,
      "epoch": 3.0192122958693566,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040999884964499196,
      "loss": 4.685,
      "mean_token_accuracy": 0.25033959597349165,
      "num_tokens": 72067889.0,
      "step": 31430
    },
    {
      "entropy": 5.07814507484436,
      "epoch": 3.0196926032660905,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040997142110211127,
      "loss": 4.6373,
      "mean_token_accuracy": 0.2456497922539711,
      "num_tokens": 72078948.0,
      "step": 31435
    },
    {
      "entropy": 5.101231527328491,
      "epoch": 3.0201729106628243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040994398942559496,
      "loss": 4.7824,
      "mean_token_accuracy": 0.23402330726385118,
      "num_tokens": 72090238.0,
      "step": 31440
    },
    {
      "entropy": 5.09940619468689,
      "epoch": 3.020653218059558,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004099165546160799,
      "loss": 4.7028,
      "mean_token_accuracy": 0.24008429050445557,
      "num_tokens": 72102848.0,
      "step": 31445
    },
    {
      "entropy": 5.169447135925293,
      "epoch": 3.021133525456292,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00040988911667420305,
      "loss": 4.6991,
      "mean_token_accuracy": 0.23964607268571853,
      "num_tokens": 72115363.0,
      "step": 31450
    },
    {
      "entropy": 5.087765312194824,
      "epoch": 3.021613832853026,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004098616756006015,
      "loss": 4.6727,
      "mean_token_accuracy": 0.2504300236701965,
      "num_tokens": 72126768.0,
      "step": 31455
    },
    {
      "entropy": 5.084184169769287,
      "epoch": 3.02209414024976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004098342313959122,
      "loss": 4.6821,
      "mean_token_accuracy": 0.2460756465792656,
      "num_tokens": 72138421.0,
      "step": 31460
    },
    {
      "entropy": 5.104829978942871,
      "epoch": 3.0225744476464937,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004098067840607725,
      "loss": 4.6915,
      "mean_token_accuracy": 0.24345028400421143,
      "num_tokens": 72149568.0,
      "step": 31465
    },
    {
      "entropy": 5.171456623077392,
      "epoch": 3.0230547550432276,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004097793335958195,
      "loss": 4.7437,
      "mean_token_accuracy": 0.24311020225286484,
      "num_tokens": 72161041.0,
      "step": 31470
    },
    {
      "entropy": 5.223282670974731,
      "epoch": 3.0235350624399615,
      "grad_norm": 1.25,
      "learning_rate": 0.00040975188000169074,
      "loss": 4.7673,
      "mean_token_accuracy": 0.2437180534005165,
      "num_tokens": 72173104.0,
      "step": 31475
    },
    {
      "entropy": 5.169146919250489,
      "epoch": 3.0240153698366954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040972442327902325,
      "loss": 4.7265,
      "mean_token_accuracy": 0.23872890919446946,
      "num_tokens": 72183549.0,
      "step": 31480
    },
    {
      "entropy": 4.9897020816802975,
      "epoch": 3.0244956772334293,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004096969634284549,
      "loss": 4.5733,
      "mean_token_accuracy": 0.2577987387776375,
      "num_tokens": 72195210.0,
      "step": 31485
    },
    {
      "entropy": 5.0926374912261965,
      "epoch": 3.024975984630163,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004096695004506229,
      "loss": 4.6451,
      "mean_token_accuracy": 0.2523143634200096,
      "num_tokens": 72205003.0,
      "step": 31490
    },
    {
      "entropy": 5.090652418136597,
      "epoch": 3.025456292026897,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040964203434616496,
      "loss": 4.6401,
      "mean_token_accuracy": 0.2419528216123581,
      "num_tokens": 72217069.0,
      "step": 31495
    },
    {
      "entropy": 5.124872732162475,
      "epoch": 3.025936599423631,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004096145651157189,
      "loss": 4.7156,
      "mean_token_accuracy": 0.24342144578695296,
      "num_tokens": 72229634.0,
      "step": 31500
    },
    {
      "entropy": 5.098457717895508,
      "epoch": 3.0264169068203652,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004095870927599223,
      "loss": 4.6152,
      "mean_token_accuracy": 0.24577573835849761,
      "num_tokens": 72241061.0,
      "step": 31505
    },
    {
      "entropy": 5.029171895980835,
      "epoch": 3.026897214217099,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040955961727941306,
      "loss": 4.6103,
      "mean_token_accuracy": 0.25175012797117235,
      "num_tokens": 72252506.0,
      "step": 31510
    },
    {
      "entropy": 5.051406526565552,
      "epoch": 3.027377521613833,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004095321386748291,
      "loss": 4.5771,
      "mean_token_accuracy": 0.24949017018079758,
      "num_tokens": 72264663.0,
      "step": 31515
    },
    {
      "entropy": 5.095618438720703,
      "epoch": 3.027857829010567,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040950465694680825,
      "loss": 4.6342,
      "mean_token_accuracy": 0.24885640442371368,
      "num_tokens": 72276665.0,
      "step": 31520
    },
    {
      "entropy": 5.103506851196289,
      "epoch": 3.0283381364073008,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040947717209598877,
      "loss": 4.653,
      "mean_token_accuracy": 0.24357238560914993,
      "num_tokens": 72287857.0,
      "step": 31525
    },
    {
      "entropy": 5.032822561264038,
      "epoch": 3.0288184438040346,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040944968412300867,
      "loss": 4.5946,
      "mean_token_accuracy": 0.252520290017128,
      "num_tokens": 72297775.0,
      "step": 31530
    },
    {
      "entropy": 5.170008134841919,
      "epoch": 3.0292987512007685,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040942219302850605,
      "loss": 4.8239,
      "mean_token_accuracy": 0.23771725296974183,
      "num_tokens": 72309350.0,
      "step": 31535
    },
    {
      "entropy": 5.078362607955933,
      "epoch": 3.0297790585975024,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004093946988131193,
      "loss": 4.6733,
      "mean_token_accuracy": 0.24496380090713502,
      "num_tokens": 72320037.0,
      "step": 31540
    },
    {
      "entropy": 5.080300235748291,
      "epoch": 3.0302593659942363,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004093672014774868,
      "loss": 4.6415,
      "mean_token_accuracy": 0.24760041385889053,
      "num_tokens": 72331871.0,
      "step": 31545
    },
    {
      "entropy": 5.069897079467774,
      "epoch": 3.03073967339097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040933970102224675,
      "loss": 4.6016,
      "mean_token_accuracy": 0.2492808535695076,
      "num_tokens": 72342792.0,
      "step": 31550
    },
    {
      "entropy": 5.10156135559082,
      "epoch": 3.031219980787704,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040931219744803774,
      "loss": 4.692,
      "mean_token_accuracy": 0.23512276262044907,
      "num_tokens": 72354466.0,
      "step": 31555
    },
    {
      "entropy": 5.156486940383911,
      "epoch": 3.031700288184438,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004092846907554985,
      "loss": 4.6732,
      "mean_token_accuracy": 0.23883960396051407,
      "num_tokens": 72366161.0,
      "step": 31560
    },
    {
      "entropy": 5.090292310714721,
      "epoch": 3.032180595581172,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040925718094526724,
      "loss": 4.6527,
      "mean_token_accuracy": 0.2509702771902084,
      "num_tokens": 72377724.0,
      "step": 31565
    },
    {
      "entropy": 5.232064437866211,
      "epoch": 3.0326609029779057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040922966801798305,
      "loss": 4.8366,
      "mean_token_accuracy": 0.230905120074749,
      "num_tokens": 72390498.0,
      "step": 31570
    },
    {
      "entropy": 5.150277423858642,
      "epoch": 3.0331412103746396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040920215197428456,
      "loss": 4.7359,
      "mean_token_accuracy": 0.2422695353627205,
      "num_tokens": 72401841.0,
      "step": 31575
    },
    {
      "entropy": 5.169709300994873,
      "epoch": 3.0336215177713735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040917463281481053,
      "loss": 4.771,
      "mean_token_accuracy": 0.23429617434740066,
      "num_tokens": 72412755.0,
      "step": 31580
    },
    {
      "entropy": 5.0675403594970705,
      "epoch": 3.034101825168108,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004091471105402,
      "loss": 4.653,
      "mean_token_accuracy": 0.2448899656534195,
      "num_tokens": 72424516.0,
      "step": 31585
    },
    {
      "entropy": 5.071693706512451,
      "epoch": 3.0345821325648417,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004091195851510918,
      "loss": 4.6208,
      "mean_token_accuracy": 0.2529422327876091,
      "num_tokens": 72435792.0,
      "step": 31590
    },
    {
      "entropy": 5.066959953308105,
      "epoch": 3.0350624399615755,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004090920566481252,
      "loss": 4.6679,
      "mean_token_accuracy": 0.25105773061513903,
      "num_tokens": 72448262.0,
      "step": 31595
    },
    {
      "entropy": 5.1736527442932125,
      "epoch": 3.0355427473583094,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004090645250319392,
      "loss": 4.6547,
      "mean_token_accuracy": 0.24791436940431594,
      "num_tokens": 72458329.0,
      "step": 31600
    },
    {
      "entropy": 5.1372878074646,
      "epoch": 3.0360230547550433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000409036990303173,
      "loss": 4.6598,
      "mean_token_accuracy": 0.25391810685396193,
      "num_tokens": 72469774.0,
      "step": 31605
    },
    {
      "entropy": 5.075024843215942,
      "epoch": 3.036503362151777,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004090094524624658,
      "loss": 4.7388,
      "mean_token_accuracy": 0.2464292138814926,
      "num_tokens": 72482619.0,
      "step": 31610
    },
    {
      "entropy": 5.1243963718414305,
      "epoch": 3.036983669548511,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040898191151045717,
      "loss": 4.7362,
      "mean_token_accuracy": 0.2372278794646263,
      "num_tokens": 72493553.0,
      "step": 31615
    },
    {
      "entropy": 5.052584886550903,
      "epoch": 3.037463976945245,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004089543674477864,
      "loss": 4.5814,
      "mean_token_accuracy": 0.2557071015238762,
      "num_tokens": 72504853.0,
      "step": 31620
    },
    {
      "entropy": 5.121501731872558,
      "epoch": 3.037944284341979,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004089268202750929,
      "loss": 4.702,
      "mean_token_accuracy": 0.2540815189480782,
      "num_tokens": 72515683.0,
      "step": 31625
    },
    {
      "entropy": 5.170930099487305,
      "epoch": 3.0384245917387127,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040889926999301634,
      "loss": 4.7423,
      "mean_token_accuracy": 0.24483331590890883,
      "num_tokens": 72525699.0,
      "step": 31630
    },
    {
      "entropy": 5.079156303405762,
      "epoch": 3.0389048991354466,
      "grad_norm": 1.0,
      "learning_rate": 0.0004088717166021964,
      "loss": 4.6554,
      "mean_token_accuracy": 0.24403244107961655,
      "num_tokens": 72536742.0,
      "step": 31635
    },
    {
      "entropy": 5.056245231628418,
      "epoch": 3.0393852065321805,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004088441601032727,
      "loss": 4.6381,
      "mean_token_accuracy": 0.25099740624427797,
      "num_tokens": 72548645.0,
      "step": 31640
    },
    {
      "entropy": 5.136969184875488,
      "epoch": 3.0398655139289144,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004088166004968849,
      "loss": 4.7396,
      "mean_token_accuracy": 0.24172220081090928,
      "num_tokens": 72560332.0,
      "step": 31645
    },
    {
      "entropy": 5.145677185058593,
      "epoch": 3.0403458213256482,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040878903778367317,
      "loss": 4.7043,
      "mean_token_accuracy": 0.24822908639907837,
      "num_tokens": 72572714.0,
      "step": 31650
    },
    {
      "entropy": 5.112582683563232,
      "epoch": 3.040826128722382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004087614719642772,
      "loss": 4.6206,
      "mean_token_accuracy": 0.25788799673318863,
      "num_tokens": 72583956.0,
      "step": 31655
    },
    {
      "entropy": 5.145257377624512,
      "epoch": 3.0413064361191164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040873390303933693,
      "loss": 4.6966,
      "mean_token_accuracy": 0.24499552994966506,
      "num_tokens": 72594429.0,
      "step": 31660
    },
    {
      "entropy": 5.118279695510864,
      "epoch": 3.0417867435158503,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040870633100949266,
      "loss": 4.7369,
      "mean_token_accuracy": 0.24089303612709045,
      "num_tokens": 72605841.0,
      "step": 31665
    },
    {
      "entropy": 5.109849834442139,
      "epoch": 3.042267050912584,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040867875587538436,
      "loss": 4.6189,
      "mean_token_accuracy": 0.25574866086244585,
      "num_tokens": 72616449.0,
      "step": 31670
    },
    {
      "entropy": 5.094426393508911,
      "epoch": 3.042747358309318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004086511776376523,
      "loss": 4.6644,
      "mean_token_accuracy": 0.24110874384641648,
      "num_tokens": 72627540.0,
      "step": 31675
    },
    {
      "entropy": 5.1538292407989506,
      "epoch": 3.043227665706052,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040862359629693684,
      "loss": 4.7956,
      "mean_token_accuracy": 0.2468473881483078,
      "num_tokens": 72638788.0,
      "step": 31680
    },
    {
      "entropy": 5.147992944717407,
      "epoch": 3.043707973102786,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004085960118538781,
      "loss": 4.758,
      "mean_token_accuracy": 0.23272383064031602,
      "num_tokens": 72650370.0,
      "step": 31685
    },
    {
      "entropy": 5.120801305770874,
      "epoch": 3.0441882804995197,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004085684243091168,
      "loss": 4.7542,
      "mean_token_accuracy": 0.24250922799110414,
      "num_tokens": 72662119.0,
      "step": 31690
    },
    {
      "entropy": 5.138534593582153,
      "epoch": 3.0446685878962536,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004085408336632933,
      "loss": 4.6909,
      "mean_token_accuracy": 0.24302243292331696,
      "num_tokens": 72673063.0,
      "step": 31695
    },
    {
      "entropy": 5.096362543106079,
      "epoch": 3.0451488952929875,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040851323991704803,
      "loss": 4.5793,
      "mean_token_accuracy": 0.2456071987748146,
      "num_tokens": 72684185.0,
      "step": 31700
    },
    {
      "entropy": 5.035220432281494,
      "epoch": 3.0456292026897214,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004084856430710219,
      "loss": 4.5884,
      "mean_token_accuracy": 0.2507237195968628,
      "num_tokens": 72695313.0,
      "step": 31705
    },
    {
      "entropy": 5.06360969543457,
      "epoch": 3.0461095100864553,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004084580431258555,
      "loss": 4.7072,
      "mean_token_accuracy": 0.2496033400297165,
      "num_tokens": 72707052.0,
      "step": 31710
    },
    {
      "entropy": 5.089043617248535,
      "epoch": 3.046589817483189,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004084304400821896,
      "loss": 4.6461,
      "mean_token_accuracy": 0.24965853989124298,
      "num_tokens": 72717833.0,
      "step": 31715
    },
    {
      "entropy": 5.1753698825836185,
      "epoch": 3.047070124879923,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004084028339406651,
      "loss": 4.709,
      "mean_token_accuracy": 0.24014876931905746,
      "num_tokens": 72729591.0,
      "step": 31720
    },
    {
      "entropy": 5.0955897808074955,
      "epoch": 3.047550432276657,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040837522470192297,
      "loss": 4.6615,
      "mean_token_accuracy": 0.2511131063103676,
      "num_tokens": 72740777.0,
      "step": 31725
    },
    {
      "entropy": 5.100281047821045,
      "epoch": 3.0480307396733908,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004083476123666041,
      "loss": 4.672,
      "mean_token_accuracy": 0.24825873523950576,
      "num_tokens": 72752093.0,
      "step": 31730
    },
    {
      "entropy": 5.164329576492309,
      "epoch": 3.048511047070125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004083199969353496,
      "loss": 4.7989,
      "mean_token_accuracy": 0.23310506343841553,
      "num_tokens": 72763366.0,
      "step": 31735
    },
    {
      "entropy": 5.139572334289551,
      "epoch": 3.048991354466859,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040829237840880075,
      "loss": 4.7157,
      "mean_token_accuracy": 0.24479717016220093,
      "num_tokens": 72776460.0,
      "step": 31740
    },
    {
      "entropy": 5.116616296768188,
      "epoch": 3.049471661863593,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040826475678759855,
      "loss": 4.7151,
      "mean_token_accuracy": 0.24522064626216888,
      "num_tokens": 72788001.0,
      "step": 31745
    },
    {
      "entropy": 5.088946580886841,
      "epoch": 3.0499519692603267,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004082371320723845,
      "loss": 4.6244,
      "mean_token_accuracy": 0.2525717094540596,
      "num_tokens": 72799640.0,
      "step": 31750
    },
    {
      "entropy": 5.095271730422974,
      "epoch": 3.0504322766570606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040820950426379986,
      "loss": 4.6842,
      "mean_token_accuracy": 0.24802108108997345,
      "num_tokens": 72810686.0,
      "step": 31755
    },
    {
      "entropy": 5.067455911636353,
      "epoch": 3.0509125840537945,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004081818733624861,
      "loss": 4.6226,
      "mean_token_accuracy": 0.2506804496049881,
      "num_tokens": 72820997.0,
      "step": 31760
    },
    {
      "entropy": 5.074386072158814,
      "epoch": 3.0513928914505284,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004081542393690847,
      "loss": 4.72,
      "mean_token_accuracy": 0.24677423089742662,
      "num_tokens": 72832044.0,
      "step": 31765
    },
    {
      "entropy": 5.080792570114136,
      "epoch": 3.0518731988472623,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004081266022842372,
      "loss": 4.658,
      "mean_token_accuracy": 0.25115212649106977,
      "num_tokens": 72842204.0,
      "step": 31770
    },
    {
      "entropy": 5.068198204040527,
      "epoch": 3.052353506243996,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040809896210858537,
      "loss": 4.6197,
      "mean_token_accuracy": 0.24891269505023955,
      "num_tokens": 72853650.0,
      "step": 31775
    },
    {
      "entropy": 5.1207318782806395,
      "epoch": 3.05283381364073,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040807131884277085,
      "loss": 4.6887,
      "mean_token_accuracy": 0.24775375574827194,
      "num_tokens": 72864692.0,
      "step": 31780
    },
    {
      "entropy": 5.0963348865509035,
      "epoch": 3.053314121037464,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004080436724874354,
      "loss": 4.6708,
      "mean_token_accuracy": 0.24319007247686386,
      "num_tokens": 72875871.0,
      "step": 31785
    },
    {
      "entropy": 5.129487752914429,
      "epoch": 3.053794428434198,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040801602304322095,
      "loss": 4.7161,
      "mean_token_accuracy": 0.2445603311061859,
      "num_tokens": 72888620.0,
      "step": 31790
    },
    {
      "entropy": 5.0917236328125,
      "epoch": 3.0542747358309317,
      "grad_norm": 1.0,
      "learning_rate": 0.00040798837051076944,
      "loss": 4.6787,
      "mean_token_accuracy": 0.24230535477399825,
      "num_tokens": 72901022.0,
      "step": 31795
    },
    {
      "entropy": 5.179385614395142,
      "epoch": 3.0547550432276656,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040796071489072286,
      "loss": 4.7912,
      "mean_token_accuracy": 0.23381830602884293,
      "num_tokens": 72913588.0,
      "step": 31800
    },
    {
      "entropy": 5.128277158737182,
      "epoch": 3.0552353506243994,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004079330561837233,
      "loss": 4.7045,
      "mean_token_accuracy": 0.23974111974239348,
      "num_tokens": 72925397.0,
      "step": 31805
    },
    {
      "entropy": 5.089568042755127,
      "epoch": 3.0557156580211333,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040790539439041287,
      "loss": 4.6566,
      "mean_token_accuracy": 0.24146271497011185,
      "num_tokens": 72936397.0,
      "step": 31810
    },
    {
      "entropy": 5.075946092605591,
      "epoch": 3.0561959654178676,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040787772951143386,
      "loss": 4.7343,
      "mean_token_accuracy": 0.24607449918985366,
      "num_tokens": 72948585.0,
      "step": 31815
    },
    {
      "entropy": 5.152942371368408,
      "epoch": 3.0566762728146015,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004078500615474285,
      "loss": 4.7001,
      "mean_token_accuracy": 0.2428433582186699,
      "num_tokens": 72961189.0,
      "step": 31820
    },
    {
      "entropy": 5.127501106262207,
      "epoch": 3.0571565802113354,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040782239049903926,
      "loss": 4.6045,
      "mean_token_accuracy": 0.25080667436122894,
      "num_tokens": 72972845.0,
      "step": 31825
    },
    {
      "entropy": 5.1660747051239015,
      "epoch": 3.0576368876080693,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040779471636690845,
      "loss": 4.6999,
      "mean_token_accuracy": 0.2392577975988388,
      "num_tokens": 72984401.0,
      "step": 31830
    },
    {
      "entropy": 5.063312196731568,
      "epoch": 3.058117195004803,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040776703915167866,
      "loss": 4.7124,
      "mean_token_accuracy": 0.2454993486404419,
      "num_tokens": 72995811.0,
      "step": 31835
    },
    {
      "entropy": 5.152497005462647,
      "epoch": 3.058597502401537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040773935885399254,
      "loss": 4.7234,
      "mean_token_accuracy": 0.24983121007680892,
      "num_tokens": 73006545.0,
      "step": 31840
    },
    {
      "entropy": 5.162114810943604,
      "epoch": 3.059077809798271,
      "grad_norm": 1.0,
      "learning_rate": 0.0004077116754744926,
      "loss": 4.6964,
      "mean_token_accuracy": 0.25045040249824524,
      "num_tokens": 73017144.0,
      "step": 31845
    },
    {
      "entropy": 5.16715292930603,
      "epoch": 3.059558117195005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040768398901382157,
      "loss": 4.7593,
      "mean_token_accuracy": 0.23923674076795579,
      "num_tokens": 73027996.0,
      "step": 31850
    },
    {
      "entropy": 5.084195709228515,
      "epoch": 3.0600384245917387,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004076562994726223,
      "loss": 4.6789,
      "mean_token_accuracy": 0.24117049276828767,
      "num_tokens": 73038949.0,
      "step": 31855
    },
    {
      "entropy": 5.155727481842041,
      "epoch": 3.0605187319884726,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004076286068515378,
      "loss": 4.7287,
      "mean_token_accuracy": 0.2453484356403351,
      "num_tokens": 73050249.0,
      "step": 31860
    },
    {
      "entropy": 5.176791095733643,
      "epoch": 3.0609990393852065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004076009111512108,
      "loss": 4.6516,
      "mean_token_accuracy": 0.2469186529517174,
      "num_tokens": 73061228.0,
      "step": 31865
    },
    {
      "entropy": 5.177712678909302,
      "epoch": 3.0614793467819403,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004075732123722844,
      "loss": 4.6445,
      "mean_token_accuracy": 0.24611999988555908,
      "num_tokens": 73070972.0,
      "step": 31870
    },
    {
      "entropy": 5.13157467842102,
      "epoch": 3.061959654178674,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004075455105154016,
      "loss": 4.7127,
      "mean_token_accuracy": 0.2384551376104355,
      "num_tokens": 73081569.0,
      "step": 31875
    },
    {
      "entropy": 5.0913749694824215,
      "epoch": 3.062439961575408,
      "grad_norm": 1.0,
      "learning_rate": 0.00040751780558120573,
      "loss": 4.707,
      "mean_token_accuracy": 0.24489471614360808,
      "num_tokens": 73092778.0,
      "step": 31880
    },
    {
      "entropy": 5.162841892242431,
      "epoch": 3.062920268972142,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004074900975703398,
      "loss": 4.801,
      "mean_token_accuracy": 0.23277629017829896,
      "num_tokens": 73104166.0,
      "step": 31885
    },
    {
      "entropy": 5.190942049026489,
      "epoch": 3.063400576368876,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004074623864834473,
      "loss": 4.7568,
      "mean_token_accuracy": 0.24359047710895537,
      "num_tokens": 73115732.0,
      "step": 31890
    },
    {
      "entropy": 5.134415531158448,
      "epoch": 3.06388088376561,
      "grad_norm": 1.125,
      "learning_rate": 0.0004074346723211715,
      "loss": 4.6648,
      "mean_token_accuracy": 0.24647997319698334,
      "num_tokens": 73126305.0,
      "step": 31895
    },
    {
      "entropy": 5.057104969024659,
      "epoch": 3.064361191162344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040740695508415583,
      "loss": 4.6306,
      "mean_token_accuracy": 0.24928771704435349,
      "num_tokens": 73138255.0,
      "step": 31900
    },
    {
      "entropy": 5.126017904281616,
      "epoch": 3.064841498559078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040737923477304386,
      "loss": 4.6103,
      "mean_token_accuracy": 0.25201036483049394,
      "num_tokens": 73148359.0,
      "step": 31905
    },
    {
      "entropy": 5.068912696838379,
      "epoch": 3.065321805955812,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00040735151138847917,
      "loss": 4.6232,
      "mean_token_accuracy": 0.24545711576938628,
      "num_tokens": 73159781.0,
      "step": 31910
    },
    {
      "entropy": 5.037901067733765,
      "epoch": 3.0658021133525457,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004073237849311053,
      "loss": 4.6817,
      "mean_token_accuracy": 0.24679249078035354,
      "num_tokens": 73171930.0,
      "step": 31915
    },
    {
      "entropy": 5.162428331375122,
      "epoch": 3.0662824207492796,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004072960554015661,
      "loss": 4.7339,
      "mean_token_accuracy": 0.24327213168144227,
      "num_tokens": 73183842.0,
      "step": 31920
    },
    {
      "entropy": 5.178036546707153,
      "epoch": 3.0667627281460135,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004072683228005055,
      "loss": 4.7719,
      "mean_token_accuracy": 0.23606629818677902,
      "num_tokens": 73195204.0,
      "step": 31925
    },
    {
      "entropy": 5.1644782543182375,
      "epoch": 3.0672430355427474,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040724058712856697,
      "loss": 4.7444,
      "mean_token_accuracy": 0.24884055703878402,
      "num_tokens": 73206829.0,
      "step": 31930
    },
    {
      "entropy": 5.04797215461731,
      "epoch": 3.0677233429394812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004072128483863948,
      "loss": 4.5982,
      "mean_token_accuracy": 0.26559618562459947,
      "num_tokens": 73217188.0,
      "step": 31935
    },
    {
      "entropy": 5.019267559051514,
      "epoch": 3.068203650336215,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004071851065746328,
      "loss": 4.603,
      "mean_token_accuracy": 0.2468065708875656,
      "num_tokens": 73228119.0,
      "step": 31940
    },
    {
      "entropy": 5.082492303848267,
      "epoch": 3.068683957732949,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004071573616939252,
      "loss": 4.7064,
      "mean_token_accuracy": 0.23784140944480897,
      "num_tokens": 73239876.0,
      "step": 31945
    },
    {
      "entropy": 5.114399099349976,
      "epoch": 3.069164265129683,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004071296137449161,
      "loss": 4.6747,
      "mean_token_accuracy": 0.250157830119133,
      "num_tokens": 73251834.0,
      "step": 31950
    },
    {
      "entropy": 5.099168491363526,
      "epoch": 3.0696445725264168,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040710186272824967,
      "loss": 4.619,
      "mean_token_accuracy": 0.24850722700357436,
      "num_tokens": 73262949.0,
      "step": 31955
    },
    {
      "entropy": 5.028237962722779,
      "epoch": 3.0701248799231506,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004070741086445703,
      "loss": 4.7057,
      "mean_token_accuracy": 0.24401334822177886,
      "num_tokens": 73274964.0,
      "step": 31960
    },
    {
      "entropy": 5.121504640579223,
      "epoch": 3.0706051873198845,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040704635149452223,
      "loss": 4.6663,
      "mean_token_accuracy": 0.24230137914419175,
      "num_tokens": 73286527.0,
      "step": 31965
    },
    {
      "entropy": 5.06420750617981,
      "epoch": 3.071085494716619,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040701859127875,
      "loss": 4.5907,
      "mean_token_accuracy": 0.25176784098148347,
      "num_tokens": 73297197.0,
      "step": 31970
    },
    {
      "entropy": 5.127895069122315,
      "epoch": 3.0715658021133527,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040699082799789814,
      "loss": 4.7133,
      "mean_token_accuracy": 0.24390658289194106,
      "num_tokens": 73308629.0,
      "step": 31975
    },
    {
      "entropy": 5.138438653945923,
      "epoch": 3.0720461095100866,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040696306165261117,
      "loss": 4.7386,
      "mean_token_accuracy": 0.2408783480525017,
      "num_tokens": 73321420.0,
      "step": 31980
    },
    {
      "entropy": 5.16006760597229,
      "epoch": 3.0725264169068205,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004069352922435337,
      "loss": 4.7304,
      "mean_token_accuracy": 0.2451960951089859,
      "num_tokens": 73332617.0,
      "step": 31985
    },
    {
      "entropy": 5.0173815250396725,
      "epoch": 3.0730067243035544,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004069075197713106,
      "loss": 4.6373,
      "mean_token_accuracy": 0.24463422298431398,
      "num_tokens": 73344687.0,
      "step": 31990
    },
    {
      "entropy": 5.062841796875,
      "epoch": 3.0734870317002883,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040687974423658655,
      "loss": 4.639,
      "mean_token_accuracy": 0.25947696566581724,
      "num_tokens": 73355885.0,
      "step": 31995
    },
    {
      "entropy": 5.073850965499878,
      "epoch": 3.073967339097022,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040685196564000644,
      "loss": 4.665,
      "mean_token_accuracy": 0.2490766391158104,
      "num_tokens": 73367147.0,
      "step": 32000
    },
    {
      "entropy": 5.244452857971192,
      "epoch": 3.074447646493756,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040682418398221517,
      "loss": 4.794,
      "mean_token_accuracy": 0.2358395680785179,
      "num_tokens": 73379780.0,
      "step": 32005
    },
    {
      "entropy": 5.08320689201355,
      "epoch": 3.07492795389049,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040679639926385783,
      "loss": 4.6774,
      "mean_token_accuracy": 0.2444912388920784,
      "num_tokens": 73392886.0,
      "step": 32010
    },
    {
      "entropy": 5.033668041229248,
      "epoch": 3.0754082612872238,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004067686114855794,
      "loss": 4.6523,
      "mean_token_accuracy": 0.24920229017734527,
      "num_tokens": 73404153.0,
      "step": 32015
    },
    {
      "entropy": 5.073902273178101,
      "epoch": 3.0758885686839577,
      "grad_norm": 1.0,
      "learning_rate": 0.00040674082064802507,
      "loss": 4.6958,
      "mean_token_accuracy": 0.2474759265780449,
      "num_tokens": 73415636.0,
      "step": 32020
    },
    {
      "entropy": 5.243733024597168,
      "epoch": 3.0763688760806915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004067130267518401,
      "loss": 4.8751,
      "mean_token_accuracy": 0.23374473452568054,
      "num_tokens": 73426349.0,
      "step": 32025
    },
    {
      "entropy": 5.146440839767456,
      "epoch": 3.0768491834774254,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004066852297976698,
      "loss": 4.678,
      "mean_token_accuracy": 0.24184914529323578,
      "num_tokens": 73437936.0,
      "step": 32030
    },
    {
      "entropy": 5.1483540534973145,
      "epoch": 3.0773294908741593,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004066574297861595,
      "loss": 4.7033,
      "mean_token_accuracy": 0.24702207297086715,
      "num_tokens": 73449818.0,
      "step": 32035
    },
    {
      "entropy": 5.005123043060303,
      "epoch": 3.077809798270893,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040662962671795454,
      "loss": 4.5293,
      "mean_token_accuracy": 0.2580830052495003,
      "num_tokens": 73459904.0,
      "step": 32040
    },
    {
      "entropy": 5.013719320297241,
      "epoch": 3.0782901056676275,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004066018205937006,
      "loss": 4.6722,
      "mean_token_accuracy": 0.2466215133666992,
      "num_tokens": 73470993.0,
      "step": 32045
    },
    {
      "entropy": 5.064469718933106,
      "epoch": 3.0787704130643614,
      "grad_norm": 1.0,
      "learning_rate": 0.0004065740114140431,
      "loss": 4.6167,
      "mean_token_accuracy": 0.24919567853212357,
      "num_tokens": 73482205.0,
      "step": 32050
    },
    {
      "entropy": 5.066756391525269,
      "epoch": 3.0792507204610953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040654619917962774,
      "loss": 4.6265,
      "mean_token_accuracy": 0.24706785529851913,
      "num_tokens": 73493990.0,
      "step": 32055
    },
    {
      "entropy": 5.156525754928589,
      "epoch": 3.079731027857829,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004065183838911003,
      "loss": 4.8115,
      "mean_token_accuracy": 0.2375935673713684,
      "num_tokens": 73505738.0,
      "step": 32060
    },
    {
      "entropy": 5.015220212936401,
      "epoch": 3.080211335254563,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004064905655491065,
      "loss": 4.5761,
      "mean_token_accuracy": 0.2562141418457031,
      "num_tokens": 73517248.0,
      "step": 32065
    },
    {
      "entropy": 5.182619428634643,
      "epoch": 3.080691642651297,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040646274415429224,
      "loss": 4.7873,
      "mean_token_accuracy": 0.2427953749895096,
      "num_tokens": 73528496.0,
      "step": 32070
    },
    {
      "entropy": 5.182440614700317,
      "epoch": 3.081171950048031,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004064349197073033,
      "loss": 4.7128,
      "mean_token_accuracy": 0.23418770879507064,
      "num_tokens": 73540672.0,
      "step": 32075
    },
    {
      "entropy": 5.1198983669281,
      "epoch": 3.0816522574447647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004064070922087859,
      "loss": 4.688,
      "mean_token_accuracy": 0.24684109538793564,
      "num_tokens": 73551683.0,
      "step": 32080
    },
    {
      "entropy": 5.0792396545410154,
      "epoch": 3.0821325648414986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040637926165938606,
      "loss": 4.6003,
      "mean_token_accuracy": 0.2504163682460785,
      "num_tokens": 73562981.0,
      "step": 32085
    },
    {
      "entropy": 5.152011728286743,
      "epoch": 3.0826128722382324,
      "grad_norm": 1.0,
      "learning_rate": 0.00040635142805974986,
      "loss": 4.782,
      "mean_token_accuracy": 0.24795447587966918,
      "num_tokens": 73574082.0,
      "step": 32090
    },
    {
      "entropy": 5.084204196929932,
      "epoch": 3.0830931796349663,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004063235914105235,
      "loss": 4.6115,
      "mean_token_accuracy": 0.24949991554021836,
      "num_tokens": 73586305.0,
      "step": 32095
    },
    {
      "entropy": 5.188805675506591,
      "epoch": 3.0835734870317,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040629575171235327,
      "loss": 4.7693,
      "mean_token_accuracy": 0.2404895007610321,
      "num_tokens": 73598036.0,
      "step": 32100
    },
    {
      "entropy": 5.0834637641906735,
      "epoch": 3.084053794428434,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004062679089658856,
      "loss": 4.6827,
      "mean_token_accuracy": 0.24890461564064026,
      "num_tokens": 73608870.0,
      "step": 32105
    },
    {
      "entropy": 5.149138975143432,
      "epoch": 3.084534101825168,
      "grad_norm": 1.0,
      "learning_rate": 0.00040624006317176685,
      "loss": 4.7092,
      "mean_token_accuracy": 0.2347562149167061,
      "num_tokens": 73619965.0,
      "step": 32110
    },
    {
      "entropy": 5.102905511856079,
      "epoch": 3.085014409221902,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040621221433064354,
      "loss": 4.5964,
      "mean_token_accuracy": 0.2559296891093254,
      "num_tokens": 73631183.0,
      "step": 32115
    },
    {
      "entropy": 5.081556415557861,
      "epoch": 3.0854947166186357,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004061843624431623,
      "loss": 4.6697,
      "mean_token_accuracy": 0.24717915058135986,
      "num_tokens": 73643550.0,
      "step": 32120
    },
    {
      "entropy": 5.150254440307617,
      "epoch": 3.08597502401537,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040615650750996956,
      "loss": 4.7525,
      "mean_token_accuracy": 0.24262361228466034,
      "num_tokens": 73654931.0,
      "step": 32125
    },
    {
      "entropy": 5.159456443786621,
      "epoch": 3.086455331412104,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040612864953171223,
      "loss": 4.7593,
      "mean_token_accuracy": 0.23840123564004898,
      "num_tokens": 73667771.0,
      "step": 32130
    },
    {
      "entropy": 5.155754899978637,
      "epoch": 3.086935638808838,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040610078850903715,
      "loss": 4.6477,
      "mean_token_accuracy": 0.25237128138542175,
      "num_tokens": 73678756.0,
      "step": 32135
    },
    {
      "entropy": 5.13864312171936,
      "epoch": 3.0874159462055717,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040607292444259094,
      "loss": 4.6495,
      "mean_token_accuracy": 0.2540803089737892,
      "num_tokens": 73689312.0,
      "step": 32140
    },
    {
      "entropy": 5.022158288955689,
      "epoch": 3.0878962536023056,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004060450573330206,
      "loss": 4.6082,
      "mean_token_accuracy": 0.24720314741134644,
      "num_tokens": 73702472.0,
      "step": 32145
    },
    {
      "entropy": 5.088761901855468,
      "epoch": 3.0883765609990395,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040601718718097325,
      "loss": 4.6851,
      "mean_token_accuracy": 0.2434731274843216,
      "num_tokens": 73714687.0,
      "step": 32150
    },
    {
      "entropy": 5.072689628601074,
      "epoch": 3.0888568683957733,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040598931398709576,
      "loss": 4.6374,
      "mean_token_accuracy": 0.238906392455101,
      "num_tokens": 73726137.0,
      "step": 32155
    },
    {
      "entropy": 5.071384763717651,
      "epoch": 3.089337175792507,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040596143775203534,
      "loss": 4.6787,
      "mean_token_accuracy": 0.24822606146335602,
      "num_tokens": 73737409.0,
      "step": 32160
    },
    {
      "entropy": 5.05735673904419,
      "epoch": 3.089817483189241,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00040593355847643933,
      "loss": 4.6456,
      "mean_token_accuracy": 0.2491337850689888,
      "num_tokens": 73750408.0,
      "step": 32165
    },
    {
      "entropy": 5.109059572219849,
      "epoch": 3.090297790585975,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004059056761609548,
      "loss": 4.6598,
      "mean_token_accuracy": 0.24477452635765076,
      "num_tokens": 73762417.0,
      "step": 32170
    },
    {
      "entropy": 5.112762880325318,
      "epoch": 3.090778097982709,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004058777908062292,
      "loss": 4.6918,
      "mean_token_accuracy": 0.24259697049856185,
      "num_tokens": 73774867.0,
      "step": 32175
    },
    {
      "entropy": 5.142957782745361,
      "epoch": 3.0912584053794427,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004058499024129099,
      "loss": 4.7528,
      "mean_token_accuracy": 0.24636638462543486,
      "num_tokens": 73785619.0,
      "step": 32180
    },
    {
      "entropy": 5.114127969741821,
      "epoch": 3.0917387127761766,
      "grad_norm": 0.90625,
      "learning_rate": 0.00040582201098164443,
      "loss": 4.7184,
      "mean_token_accuracy": 0.24315282106399536,
      "num_tokens": 73797520.0,
      "step": 32185
    },
    {
      "entropy": 5.069916200637818,
      "epoch": 3.0922190201729105,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040579411651308034,
      "loss": 4.6105,
      "mean_token_accuracy": 0.2477358475327492,
      "num_tokens": 73809135.0,
      "step": 32190
    },
    {
      "entropy": 5.02493805885315,
      "epoch": 3.0926993275696444,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040576621900786523,
      "loss": 4.6026,
      "mean_token_accuracy": 0.25245596617460253,
      "num_tokens": 73820126.0,
      "step": 32195
    },
    {
      "entropy": 5.226872301101684,
      "epoch": 3.0931796349663783,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004057383184666468,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22878171354532242,
      "num_tokens": 73832788.0,
      "step": 32200
    },
    {
      "entropy": 5.1289506435394285,
      "epoch": 3.0936599423631126,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040571041489007286,
      "loss": 4.687,
      "mean_token_accuracy": 0.2480057179927826,
      "num_tokens": 73845322.0,
      "step": 32205
    },
    {
      "entropy": 5.076383495330811,
      "epoch": 3.0941402497598465,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040568250827879127,
      "loss": 4.7256,
      "mean_token_accuracy": 0.24741925299167633,
      "num_tokens": 73857818.0,
      "step": 32210
    },
    {
      "entropy": 4.984539747238159,
      "epoch": 3.0946205571565804,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004056545986334497,
      "loss": 4.5494,
      "mean_token_accuracy": 0.25599558651447296,
      "num_tokens": 73869373.0,
      "step": 32215
    },
    {
      "entropy": 5.074178218841553,
      "epoch": 3.0951008645533142,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004056266859546965,
      "loss": 4.6354,
      "mean_token_accuracy": 0.2471168965101242,
      "num_tokens": 73879219.0,
      "step": 32220
    },
    {
      "entropy": 5.076261568069458,
      "epoch": 3.095581171950048,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004055987702431795,
      "loss": 4.7028,
      "mean_token_accuracy": 0.24605603963136674,
      "num_tokens": 73890287.0,
      "step": 32225
    },
    {
      "entropy": 5.082270383834839,
      "epoch": 3.096061479346782,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040557085149954677,
      "loss": 4.6849,
      "mean_token_accuracy": 0.2408321261405945,
      "num_tokens": 73902798.0,
      "step": 32230
    },
    {
      "entropy": 5.139325428009033,
      "epoch": 3.096541786743516,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040554292972444663,
      "loss": 4.6916,
      "mean_token_accuracy": 0.2455345645546913,
      "num_tokens": 73914042.0,
      "step": 32235
    },
    {
      "entropy": 5.03339900970459,
      "epoch": 3.0970220941402498,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040551500491852735,
      "loss": 4.5636,
      "mean_token_accuracy": 0.2592805415391922,
      "num_tokens": 73926682.0,
      "step": 32240
    },
    {
      "entropy": 5.0699278831481935,
      "epoch": 3.0975024015369836,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004054870770824371,
      "loss": 4.681,
      "mean_token_accuracy": 0.24787142425775527,
      "num_tokens": 73937657.0,
      "step": 32245
    },
    {
      "entropy": 5.080751132965088,
      "epoch": 3.0979827089337175,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040545914621682445,
      "loss": 4.6802,
      "mean_token_accuracy": 0.24944050759077072,
      "num_tokens": 73950234.0,
      "step": 32250
    },
    {
      "entropy": 5.175245761871338,
      "epoch": 3.0984630163304514,
      "grad_norm": 1.0,
      "learning_rate": 0.0004054312123223378,
      "loss": 4.7079,
      "mean_token_accuracy": 0.2386370927095413,
      "num_tokens": 73961404.0,
      "step": 32255
    },
    {
      "entropy": 5.086889314651489,
      "epoch": 3.0989433237271853,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040540327539962567,
      "loss": 4.6809,
      "mean_token_accuracy": 0.24292728453874587,
      "num_tokens": 73973189.0,
      "step": 32260
    },
    {
      "entropy": 5.083288478851318,
      "epoch": 3.099423631123919,
      "grad_norm": 0.890625,
      "learning_rate": 0.00040537533544933674,
      "loss": 4.6672,
      "mean_token_accuracy": 0.24564946293830872,
      "num_tokens": 73985303.0,
      "step": 32265
    },
    {
      "entropy": 5.069171810150147,
      "epoch": 3.099903938520653,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004053473924721197,
      "loss": 4.6285,
      "mean_token_accuracy": 0.24723270982503892,
      "num_tokens": 73997008.0,
      "step": 32270
    },
    {
      "entropy": 5.104254961013794,
      "epoch": 3.100384245917387,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004053194464686232,
      "loss": 4.6772,
      "mean_token_accuracy": 0.24046698659658433,
      "num_tokens": 74007752.0,
      "step": 32275
    },
    {
      "entropy": 5.079176664352417,
      "epoch": 3.1008645533141213,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004052914974394961,
      "loss": 4.696,
      "mean_token_accuracy": 0.24078585505485534,
      "num_tokens": 74020247.0,
      "step": 32280
    },
    {
      "entropy": 5.128537702560425,
      "epoch": 3.101344860710855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040526354538538735,
      "loss": 4.6907,
      "mean_token_accuracy": 0.24463727921247483,
      "num_tokens": 74030823.0,
      "step": 32285
    },
    {
      "entropy": 5.062671804428101,
      "epoch": 3.101825168107589,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004052355903069459,
      "loss": 4.6235,
      "mean_token_accuracy": 0.24839554727077484,
      "num_tokens": 74042157.0,
      "step": 32290
    },
    {
      "entropy": 5.073385715484619,
      "epoch": 3.102305475504323,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004052076322048207,
      "loss": 4.7336,
      "mean_token_accuracy": 0.24195914566516877,
      "num_tokens": 74052772.0,
      "step": 32295
    },
    {
      "entropy": 5.103564453125,
      "epoch": 3.1027857829010568,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040517967107966095,
      "loss": 4.6672,
      "mean_token_accuracy": 0.24935964792966842,
      "num_tokens": 74063694.0,
      "step": 32300
    },
    {
      "entropy": 5.0419079780578615,
      "epoch": 3.1032660902977907,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040515170693211584,
      "loss": 4.5517,
      "mean_token_accuracy": 0.24688998460769654,
      "num_tokens": 74075249.0,
      "step": 32305
    },
    {
      "entropy": 5.082829523086548,
      "epoch": 3.1037463976945245,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004051237397628345,
      "loss": 4.6977,
      "mean_token_accuracy": 0.2370525985956192,
      "num_tokens": 74086495.0,
      "step": 32310
    },
    {
      "entropy": 5.127979516983032,
      "epoch": 3.1042267050912584,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004050957695724663,
      "loss": 4.7516,
      "mean_token_accuracy": 0.23851024508476257,
      "num_tokens": 74097812.0,
      "step": 32315
    },
    {
      "entropy": 5.016599559783936,
      "epoch": 3.1047070124879923,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004050677963616607,
      "loss": 4.5544,
      "mean_token_accuracy": 0.26197843849658964,
      "num_tokens": 74108680.0,
      "step": 32320
    },
    {
      "entropy": 5.03653621673584,
      "epoch": 3.105187319884726,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040503982013106706,
      "loss": 4.663,
      "mean_token_accuracy": 0.25310470908880234,
      "num_tokens": 74120666.0,
      "step": 32325
    },
    {
      "entropy": 5.046400880813598,
      "epoch": 3.10566762728146,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000405011840881335,
      "loss": 4.6515,
      "mean_token_accuracy": 0.24986546039581298,
      "num_tokens": 74133092.0,
      "step": 32330
    },
    {
      "entropy": 5.128210926055909,
      "epoch": 3.106147934678194,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004049838586131139,
      "loss": 4.7417,
      "mean_token_accuracy": 0.24086080640554428,
      "num_tokens": 74145109.0,
      "step": 32335
    },
    {
      "entropy": 5.09159836769104,
      "epoch": 3.106628242074928,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004049558733270537,
      "loss": 4.6715,
      "mean_token_accuracy": 0.24745867252349854,
      "num_tokens": 74156505.0,
      "step": 32340
    },
    {
      "entropy": 5.1186858177185055,
      "epoch": 3.1071085494716617,
      "grad_norm": 1.03125,
      "learning_rate": 0.000404927885023804,
      "loss": 4.7292,
      "mean_token_accuracy": 0.2433025896549225,
      "num_tokens": 74168213.0,
      "step": 32345
    },
    {
      "entropy": 5.0924906730651855,
      "epoch": 3.1075888568683956,
      "grad_norm": 0.890625,
      "learning_rate": 0.00040489989370401456,
      "loss": 4.6553,
      "mean_token_accuracy": 0.24501588344573974,
      "num_tokens": 74178758.0,
      "step": 32350
    },
    {
      "entropy": 5.095537996292114,
      "epoch": 3.1080691642651295,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004048718993683353,
      "loss": 4.6539,
      "mean_token_accuracy": 0.25398500114679334,
      "num_tokens": 74189716.0,
      "step": 32355
    },
    {
      "entropy": 5.080453491210937,
      "epoch": 3.108549471661864,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040484390201741627,
      "loss": 4.674,
      "mean_token_accuracy": 0.24516672343015672,
      "num_tokens": 74202225.0,
      "step": 32360
    },
    {
      "entropy": 5.023632001876831,
      "epoch": 3.1090297790585977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004048159016519073,
      "loss": 4.5888,
      "mean_token_accuracy": 0.24939750134944916,
      "num_tokens": 74212573.0,
      "step": 32365
    },
    {
      "entropy": 5.06605167388916,
      "epoch": 3.1095100864553316,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004047878982724586,
      "loss": 4.6795,
      "mean_token_accuracy": 0.24884903877973558,
      "num_tokens": 74223415.0,
      "step": 32370
    },
    {
      "entropy": 5.048687219619751,
      "epoch": 3.1099903938520654,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040475989187972034,
      "loss": 4.6616,
      "mean_token_accuracy": 0.24700356274843216,
      "num_tokens": 74235046.0,
      "step": 32375
    },
    {
      "entropy": 5.0443727493286135,
      "epoch": 3.1104707012487993,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040473188247434265,
      "loss": 4.6773,
      "mean_token_accuracy": 0.2466394141316414,
      "num_tokens": 74247286.0,
      "step": 32380
    },
    {
      "entropy": 5.068221664428711,
      "epoch": 3.110951008645533,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040470387005697587,
      "loss": 4.6369,
      "mean_token_accuracy": 0.24985899180173873,
      "num_tokens": 74258753.0,
      "step": 32385
    },
    {
      "entropy": 5.082240772247315,
      "epoch": 3.111431316042267,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004046758546282704,
      "loss": 4.7168,
      "mean_token_accuracy": 0.24989415407180787,
      "num_tokens": 74268878.0,
      "step": 32390
    },
    {
      "entropy": 5.101941680908203,
      "epoch": 3.111911623439001,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004046478361888766,
      "loss": 4.5948,
      "mean_token_accuracy": 0.2488268107175827,
      "num_tokens": 74280912.0,
      "step": 32395
    },
    {
      "entropy": 5.078135538101196,
      "epoch": 3.112391930835735,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000404619814739445,
      "loss": 4.6633,
      "mean_token_accuracy": 0.2501836150884628,
      "num_tokens": 74291960.0,
      "step": 32400
    },
    {
      "entropy": 5.079361915588379,
      "epoch": 3.1128722382324687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004045917902806263,
      "loss": 4.6039,
      "mean_token_accuracy": 0.24740735739469527,
      "num_tokens": 74302285.0,
      "step": 32405
    },
    {
      "entropy": 5.122711896896362,
      "epoch": 3.1133525456292026,
      "grad_norm": 1.15625,
      "learning_rate": 0.000404563762813071,
      "loss": 4.7448,
      "mean_token_accuracy": 0.24058245420455932,
      "num_tokens": 74313833.0,
      "step": 32410
    },
    {
      "entropy": 5.204500579833985,
      "epoch": 3.1138328530259365,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004045357323374298,
      "loss": 4.8209,
      "mean_token_accuracy": 0.23884305655956267,
      "num_tokens": 74324910.0,
      "step": 32415
    },
    {
      "entropy": 5.137303972244263,
      "epoch": 3.1143131604226704,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040450769885435364,
      "loss": 4.7043,
      "mean_token_accuracy": 0.24388520568609237,
      "num_tokens": 74337860.0,
      "step": 32420
    },
    {
      "entropy": 5.02867865562439,
      "epoch": 3.1147934678194042,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040447966236449313,
      "loss": 4.5701,
      "mean_token_accuracy": 0.252069029211998,
      "num_tokens": 74348769.0,
      "step": 32425
    },
    {
      "entropy": 5.069629669189453,
      "epoch": 3.115273775216138,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040445162286849935,
      "loss": 4.6395,
      "mean_token_accuracy": 0.25150172263383863,
      "num_tokens": 74360007.0,
      "step": 32430
    },
    {
      "entropy": 5.08478479385376,
      "epoch": 3.115754082612872,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040442358036702343,
      "loss": 4.6562,
      "mean_token_accuracy": 0.2485018864274025,
      "num_tokens": 74369455.0,
      "step": 32435
    },
    {
      "entropy": 5.107403564453125,
      "epoch": 3.1162343900096063,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004043955348607161,
      "loss": 4.6436,
      "mean_token_accuracy": 0.2517516478896141,
      "num_tokens": 74381696.0,
      "step": 32440
    },
    {
      "entropy": 5.134664392471313,
      "epoch": 3.11671469740634,
      "grad_norm": 1.0,
      "learning_rate": 0.0004043674863502288,
      "loss": 4.7021,
      "mean_token_accuracy": 0.24264785498380662,
      "num_tokens": 74392300.0,
      "step": 32445
    },
    {
      "entropy": 5.071734762191772,
      "epoch": 3.117195004803074,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040433943483621253,
      "loss": 4.6874,
      "mean_token_accuracy": 0.2533950537443161,
      "num_tokens": 74403953.0,
      "step": 32450
    },
    {
      "entropy": 5.120111131668091,
      "epoch": 3.117675312199808,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004043113803193187,
      "loss": 4.7261,
      "mean_token_accuracy": 0.2440980166196823,
      "num_tokens": 74415789.0,
      "step": 32455
    },
    {
      "entropy": 5.1853724956512455,
      "epoch": 3.118155619596542,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040428332280019864,
      "loss": 4.7968,
      "mean_token_accuracy": 0.23154036104679107,
      "num_tokens": 74428173.0,
      "step": 32460
    },
    {
      "entropy": 5.029074954986572,
      "epoch": 3.1186359269932757,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004042552622795036,
      "loss": 4.5396,
      "mean_token_accuracy": 0.2573298677802086,
      "num_tokens": 74439623.0,
      "step": 32465
    },
    {
      "entropy": 5.110775279998779,
      "epoch": 3.1191162343900096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004042271987578852,
      "loss": 4.752,
      "mean_token_accuracy": 0.24099194407463073,
      "num_tokens": 74451194.0,
      "step": 32470
    },
    {
      "entropy": 5.088257646560669,
      "epoch": 3.1195965417867435,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00040419913223599505,
      "loss": 4.6461,
      "mean_token_accuracy": 0.24991240948438645,
      "num_tokens": 74462395.0,
      "step": 32475
    },
    {
      "entropy": 5.017224311828613,
      "epoch": 3.1200768491834774,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040417106271448464,
      "loss": 4.5977,
      "mean_token_accuracy": 0.2553364664316177,
      "num_tokens": 74473332.0,
      "step": 32480
    },
    {
      "entropy": 5.187587547302246,
      "epoch": 3.1205571565802113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004041429901940057,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23021783381700517,
      "num_tokens": 74485605.0,
      "step": 32485
    },
    {
      "entropy": 5.0407788276672365,
      "epoch": 3.121037463976945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040411491467521,
      "loss": 4.5931,
      "mean_token_accuracy": 0.2587581902742386,
      "num_tokens": 74496901.0,
      "step": 32490
    },
    {
      "entropy": 5.127786493301391,
      "epoch": 3.121517771373679,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004040868361587494,
      "loss": 4.7818,
      "mean_token_accuracy": 0.24059900194406508,
      "num_tokens": 74508067.0,
      "step": 32495
    },
    {
      "entropy": 5.135271263122559,
      "epoch": 3.121998078770413,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004040587546452758,
      "loss": 4.7072,
      "mean_token_accuracy": 0.2363157793879509,
      "num_tokens": 74519991.0,
      "step": 32500
    },
    {
      "entropy": 5.128338861465454,
      "epoch": 3.122478386167147,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040403067013544116,
      "loss": 4.6837,
      "mean_token_accuracy": 0.24221144765615463,
      "num_tokens": 74532462.0,
      "step": 32505
    },
    {
      "entropy": 5.114025545120239,
      "epoch": 3.1229586935638807,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040400258262989744,
      "loss": 4.7095,
      "mean_token_accuracy": 0.24701208770275115,
      "num_tokens": 74543161.0,
      "step": 32510
    },
    {
      "entropy": 5.112951946258545,
      "epoch": 3.123439000960615,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040397449212929676,
      "loss": 4.6974,
      "mean_token_accuracy": 0.24052483588457108,
      "num_tokens": 74554904.0,
      "step": 32515
    },
    {
      "entropy": 5.12147626876831,
      "epoch": 3.123919308357349,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004039463986342914,
      "loss": 4.6501,
      "mean_token_accuracy": 0.24824616461992263,
      "num_tokens": 74566632.0,
      "step": 32520
    },
    {
      "entropy": 5.107873344421387,
      "epoch": 3.1243996157540828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040391830214553365,
      "loss": 4.7175,
      "mean_token_accuracy": 0.24429922550916672,
      "num_tokens": 74578151.0,
      "step": 32525
    },
    {
      "entropy": 5.105573844909668,
      "epoch": 3.1248799231508166,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004038902026636756,
      "loss": 4.7151,
      "mean_token_accuracy": 0.24164052605628966,
      "num_tokens": 74589615.0,
      "step": 32530
    },
    {
      "entropy": 5.10127739906311,
      "epoch": 3.1253602305475505,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004038621001893698,
      "loss": 4.676,
      "mean_token_accuracy": 0.2416013926267624,
      "num_tokens": 74600549.0,
      "step": 32535
    },
    {
      "entropy": 5.078930377960205,
      "epoch": 3.1258405379442844,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040383399472326874,
      "loss": 4.6378,
      "mean_token_accuracy": 0.25172407925128937,
      "num_tokens": 74612449.0,
      "step": 32540
    },
    {
      "entropy": 5.006037473678589,
      "epoch": 3.1263208453410183,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040380588626602484,
      "loss": 4.583,
      "mean_token_accuracy": 0.25625180304050443,
      "num_tokens": 74623804.0,
      "step": 32545
    },
    {
      "entropy": 5.062789726257324,
      "epoch": 3.126801152737752,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004037777748182907,
      "loss": 4.6886,
      "mean_token_accuracy": 0.2500510856509209,
      "num_tokens": 74636711.0,
      "step": 32550
    },
    {
      "entropy": 5.165604543685913,
      "epoch": 3.127281460134486,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004037496603807191,
      "loss": 4.7005,
      "mean_token_accuracy": 0.24427812546491623,
      "num_tokens": 74649037.0,
      "step": 32555
    },
    {
      "entropy": 5.138341283798217,
      "epoch": 3.12776176753122,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004037215429539626,
      "loss": 4.7124,
      "mean_token_accuracy": 0.24061508029699324,
      "num_tokens": 74660730.0,
      "step": 32560
    },
    {
      "entropy": 5.040322399139404,
      "epoch": 3.128242074927954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040369342253867413,
      "loss": 4.6455,
      "mean_token_accuracy": 0.2566502928733826,
      "num_tokens": 74672558.0,
      "step": 32565
    },
    {
      "entropy": 5.042397165298462,
      "epoch": 3.1287223823246877,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004036652991355066,
      "loss": 4.604,
      "mean_token_accuracy": 0.2589424893260002,
      "num_tokens": 74684777.0,
      "step": 32570
    },
    {
      "entropy": 5.147999286651611,
      "epoch": 3.1292026897214216,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004036371727451128,
      "loss": 4.6961,
      "mean_token_accuracy": 0.2440796747803688,
      "num_tokens": 74695706.0,
      "step": 32575
    },
    {
      "entropy": 5.11173644065857,
      "epoch": 3.1296829971181555,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040360904336814586,
      "loss": 4.7074,
      "mean_token_accuracy": 0.23528954088687898,
      "num_tokens": 74708277.0,
      "step": 32580
    },
    {
      "entropy": 5.130792856216431,
      "epoch": 3.1301633045148893,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004035809110052588,
      "loss": 4.6927,
      "mean_token_accuracy": 0.24281734526157378,
      "num_tokens": 74719188.0,
      "step": 32585
    },
    {
      "entropy": 5.1100115299224855,
      "epoch": 3.1306436119116237,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004035527756571048,
      "loss": 4.7129,
      "mean_token_accuracy": 0.24252479821443557,
      "num_tokens": 74730170.0,
      "step": 32590
    },
    {
      "entropy": 5.05640811920166,
      "epoch": 3.1311239193083575,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00040352463732433707,
      "loss": 4.6009,
      "mean_token_accuracy": 0.24299954026937484,
      "num_tokens": 74741212.0,
      "step": 32595
    },
    {
      "entropy": 5.119296360015869,
      "epoch": 3.1316042267050914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040349649600760894,
      "loss": 4.6695,
      "mean_token_accuracy": 0.24674908369779586,
      "num_tokens": 74752323.0,
      "step": 32600
    },
    {
      "entropy": 5.112207841873169,
      "epoch": 3.1320845341018253,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004034683517075737,
      "loss": 4.7003,
      "mean_token_accuracy": 0.2460939481854439,
      "num_tokens": 74763462.0,
      "step": 32605
    },
    {
      "entropy": 5.1428831100463865,
      "epoch": 3.132564841498559,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040344020442488476,
      "loss": 4.6952,
      "mean_token_accuracy": 0.2513548001646996,
      "num_tokens": 74774701.0,
      "step": 32610
    },
    {
      "entropy": 5.141017627716065,
      "epoch": 3.133045148895293,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040341205416019577,
      "loss": 4.7136,
      "mean_token_accuracy": 0.24167830049991607,
      "num_tokens": 74786378.0,
      "step": 32615
    },
    {
      "entropy": 5.124266767501831,
      "epoch": 3.133525456292027,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004033839009141601,
      "loss": 4.6598,
      "mean_token_accuracy": 0.2477421835064888,
      "num_tokens": 74798842.0,
      "step": 32620
    },
    {
      "entropy": 5.109239530563355,
      "epoch": 3.134005763688761,
      "grad_norm": 0.90625,
      "learning_rate": 0.00040335574468743145,
      "loss": 4.6409,
      "mean_token_accuracy": 0.2423310846090317,
      "num_tokens": 74809707.0,
      "step": 32625
    },
    {
      "entropy": 5.045934009552002,
      "epoch": 3.1344860710854947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004033275854806636,
      "loss": 4.6839,
      "mean_token_accuracy": 0.24761220663785935,
      "num_tokens": 74821075.0,
      "step": 32630
    },
    {
      "entropy": 5.110152244567871,
      "epoch": 3.1349663784822286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004032994232945103,
      "loss": 4.7449,
      "mean_token_accuracy": 0.23715028464794158,
      "num_tokens": 74833096.0,
      "step": 32635
    },
    {
      "entropy": 5.146285533905029,
      "epoch": 3.1354466858789625,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004032712581296253,
      "loss": 4.7021,
      "mean_token_accuracy": 0.24899860322475434,
      "num_tokens": 74844512.0,
      "step": 32640
    },
    {
      "entropy": 5.148516321182251,
      "epoch": 3.1359269932756964,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040324308998666267,
      "loss": 4.7132,
      "mean_token_accuracy": 0.253988641500473,
      "num_tokens": 74854253.0,
      "step": 32645
    },
    {
      "entropy": 5.071900701522827,
      "epoch": 3.1364073006724302,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040321491886627614,
      "loss": 4.7225,
      "mean_token_accuracy": 0.24385782033205033,
      "num_tokens": 74865772.0,
      "step": 32650
    },
    {
      "entropy": 5.1008813858032225,
      "epoch": 3.136887608069164,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040318674476912006,
      "loss": 4.6954,
      "mean_token_accuracy": 0.2440424755215645,
      "num_tokens": 74875773.0,
      "step": 32655
    },
    {
      "entropy": 5.155717039108277,
      "epoch": 3.137367915465898,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004031585676958483,
      "loss": 4.7607,
      "mean_token_accuracy": 0.2399914652109146,
      "num_tokens": 74887508.0,
      "step": 32660
    },
    {
      "entropy": 5.061465883255005,
      "epoch": 3.1378482228626323,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040313038764711517,
      "loss": 4.541,
      "mean_token_accuracy": 0.26391522735357287,
      "num_tokens": 74897350.0,
      "step": 32665
    },
    {
      "entropy": 4.986086034774781,
      "epoch": 3.138328530259366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040310220462357494,
      "loss": 4.5698,
      "mean_token_accuracy": 0.2612107306718826,
      "num_tokens": 74908697.0,
      "step": 32670
    },
    {
      "entropy": 4.987684202194214,
      "epoch": 3.1388088376561,
      "grad_norm": 1.0,
      "learning_rate": 0.0004030740186258819,
      "loss": 4.6502,
      "mean_token_accuracy": 0.24893118143081666,
      "num_tokens": 74919012.0,
      "step": 32675
    },
    {
      "entropy": 5.07190842628479,
      "epoch": 3.139289145052834,
      "grad_norm": 0.875,
      "learning_rate": 0.0004030458296546905,
      "loss": 4.701,
      "mean_token_accuracy": 0.24671801924705505,
      "num_tokens": 74931266.0,
      "step": 32680
    },
    {
      "entropy": 5.126935911178589,
      "epoch": 3.139769452449568,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040301763771065504,
      "loss": 4.7295,
      "mean_token_accuracy": 0.24547708332538604,
      "num_tokens": 74942860.0,
      "step": 32685
    },
    {
      "entropy": 5.047345590591431,
      "epoch": 3.1402497598463017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004029894427944302,
      "loss": 4.6065,
      "mean_token_accuracy": 0.2547300338745117,
      "num_tokens": 74955441.0,
      "step": 32690
    },
    {
      "entropy": 5.010250091552734,
      "epoch": 3.1407300672430356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040296124490667065,
      "loss": 4.6166,
      "mean_token_accuracy": 0.2528986126184464,
      "num_tokens": 74966830.0,
      "step": 32695
    },
    {
      "entropy": 5.1572239875793455,
      "epoch": 3.1412103746397695,
      "grad_norm": 1.0,
      "learning_rate": 0.0004029330440480308,
      "loss": 4.6889,
      "mean_token_accuracy": 0.24060844182968139,
      "num_tokens": 74978380.0,
      "step": 32700
    },
    {
      "entropy": 5.140653133392334,
      "epoch": 3.1416906820365034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004029048402191656,
      "loss": 4.7318,
      "mean_token_accuracy": 0.24110923409461976,
      "num_tokens": 74990072.0,
      "step": 32705
    },
    {
      "entropy": 5.0691001415252686,
      "epoch": 3.1421709894332372,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004028766334207299,
      "loss": 4.6661,
      "mean_token_accuracy": 0.24485287219285964,
      "num_tokens": 75000784.0,
      "step": 32710
    },
    {
      "entropy": 5.101446914672851,
      "epoch": 3.142651296829971,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004028484236533784,
      "loss": 4.6894,
      "mean_token_accuracy": 0.24176436960697173,
      "num_tokens": 75012709.0,
      "step": 32715
    },
    {
      "entropy": 5.0356689453125,
      "epoch": 3.143131604226705,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040282021091776624,
      "loss": 4.6091,
      "mean_token_accuracy": 0.2523089095950127,
      "num_tokens": 75023644.0,
      "step": 32720
    },
    {
      "entropy": 5.0615644454956055,
      "epoch": 3.143611911623439,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004027919952145482,
      "loss": 4.638,
      "mean_token_accuracy": 0.24897396862506865,
      "num_tokens": 75035706.0,
      "step": 32725
    },
    {
      "entropy": 5.142883825302124,
      "epoch": 3.1440922190201728,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004027637765443795,
      "loss": 4.6766,
      "mean_token_accuracy": 0.24526393711566924,
      "num_tokens": 75046447.0,
      "step": 32730
    },
    {
      "entropy": 5.054813718795776,
      "epoch": 3.1445725264169067,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040273555490791534,
      "loss": 4.6474,
      "mean_token_accuracy": 0.24661665111780168,
      "num_tokens": 75058515.0,
      "step": 32735
    },
    {
      "entropy": 5.101583862304688,
      "epoch": 3.1450528338136405,
      "grad_norm": 1.0,
      "learning_rate": 0.0004027073303058109,
      "loss": 4.6934,
      "mean_token_accuracy": 0.24538722783327102,
      "num_tokens": 75068638.0,
      "step": 32740
    },
    {
      "entropy": 5.15917067527771,
      "epoch": 3.1455331412103744,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004026791027387214,
      "loss": 4.7165,
      "mean_token_accuracy": 0.2382655307650566,
      "num_tokens": 75080711.0,
      "step": 32745
    },
    {
      "entropy": 5.227984380722046,
      "epoch": 3.1460134486071087,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004026508722073024,
      "loss": 4.7642,
      "mean_token_accuracy": 0.2419245347380638,
      "num_tokens": 75090337.0,
      "step": 32750
    },
    {
      "entropy": 5.077029275894165,
      "epoch": 3.1464937560038426,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040262263871220904,
      "loss": 4.6484,
      "mean_token_accuracy": 0.24441083818674086,
      "num_tokens": 75101446.0,
      "step": 32755
    },
    {
      "entropy": 5.030180406570435,
      "epoch": 3.1469740634005765,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004025944022540971,
      "loss": 4.6308,
      "mean_token_accuracy": 0.25184379070997237,
      "num_tokens": 75112070.0,
      "step": 32760
    },
    {
      "entropy": 5.067906904220581,
      "epoch": 3.1474543707973104,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040256616283362195,
      "loss": 4.6466,
      "mean_token_accuracy": 0.252499932050705,
      "num_tokens": 75123635.0,
      "step": 32765
    },
    {
      "entropy": 5.149288702011108,
      "epoch": 3.1479346781940443,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040253792045143926,
      "loss": 4.6993,
      "mean_token_accuracy": 0.2474423348903656,
      "num_tokens": 75135118.0,
      "step": 32770
    },
    {
      "entropy": 5.10075798034668,
      "epoch": 3.148414985590778,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004025096751082048,
      "loss": 4.7742,
      "mean_token_accuracy": 0.23826517015695572,
      "num_tokens": 75147319.0,
      "step": 32775
    },
    {
      "entropy": 5.0174072265625,
      "epoch": 3.148895292987512,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004024814268045743,
      "loss": 4.5625,
      "mean_token_accuracy": 0.26127906441688536,
      "num_tokens": 75158185.0,
      "step": 32780
    },
    {
      "entropy": 5.0645973682403564,
      "epoch": 3.149375600384246,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040245317554120363,
      "loss": 4.5893,
      "mean_token_accuracy": 0.2494758501648903,
      "num_tokens": 75169502.0,
      "step": 32785
    },
    {
      "entropy": 5.0782966136932375,
      "epoch": 3.14985590778098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004024249213187487,
      "loss": 4.7269,
      "mean_token_accuracy": 0.23653388470411302,
      "num_tokens": 75180567.0,
      "step": 32790
    },
    {
      "entropy": 5.091277837753296,
      "epoch": 3.1503362151777137,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004023966641378655,
      "loss": 4.6603,
      "mean_token_accuracy": 0.2406967356801033,
      "num_tokens": 75191622.0,
      "step": 32795
    },
    {
      "entropy": 5.063860654830933,
      "epoch": 3.1508165225744476,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040236840399920996,
      "loss": 4.7128,
      "mean_token_accuracy": 0.23893865048885346,
      "num_tokens": 75203306.0,
      "step": 32800
    },
    {
      "entropy": 5.083482503890991,
      "epoch": 3.1512968299711814,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040234014090343833,
      "loss": 4.628,
      "mean_token_accuracy": 0.2469210609793663,
      "num_tokens": 75214471.0,
      "step": 32805
    },
    {
      "entropy": 5.137204551696778,
      "epoch": 3.1517771373679153,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004023118748512068,
      "loss": 4.7002,
      "mean_token_accuracy": 0.24707604497671126,
      "num_tokens": 75226657.0,
      "step": 32810
    },
    {
      "entropy": 5.031982326507569,
      "epoch": 3.152257444764649,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004022836058431715,
      "loss": 4.6468,
      "mean_token_accuracy": 0.25368584245443343,
      "num_tokens": 75237850.0,
      "step": 32815
    },
    {
      "entropy": 5.106040525436401,
      "epoch": 3.152737752161383,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040225533387998883,
      "loss": 4.6183,
      "mean_token_accuracy": 0.24920148104429246,
      "num_tokens": 75248604.0,
      "step": 32820
    },
    {
      "entropy": 5.123107099533081,
      "epoch": 3.1532180595581174,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004022270589623152,
      "loss": 4.6201,
      "mean_token_accuracy": 0.24693459123373032,
      "num_tokens": 75259745.0,
      "step": 32825
    },
    {
      "entropy": 5.131017065048217,
      "epoch": 3.1536983669548513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000402198781090807,
      "loss": 4.8347,
      "mean_token_accuracy": 0.23322611302137375,
      "num_tokens": 75272421.0,
      "step": 32830
    },
    {
      "entropy": 5.117351770401001,
      "epoch": 3.154178674351585,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004021705002661208,
      "loss": 4.715,
      "mean_token_accuracy": 0.23647909015417098,
      "num_tokens": 75283753.0,
      "step": 32835
    },
    {
      "entropy": 5.107088422775268,
      "epoch": 3.154658981748319,
      "grad_norm": 0.875,
      "learning_rate": 0.0004021422164889133,
      "loss": 4.6487,
      "mean_token_accuracy": 0.2431615799665451,
      "num_tokens": 75294732.0,
      "step": 32840
    },
    {
      "entropy": 5.034707260131836,
      "epoch": 3.155139289145053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000402113929759841,
      "loss": 4.6733,
      "mean_token_accuracy": 0.24563550800085068,
      "num_tokens": 75306707.0,
      "step": 32845
    },
    {
      "entropy": 5.033975267410279,
      "epoch": 3.155619596541787,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00040208564007956075,
      "loss": 4.6397,
      "mean_token_accuracy": 0.2488164037466049,
      "num_tokens": 75317663.0,
      "step": 32850
    },
    {
      "entropy": 5.074614858627319,
      "epoch": 3.1560999039385207,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004020573474487293,
      "loss": 4.6704,
      "mean_token_accuracy": 0.24556121826171876,
      "num_tokens": 75328700.0,
      "step": 32855
    },
    {
      "entropy": 5.1250951290130615,
      "epoch": 3.1565802113352546,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040202905186800347,
      "loss": 4.6765,
      "mean_token_accuracy": 0.2513749971985817,
      "num_tokens": 75340233.0,
      "step": 32860
    },
    {
      "entropy": 5.049603891372681,
      "epoch": 3.1570605187319885,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004020007533380403,
      "loss": 4.5888,
      "mean_token_accuracy": 0.255260144174099,
      "num_tokens": 75351595.0,
      "step": 32865
    },
    {
      "entropy": 5.102052688598633,
      "epoch": 3.1575408261287223,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004019724518594967,
      "loss": 4.6829,
      "mean_token_accuracy": 0.25028142631053923,
      "num_tokens": 75362240.0,
      "step": 32870
    },
    {
      "entropy": 5.031745719909668,
      "epoch": 3.158021133525456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004019441474330298,
      "loss": 4.6171,
      "mean_token_accuracy": 0.2550402611494064,
      "num_tokens": 75373210.0,
      "step": 32875
    },
    {
      "entropy": 5.103553676605225,
      "epoch": 3.15850144092219,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040191584005929684,
      "loss": 4.7029,
      "mean_token_accuracy": 0.2420038640499115,
      "num_tokens": 75386320.0,
      "step": 32880
    },
    {
      "entropy": 5.107150268554688,
      "epoch": 3.158981748318924,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004018875297389549,
      "loss": 4.6858,
      "mean_token_accuracy": 0.2431190714240074,
      "num_tokens": 75397729.0,
      "step": 32885
    },
    {
      "entropy": 5.071495056152344,
      "epoch": 3.159462055715658,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040185921647266126,
      "loss": 4.6159,
      "mean_token_accuracy": 0.2524034813046455,
      "num_tokens": 75408508.0,
      "step": 32890
    },
    {
      "entropy": 5.072680807113647,
      "epoch": 3.1599423631123917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040183090026107326,
      "loss": 4.6566,
      "mean_token_accuracy": 0.2488311469554901,
      "num_tokens": 75419960.0,
      "step": 32895
    },
    {
      "entropy": 5.083609580993652,
      "epoch": 3.160422670509126,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040180258110484847,
      "loss": 4.7062,
      "mean_token_accuracy": 0.24170590937137604,
      "num_tokens": 75431115.0,
      "step": 32900
    },
    {
      "entropy": 5.012141752243042,
      "epoch": 3.16090297790586,
      "grad_norm": 1.0,
      "learning_rate": 0.0004017742590046442,
      "loss": 4.59,
      "mean_token_accuracy": 0.25889546871185304,
      "num_tokens": 75441917.0,
      "step": 32905
    },
    {
      "entropy": 5.062848901748657,
      "epoch": 3.161383285302594,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040174593396111814,
      "loss": 4.7008,
      "mean_token_accuracy": 0.2520574316382408,
      "num_tokens": 75452514.0,
      "step": 32910
    },
    {
      "entropy": 5.226338529586792,
      "epoch": 3.1618635926993277,
      "grad_norm": 1.0,
      "learning_rate": 0.00040171760597492785,
      "loss": 4.7697,
      "mean_token_accuracy": 0.2404971241950989,
      "num_tokens": 75463511.0,
      "step": 32915
    },
    {
      "entropy": 5.080031490325927,
      "epoch": 3.1623439000960616,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040168927504673094,
      "loss": 4.6578,
      "mean_token_accuracy": 0.24157734513282775,
      "num_tokens": 75474595.0,
      "step": 32920
    },
    {
      "entropy": 5.123180150985718,
      "epoch": 3.1628242074927955,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004016609411771853,
      "loss": 4.6848,
      "mean_token_accuracy": 0.24243765324354172,
      "num_tokens": 75485639.0,
      "step": 32925
    },
    {
      "entropy": 5.121702241897583,
      "epoch": 3.1633045148895294,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040163260436694876,
      "loss": 4.6948,
      "mean_token_accuracy": 0.23780447840690613,
      "num_tokens": 75497280.0,
      "step": 32930
    },
    {
      "entropy": 5.134649610519409,
      "epoch": 3.1637848222862632,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004016042646166791,
      "loss": 4.6873,
      "mean_token_accuracy": 0.251137538254261,
      "num_tokens": 75509134.0,
      "step": 32935
    },
    {
      "entropy": 5.025428247451782,
      "epoch": 3.164265129682997,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004015759219270344,
      "loss": 4.6251,
      "mean_token_accuracy": 0.24812544584274293,
      "num_tokens": 75520166.0,
      "step": 32940
    },
    {
      "entropy": 5.032694005966187,
      "epoch": 3.164745437079731,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004015475762986726,
      "loss": 4.6173,
      "mean_token_accuracy": 0.2477185159921646,
      "num_tokens": 75530698.0,
      "step": 32945
    },
    {
      "entropy": 5.042399787902832,
      "epoch": 3.165225744476465,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040151922773225187,
      "loss": 4.7224,
      "mean_token_accuracy": 0.24196990877389907,
      "num_tokens": 75542232.0,
      "step": 32950
    },
    {
      "entropy": 5.086015367507935,
      "epoch": 3.1657060518731988,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004014908762284303,
      "loss": 4.6275,
      "mean_token_accuracy": 0.25065446496009824,
      "num_tokens": 75553650.0,
      "step": 32955
    },
    {
      "entropy": 5.114188003540039,
      "epoch": 3.1661863592699326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040146252178786633,
      "loss": 4.6624,
      "mean_token_accuracy": 0.2500924378633499,
      "num_tokens": 75565877.0,
      "step": 32960
    },
    {
      "entropy": 5.070130109786987,
      "epoch": 3.1666666666666665,
      "grad_norm": 1.015625,
      "learning_rate": 0.000401434164411218,
      "loss": 4.5413,
      "mean_token_accuracy": 0.26022217720746993,
      "num_tokens": 75578294.0,
      "step": 32965
    },
    {
      "entropy": 5.081899166107178,
      "epoch": 3.1671469740634004,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040140580409914385,
      "loss": 4.701,
      "mean_token_accuracy": 0.24417033344507216,
      "num_tokens": 75589228.0,
      "step": 32970
    },
    {
      "entropy": 5.079851150512695,
      "epoch": 3.1676272814601343,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040137744085230227,
      "loss": 4.7082,
      "mean_token_accuracy": 0.24570702463388444,
      "num_tokens": 75600443.0,
      "step": 32975
    },
    {
      "entropy": 5.156255054473877,
      "epoch": 3.168107588856868,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004013490746713518,
      "loss": 4.6379,
      "mean_token_accuracy": 0.24483694434165953,
      "num_tokens": 75612053.0,
      "step": 32980
    },
    {
      "entropy": 5.1414406299591064,
      "epoch": 3.1685878962536025,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040132070555695096,
      "loss": 4.7011,
      "mean_token_accuracy": 0.23629581034183503,
      "num_tokens": 75624084.0,
      "step": 32985
    },
    {
      "entropy": 5.1290970802307125,
      "epoch": 3.1690682036503364,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040129233350975847,
      "loss": 4.6993,
      "mean_token_accuracy": 0.24040416926145552,
      "num_tokens": 75634863.0,
      "step": 32990
    },
    {
      "entropy": 4.959068679809571,
      "epoch": 3.1695485110470702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040126395853043293,
      "loss": 4.5219,
      "mean_token_accuracy": 0.2596110373735428,
      "num_tokens": 75645194.0,
      "step": 32995
    },
    {
      "entropy": 5.01206407546997,
      "epoch": 3.170028818443804,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004012355806196332,
      "loss": 4.6085,
      "mean_token_accuracy": 0.25525314211845396,
      "num_tokens": 75657444.0,
      "step": 33000
    },
    {
      "epoch": 3.170028818443804,
      "eval_entropy": 4.90025024097876,
      "eval_loss": 4.823419570922852,
      "eval_mean_token_accuracy": 0.24615764812357915,
      "eval_num_tokens": 75657444.0,
      "eval_runtime": 26.491,
      "eval_samples_per_second": 1238.724,
      "eval_steps_per_second": 154.845,
      "step": 33000
    },
    {
      "entropy": 5.141807126998901,
      "epoch": 3.170509125840538,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040120719977801823,
      "loss": 4.7145,
      "mean_token_accuracy": 0.2426938906311989,
      "num_tokens": 75670649.0,
      "step": 33005
    },
    {
      "entropy": 5.144469785690307,
      "epoch": 3.170989433237272,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040117881600624676,
      "loss": 4.7134,
      "mean_token_accuracy": 0.2372470960021019,
      "num_tokens": 75682222.0,
      "step": 33010
    },
    {
      "entropy": 5.11884298324585,
      "epoch": 3.1714697406340058,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040115042930497787,
      "loss": 4.7943,
      "mean_token_accuracy": 0.23773314505815507,
      "num_tokens": 75694644.0,
      "step": 33015
    },
    {
      "entropy": 5.196475601196289,
      "epoch": 3.1719500480307397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040112203967487066,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23511813431978226,
      "num_tokens": 75705736.0,
      "step": 33020
    },
    {
      "entropy": 5.055683660507202,
      "epoch": 3.1724303554274735,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040109364711658416,
      "loss": 4.6301,
      "mean_token_accuracy": 0.2537052363157272,
      "num_tokens": 75716829.0,
      "step": 33025
    },
    {
      "entropy": 5.095101976394654,
      "epoch": 3.1729106628242074,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040106525163077756,
      "loss": 4.713,
      "mean_token_accuracy": 0.24553814977407457,
      "num_tokens": 75727304.0,
      "step": 33030
    },
    {
      "entropy": 5.056622886657715,
      "epoch": 3.1733909702209413,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004010368532181102,
      "loss": 4.6906,
      "mean_token_accuracy": 0.24569199830293656,
      "num_tokens": 75740346.0,
      "step": 33035
    },
    {
      "entropy": 5.089049434661865,
      "epoch": 3.173871277617675,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004010084518792413,
      "loss": 4.6594,
      "mean_token_accuracy": 0.2459271416068077,
      "num_tokens": 75751808.0,
      "step": 33040
    },
    {
      "entropy": 5.12053918838501,
      "epoch": 3.174351585014409,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040098004761483037,
      "loss": 4.6308,
      "mean_token_accuracy": 0.25295125097036364,
      "num_tokens": 75763635.0,
      "step": 33045
    },
    {
      "entropy": 5.232107639312744,
      "epoch": 3.174831892411143,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004009516404255368,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22956520915031434,
      "num_tokens": 75775869.0,
      "step": 33050
    },
    {
      "entropy": 5.055005407333374,
      "epoch": 3.175312199807877,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004009232303120202,
      "loss": 4.6813,
      "mean_token_accuracy": 0.24838326126337051,
      "num_tokens": 75787263.0,
      "step": 33055
    },
    {
      "entropy": 5.203339719772339,
      "epoch": 3.175792507204611,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040089481727494,
      "loss": 4.7271,
      "mean_token_accuracy": 0.23913052380084993,
      "num_tokens": 75800128.0,
      "step": 33060
    },
    {
      "entropy": 5.108438158035279,
      "epoch": 3.176272814601345,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000400866401314956,
      "loss": 4.669,
      "mean_token_accuracy": 0.2517004907131195,
      "num_tokens": 75812151.0,
      "step": 33065
    },
    {
      "entropy": 5.134654235839844,
      "epoch": 3.176753121998079,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040083798243272797,
      "loss": 4.7253,
      "mean_token_accuracy": 0.24192118495702744,
      "num_tokens": 75823416.0,
      "step": 33070
    },
    {
      "entropy": 5.106930685043335,
      "epoch": 3.177233429394813,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00040080956062891554,
      "loss": 4.6875,
      "mean_token_accuracy": 0.24201093018054962,
      "num_tokens": 75835225.0,
      "step": 33075
    },
    {
      "entropy": 5.104752063751221,
      "epoch": 3.1777137367915467,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040078113590417887,
      "loss": 4.7385,
      "mean_token_accuracy": 0.2397340252995491,
      "num_tokens": 75846386.0,
      "step": 33080
    },
    {
      "entropy": 5.1819798946380615,
      "epoch": 3.1781940441882806,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040075270825917753,
      "loss": 4.7312,
      "mean_token_accuracy": 0.24471127688884736,
      "num_tokens": 75857163.0,
      "step": 33085
    },
    {
      "entropy": 5.097684335708618,
      "epoch": 3.1786743515850144,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004007242776945718,
      "loss": 4.6223,
      "mean_token_accuracy": 0.2556991681456566,
      "num_tokens": 75868054.0,
      "step": 33090
    },
    {
      "entropy": 5.077585363388062,
      "epoch": 3.1791546589817483,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00040069584421102174,
      "loss": 4.7376,
      "mean_token_accuracy": 0.24238463640213012,
      "num_tokens": 75879323.0,
      "step": 33095
    },
    {
      "entropy": 5.075553035736084,
      "epoch": 3.179634966378482,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040066740780918725,
      "loss": 4.6742,
      "mean_token_accuracy": 0.24375557154417038,
      "num_tokens": 75891324.0,
      "step": 33100
    },
    {
      "entropy": 5.120955848693848,
      "epoch": 3.180115273775216,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004006389684897288,
      "loss": 4.7434,
      "mean_token_accuracy": 0.2384619951248169,
      "num_tokens": 75902630.0,
      "step": 33105
    },
    {
      "entropy": 5.098286104202271,
      "epoch": 3.18059558117195,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004006105262533066,
      "loss": 4.6817,
      "mean_token_accuracy": 0.24984826296567916,
      "num_tokens": 75914080.0,
      "step": 33110
    },
    {
      "entropy": 5.211499357223511,
      "epoch": 3.181075888568684,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004005820811005809,
      "loss": 4.7259,
      "mean_token_accuracy": 0.2433484375476837,
      "num_tokens": 75925418.0,
      "step": 33115
    },
    {
      "entropy": 5.081755685806274,
      "epoch": 3.1815561959654177,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040055363303221226,
      "loss": 4.6576,
      "mean_token_accuracy": 0.24909963458776474,
      "num_tokens": 75935450.0,
      "step": 33120
    },
    {
      "entropy": 4.995599222183228,
      "epoch": 3.1820365033621516,
      "grad_norm": 0.953125,
      "learning_rate": 0.000400525182048861,
      "loss": 4.6091,
      "mean_token_accuracy": 0.25128038674592973,
      "num_tokens": 75947198.0,
      "step": 33125
    },
    {
      "entropy": 5.0536572456359865,
      "epoch": 3.1825168107588855,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040049672815118775,
      "loss": 4.634,
      "mean_token_accuracy": 0.25237257927656176,
      "num_tokens": 75958303.0,
      "step": 33130
    },
    {
      "entropy": 5.0861546993255615,
      "epoch": 3.18299711815562,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040046827133985316,
      "loss": 4.6629,
      "mean_token_accuracy": 0.24600543230772018,
      "num_tokens": 75968590.0,
      "step": 33135
    },
    {
      "entropy": 5.075027084350586,
      "epoch": 3.1834774255523537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040043981161551784,
      "loss": 4.626,
      "mean_token_accuracy": 0.24617850333452224,
      "num_tokens": 75980518.0,
      "step": 33140
    },
    {
      "entropy": 5.044564294815063,
      "epoch": 3.1839577329490876,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004004113489788426,
      "loss": 4.7526,
      "mean_token_accuracy": 0.24929489195346832,
      "num_tokens": 75993160.0,
      "step": 33145
    },
    {
      "entropy": 4.994883584976196,
      "epoch": 3.1844380403458215,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040038288343048823,
      "loss": 4.5586,
      "mean_token_accuracy": 0.2603205740451813,
      "num_tokens": 76003651.0,
      "step": 33150
    },
    {
      "entropy": 5.1404444694519045,
      "epoch": 3.1849183477425553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040035441497111564,
      "loss": 4.6184,
      "mean_token_accuracy": 0.2517063349485397,
      "num_tokens": 76014604.0,
      "step": 33155
    },
    {
      "entropy": 5.047126436233521,
      "epoch": 3.185398655139289,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040032594360138576,
      "loss": 4.6234,
      "mean_token_accuracy": 0.25169131308794024,
      "num_tokens": 76025821.0,
      "step": 33160
    },
    {
      "entropy": 5.055217313766479,
      "epoch": 3.185878962536023,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004002974693219595,
      "loss": 4.7022,
      "mean_token_accuracy": 0.2409898594021797,
      "num_tokens": 76038294.0,
      "step": 33165
    },
    {
      "entropy": 5.03799614906311,
      "epoch": 3.186359269932757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040026899213349814,
      "loss": 4.6177,
      "mean_token_accuracy": 0.2599205791950226,
      "num_tokens": 76049557.0,
      "step": 33170
    },
    {
      "entropy": 5.006697082519532,
      "epoch": 3.186839577329491,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004002405120366628,
      "loss": 4.5325,
      "mean_token_accuracy": 0.2612086981534958,
      "num_tokens": 76060167.0,
      "step": 33175
    },
    {
      "entropy": 5.000286722183228,
      "epoch": 3.1873198847262247,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040021202903211454,
      "loss": 4.6382,
      "mean_token_accuracy": 0.2431085303425789,
      "num_tokens": 76071214.0,
      "step": 33180
    },
    {
      "entropy": 5.154953384399414,
      "epoch": 3.1878001921229586,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004001835431205149,
      "loss": 4.6996,
      "mean_token_accuracy": 0.24325444996356965,
      "num_tokens": 76082943.0,
      "step": 33185
    },
    {
      "entropy": 5.239356899261475,
      "epoch": 3.1882804995196925,
      "grad_norm": 1.0,
      "learning_rate": 0.00040015505430252506,
      "loss": 4.806,
      "mean_token_accuracy": 0.23713841736316682,
      "num_tokens": 76095298.0,
      "step": 33190
    },
    {
      "entropy": 5.206976652145386,
      "epoch": 3.1887608069164264,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040012656257880645,
      "loss": 4.8355,
      "mean_token_accuracy": 0.22923275232315063,
      "num_tokens": 76106260.0,
      "step": 33195
    },
    {
      "entropy": 5.041665744781494,
      "epoch": 3.1892411143131603,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040009806795002076,
      "loss": 4.6384,
      "mean_token_accuracy": 0.24793130308389663,
      "num_tokens": 76117655.0,
      "step": 33200
    },
    {
      "entropy": 5.048253679275513,
      "epoch": 3.189721421709894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004000695704168292,
      "loss": 4.6768,
      "mean_token_accuracy": 0.24945951253175735,
      "num_tokens": 76128553.0,
      "step": 33205
    },
    {
      "entropy": 5.114042615890503,
      "epoch": 3.1902017291066285,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004000410699798937,
      "loss": 4.6803,
      "mean_token_accuracy": 0.24873627126216888,
      "num_tokens": 76138060.0,
      "step": 33210
    },
    {
      "entropy": 5.1205408573150635,
      "epoch": 3.1906820365033624,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00040001256663987585,
      "loss": 4.6636,
      "mean_token_accuracy": 0.2493976339697838,
      "num_tokens": 76150287.0,
      "step": 33215
    },
    {
      "entropy": 5.0660097122192385,
      "epoch": 3.1911623439000962,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039998406039743736,
      "loss": 4.6605,
      "mean_token_accuracy": 0.24754346311092376,
      "num_tokens": 76162940.0,
      "step": 33220
    },
    {
      "entropy": 5.132634210586548,
      "epoch": 3.19164265129683,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003999555512532401,
      "loss": 4.7464,
      "mean_token_accuracy": 0.24161824882030486,
      "num_tokens": 76174620.0,
      "step": 33225
    },
    {
      "entropy": 5.100732755661011,
      "epoch": 3.192122958693564,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003999270392079461,
      "loss": 4.7162,
      "mean_token_accuracy": 0.24494210332632066,
      "num_tokens": 76186499.0,
      "step": 33230
    },
    {
      "entropy": 5.1230292320251465,
      "epoch": 3.192603266090298,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003998985242622171,
      "loss": 4.6516,
      "mean_token_accuracy": 0.24661037027835847,
      "num_tokens": 76197752.0,
      "step": 33235
    },
    {
      "entropy": 5.050289869308472,
      "epoch": 3.1930835734870318,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003998700064167153,
      "loss": 4.6691,
      "mean_token_accuracy": 0.24695138335227967,
      "num_tokens": 76208796.0,
      "step": 33240
    },
    {
      "entropy": 5.045623874664306,
      "epoch": 3.1935638808837656,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003998414856721027,
      "loss": 4.5585,
      "mean_token_accuracy": 0.25555354803800584,
      "num_tokens": 76219605.0,
      "step": 33245
    },
    {
      "entropy": 5.1261707782745365,
      "epoch": 3.1940441882804995,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003998129620290415,
      "loss": 4.733,
      "mean_token_accuracy": 0.24242354929447174,
      "num_tokens": 76230088.0,
      "step": 33250
    },
    {
      "entropy": 5.197114896774292,
      "epoch": 3.1945244956772334,
      "grad_norm": 0.90625,
      "learning_rate": 0.00039978443548819393,
      "loss": 4.7798,
      "mean_token_accuracy": 0.2348928853869438,
      "num_tokens": 76243039.0,
      "step": 33255
    },
    {
      "entropy": 5.118587684631348,
      "epoch": 3.1950048030739673,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003997559060502224,
      "loss": 4.6248,
      "mean_token_accuracy": 0.2463693454861641,
      "num_tokens": 76255324.0,
      "step": 33260
    },
    {
      "entropy": 5.063439846038818,
      "epoch": 3.195485110470701,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003997273737157891,
      "loss": 4.5914,
      "mean_token_accuracy": 0.2514140114188194,
      "num_tokens": 76266054.0,
      "step": 33265
    },
    {
      "entropy": 4.995844030380249,
      "epoch": 3.195965417867435,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039969883848555647,
      "loss": 4.5748,
      "mean_token_accuracy": 0.25117960572242737,
      "num_tokens": 76277212.0,
      "step": 33270
    },
    {
      "entropy": 5.094616794586182,
      "epoch": 3.196445725264169,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003996703003601872,
      "loss": 4.6975,
      "mean_token_accuracy": 0.24443210512399674,
      "num_tokens": 76288707.0,
      "step": 33275
    },
    {
      "entropy": 5.14540810585022,
      "epoch": 3.196926032660903,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039964175934034375,
      "loss": 4.7218,
      "mean_token_accuracy": 0.2444423496723175,
      "num_tokens": 76300628.0,
      "step": 33280
    },
    {
      "entropy": 5.148494529724121,
      "epoch": 3.1974063400576367,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003996132154266887,
      "loss": 4.7501,
      "mean_token_accuracy": 0.23234175890684128,
      "num_tokens": 76314025.0,
      "step": 33285
    },
    {
      "entropy": 5.102265167236328,
      "epoch": 3.1978866474543706,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003995846686198849,
      "loss": 4.6735,
      "mean_token_accuracy": 0.24648328870534897,
      "num_tokens": 76326053.0,
      "step": 33290
    },
    {
      "entropy": 5.162123775482177,
      "epoch": 3.198366954851105,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003995561189205949,
      "loss": 4.7302,
      "mean_token_accuracy": 0.24104377627372742,
      "num_tokens": 76336738.0,
      "step": 33295
    },
    {
      "entropy": 5.0670849800109865,
      "epoch": 3.1988472622478388,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003995275663294818,
      "loss": 4.6695,
      "mean_token_accuracy": 0.2402460142970085,
      "num_tokens": 76347848.0,
      "step": 33300
    },
    {
      "entropy": 5.0983888626098635,
      "epoch": 3.1993275696445727,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003994990108472084,
      "loss": 4.7333,
      "mean_token_accuracy": 0.24132407009601592,
      "num_tokens": 76359526.0,
      "step": 33305
    },
    {
      "entropy": 5.1426183700561525,
      "epoch": 3.1998078770413065,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039947045247443755,
      "loss": 4.6431,
      "mean_token_accuracy": 0.25180455297231674,
      "num_tokens": 76370252.0,
      "step": 33310
    },
    {
      "entropy": 5.203094387054444,
      "epoch": 3.2002881844380404,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039944189121183247,
      "loss": 4.7463,
      "mean_token_accuracy": 0.2385396122932434,
      "num_tokens": 76381655.0,
      "step": 33315
    },
    {
      "entropy": 5.057973194122314,
      "epoch": 3.2007684918347743,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039941332706005617,
      "loss": 4.673,
      "mean_token_accuracy": 0.24321062117815018,
      "num_tokens": 76392489.0,
      "step": 33320
    },
    {
      "entropy": 5.096010398864746,
      "epoch": 3.201248799231508,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039938476001977175,
      "loss": 4.6965,
      "mean_token_accuracy": 0.24710288494825364,
      "num_tokens": 76403959.0,
      "step": 33325
    },
    {
      "entropy": 5.09407148361206,
      "epoch": 3.201729106628242,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039935619009164264,
      "loss": 4.6041,
      "mean_token_accuracy": 0.25010435581207274,
      "num_tokens": 76414783.0,
      "step": 33330
    },
    {
      "entropy": 5.130188751220703,
      "epoch": 3.202209414024976,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003993276172763321,
      "loss": 4.6028,
      "mean_token_accuracy": 0.25669704377651215,
      "num_tokens": 76426873.0,
      "step": 33335
    },
    {
      "entropy": 5.071373510360718,
      "epoch": 3.20268972142171,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039929904157450343,
      "loss": 4.63,
      "mean_token_accuracy": 0.250645224750042,
      "num_tokens": 76436821.0,
      "step": 33340
    },
    {
      "entropy": 5.065137815475464,
      "epoch": 3.2031700288184437,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039927046298682007,
      "loss": 4.7263,
      "mean_token_accuracy": 0.24606222212314605,
      "num_tokens": 76449083.0,
      "step": 33345
    },
    {
      "entropy": 5.096722221374511,
      "epoch": 3.2036503362151776,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003992418815139456,
      "loss": 4.5887,
      "mean_token_accuracy": 0.2543162390589714,
      "num_tokens": 76460489.0,
      "step": 33350
    },
    {
      "entropy": 5.091054916381836,
      "epoch": 3.2041306436119115,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039921329715654355,
      "loss": 4.6951,
      "mean_token_accuracy": 0.2443339303135872,
      "num_tokens": 76472553.0,
      "step": 33355
    },
    {
      "entropy": 5.130966854095459,
      "epoch": 3.2046109510086453,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003991847099152775,
      "loss": 4.719,
      "mean_token_accuracy": 0.23975479304790498,
      "num_tokens": 76484348.0,
      "step": 33360
    },
    {
      "entropy": 5.140629720687866,
      "epoch": 3.2050912584053792,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003991561197908114,
      "loss": 4.6794,
      "mean_token_accuracy": 0.24721895009279252,
      "num_tokens": 76496157.0,
      "step": 33365
    },
    {
      "entropy": 5.113038015365601,
      "epoch": 3.2055715658021136,
      "grad_norm": 1.0,
      "learning_rate": 0.0003991275267838088,
      "loss": 4.7539,
      "mean_token_accuracy": 0.23485117405653,
      "num_tokens": 76507924.0,
      "step": 33370
    },
    {
      "entropy": 5.064083576202393,
      "epoch": 3.2060518731988474,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039909893089493353,
      "loss": 4.636,
      "mean_token_accuracy": 0.2560821622610092,
      "num_tokens": 76519542.0,
      "step": 33375
    },
    {
      "entropy": 5.009941577911377,
      "epoch": 3.2065321805955813,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039907033212484966,
      "loss": 4.5515,
      "mean_token_accuracy": 0.2573045343160629,
      "num_tokens": 76532248.0,
      "step": 33380
    },
    {
      "entropy": 5.10320291519165,
      "epoch": 3.207012487992315,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003990417304742211,
      "loss": 4.6784,
      "mean_token_accuracy": 0.2470979794859886,
      "num_tokens": 76544726.0,
      "step": 33385
    },
    {
      "entropy": 5.192077255249023,
      "epoch": 3.207492795389049,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003990131259437119,
      "loss": 4.7602,
      "mean_token_accuracy": 0.2377001389861107,
      "num_tokens": 76557525.0,
      "step": 33390
    },
    {
      "entropy": 5.112313890457154,
      "epoch": 3.207973102785783,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003989845185339861,
      "loss": 4.6964,
      "mean_token_accuracy": 0.24188895523548126,
      "num_tokens": 76569487.0,
      "step": 33395
    },
    {
      "entropy": 5.137678098678589,
      "epoch": 3.208453410182517,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003989559082457079,
      "loss": 4.6923,
      "mean_token_accuracy": 0.2426787093281746,
      "num_tokens": 76581289.0,
      "step": 33400
    },
    {
      "entropy": 5.071100568771362,
      "epoch": 3.2089337175792507,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039892729507954173,
      "loss": 4.57,
      "mean_token_accuracy": 0.2558321446180344,
      "num_tokens": 76592542.0,
      "step": 33405
    },
    {
      "entropy": 5.1090906143188475,
      "epoch": 3.2094140249759846,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039889867903615165,
      "loss": 4.7784,
      "mean_token_accuracy": 0.23141699135303498,
      "num_tokens": 76604264.0,
      "step": 33410
    },
    {
      "entropy": 5.182481336593628,
      "epoch": 3.2098943323727185,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039887006011620217,
      "loss": 4.8018,
      "mean_token_accuracy": 0.23889685571193695,
      "num_tokens": 76615203.0,
      "step": 33415
    },
    {
      "entropy": 5.070486736297608,
      "epoch": 3.2103746397694524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039884143832035775,
      "loss": 4.6386,
      "mean_token_accuracy": 0.24763466864824296,
      "num_tokens": 76626816.0,
      "step": 33420
    },
    {
      "entropy": 5.0651082515716555,
      "epoch": 3.2108549471661862,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003988128136492828,
      "loss": 4.6599,
      "mean_token_accuracy": 0.24385498017072677,
      "num_tokens": 76638126.0,
      "step": 33425
    },
    {
      "entropy": 5.114804077148437,
      "epoch": 3.21133525456292,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000398784186103642,
      "loss": 4.6234,
      "mean_token_accuracy": 0.24530576914548874,
      "num_tokens": 76648874.0,
      "step": 33430
    },
    {
      "entropy": 5.134537649154663,
      "epoch": 3.211815561959654,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039875555568409996,
      "loss": 4.662,
      "mean_token_accuracy": 0.2421793147921562,
      "num_tokens": 76659548.0,
      "step": 33435
    },
    {
      "entropy": 5.034884786605835,
      "epoch": 3.212295869356388,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003987269223913214,
      "loss": 4.6421,
      "mean_token_accuracy": 0.25084117501974107,
      "num_tokens": 76671430.0,
      "step": 33440
    },
    {
      "entropy": 5.048752212524414,
      "epoch": 3.212776176753122,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039869828622597105,
      "loss": 4.6914,
      "mean_token_accuracy": 0.24277795255184173,
      "num_tokens": 76684169.0,
      "step": 33445
    },
    {
      "entropy": 5.117694759368897,
      "epoch": 3.213256484149856,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039866964718871385,
      "loss": 4.6568,
      "mean_token_accuracy": 0.2462952494621277,
      "num_tokens": 76694029.0,
      "step": 33450
    },
    {
      "entropy": 5.161125707626343,
      "epoch": 3.21373679154659,
      "grad_norm": 1.0,
      "learning_rate": 0.0003986410052802146,
      "loss": 4.7565,
      "mean_token_accuracy": 0.24014824479818345,
      "num_tokens": 76704889.0,
      "step": 33455
    },
    {
      "entropy": 5.152541017532348,
      "epoch": 3.214217098943324,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039861236050113845,
      "loss": 4.7696,
      "mean_token_accuracy": 0.24292737394571304,
      "num_tokens": 76716157.0,
      "step": 33460
    },
    {
      "entropy": 5.1150062084198,
      "epoch": 3.2146974063400577,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003985837128521503,
      "loss": 4.6074,
      "mean_token_accuracy": 0.25371393412351606,
      "num_tokens": 76727592.0,
      "step": 33465
    },
    {
      "entropy": 5.066445827484131,
      "epoch": 3.2151777137367916,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003985550623339153,
      "loss": 4.6817,
      "mean_token_accuracy": 0.23962051272392274,
      "num_tokens": 76740215.0,
      "step": 33470
    },
    {
      "entropy": 5.189021492004395,
      "epoch": 3.2156580211335255,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003985264089470987,
      "loss": 4.7846,
      "mean_token_accuracy": 0.23537371903657914,
      "num_tokens": 76751887.0,
      "step": 33475
    },
    {
      "entropy": 5.062526988983154,
      "epoch": 3.2161383285302594,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039849775269236556,
      "loss": 4.6299,
      "mean_token_accuracy": 0.24845572263002397,
      "num_tokens": 76762969.0,
      "step": 33480
    },
    {
      "entropy": 5.079005336761474,
      "epoch": 3.2166186359269933,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039846909357038135,
      "loss": 4.6574,
      "mean_token_accuracy": 0.24266939610242844,
      "num_tokens": 76773203.0,
      "step": 33485
    },
    {
      "entropy": 5.175133466720581,
      "epoch": 3.217098943323727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003984404315818115,
      "loss": 4.7653,
      "mean_token_accuracy": 0.23889745473861695,
      "num_tokens": 76785416.0,
      "step": 33490
    },
    {
      "entropy": 5.087816762924194,
      "epoch": 3.217579250720461,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039841176672732127,
      "loss": 4.6839,
      "mean_token_accuracy": 0.24851286858320237,
      "num_tokens": 76797875.0,
      "step": 33495
    },
    {
      "entropy": 5.0979687690734865,
      "epoch": 3.218059558117195,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003983830990075763,
      "loss": 4.6981,
      "mean_token_accuracy": 0.24648612439632417,
      "num_tokens": 76808937.0,
      "step": 33500
    },
    {
      "entropy": 5.134464693069458,
      "epoch": 3.218539865513929,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039835442842324216,
      "loss": 4.6708,
      "mean_token_accuracy": 0.24796090871095658,
      "num_tokens": 76820616.0,
      "step": 33505
    },
    {
      "entropy": 5.118542146682739,
      "epoch": 3.2190201729106627,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039832575497498454,
      "loss": 4.7355,
      "mean_token_accuracy": 0.2428065076470375,
      "num_tokens": 76832589.0,
      "step": 33510
    },
    {
      "entropy": 5.031345510482788,
      "epoch": 3.2195004803073966,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039829707866346895,
      "loss": 4.6006,
      "mean_token_accuracy": 0.25125529170036315,
      "num_tokens": 76844425.0,
      "step": 33515
    },
    {
      "entropy": 5.046614503860473,
      "epoch": 3.219980787704131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003982683994893614,
      "loss": 4.6405,
      "mean_token_accuracy": 0.24919351637363435,
      "num_tokens": 76856140.0,
      "step": 33520
    },
    {
      "entropy": 5.096798133850098,
      "epoch": 3.2204610951008648,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039823971745332764,
      "loss": 4.6812,
      "mean_token_accuracy": 0.2499474033713341,
      "num_tokens": 76868423.0,
      "step": 33525
    },
    {
      "entropy": 5.2181943416595455,
      "epoch": 3.2209414024975986,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003982110325560336,
      "loss": 4.772,
      "mean_token_accuracy": 0.2378716066479683,
      "num_tokens": 76879817.0,
      "step": 33530
    },
    {
      "entropy": 5.179190397262573,
      "epoch": 3.2214217098943325,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003981823447981453,
      "loss": 4.7231,
      "mean_token_accuracy": 0.2435617998242378,
      "num_tokens": 76892679.0,
      "step": 33535
    },
    {
      "entropy": 5.1111479759216305,
      "epoch": 3.2219020172910664,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039815365418032855,
      "loss": 4.6882,
      "mean_token_accuracy": 0.24665078222751619,
      "num_tokens": 76904563.0,
      "step": 33540
    },
    {
      "entropy": 5.117777395248413,
      "epoch": 3.2223823246878003,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039812496070324983,
      "loss": 4.7068,
      "mean_token_accuracy": 0.24908678978681564,
      "num_tokens": 76917708.0,
      "step": 33545
    },
    {
      "entropy": 5.107707071304321,
      "epoch": 3.222862632084534,
      "grad_norm": 0.921875,
      "learning_rate": 0.000398096264367575,
      "loss": 4.6854,
      "mean_token_accuracy": 0.2504700481891632,
      "num_tokens": 76929254.0,
      "step": 33550
    },
    {
      "entropy": 5.017084980010987,
      "epoch": 3.223342939481268,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003980675651739705,
      "loss": 4.6549,
      "mean_token_accuracy": 0.2510343000292778,
      "num_tokens": 76940320.0,
      "step": 33555
    },
    {
      "entropy": 5.066117572784424,
      "epoch": 3.223823246878002,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039803886312310253,
      "loss": 4.6128,
      "mean_token_accuracy": 0.24666896015405654,
      "num_tokens": 76952403.0,
      "step": 33560
    },
    {
      "entropy": 5.093769788742065,
      "epoch": 3.224303554274736,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039801015821563755,
      "loss": 4.6989,
      "mean_token_accuracy": 0.25204372107982637,
      "num_tokens": 76963690.0,
      "step": 33565
    },
    {
      "entropy": 5.1541379451751705,
      "epoch": 3.2247838616714697,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003979814504522419,
      "loss": 4.6855,
      "mean_token_accuracy": 0.2436726927757263,
      "num_tokens": 76974451.0,
      "step": 33570
    },
    {
      "entropy": 5.181416416168213,
      "epoch": 3.2252641690682036,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039795273983358223,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23880098164081573,
      "num_tokens": 76986308.0,
      "step": 33575
    },
    {
      "entropy": 5.128559017181397,
      "epoch": 3.2257444764649374,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039792402636032497,
      "loss": 4.7338,
      "mean_token_accuracy": 0.24017376601696014,
      "num_tokens": 76999626.0,
      "step": 33580
    },
    {
      "entropy": 5.168804883956909,
      "epoch": 3.2262247838616713,
      "grad_norm": 0.84375,
      "learning_rate": 0.00039789531003313696,
      "loss": 4.7467,
      "mean_token_accuracy": 0.24260976910591125,
      "num_tokens": 77011461.0,
      "step": 33585
    },
    {
      "entropy": 5.1362062931060795,
      "epoch": 3.226705091258405,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003978665908526846,
      "loss": 4.6873,
      "mean_token_accuracy": 0.24237828105688095,
      "num_tokens": 77022684.0,
      "step": 33590
    },
    {
      "entropy": 5.08912205696106,
      "epoch": 3.227185398655139,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003978378688196349,
      "loss": 4.666,
      "mean_token_accuracy": 0.24796006232500076,
      "num_tokens": 77034827.0,
      "step": 33595
    },
    {
      "entropy": 5.040789604187012,
      "epoch": 3.227665706051873,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003978091439346546,
      "loss": 4.6473,
      "mean_token_accuracy": 0.2517272591590881,
      "num_tokens": 77047446.0,
      "step": 33600
    },
    {
      "entropy": 5.0692168235778805,
      "epoch": 3.2281460134486073,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003977804161984108,
      "loss": 4.5856,
      "mean_token_accuracy": 0.2541586130857468,
      "num_tokens": 77057907.0,
      "step": 33605
    },
    {
      "entropy": 5.008700799942017,
      "epoch": 3.228626320845341,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003977516856115702,
      "loss": 4.6044,
      "mean_token_accuracy": 0.25199479311704637,
      "num_tokens": 77069276.0,
      "step": 33610
    },
    {
      "entropy": 5.094367837905883,
      "epoch": 3.229106628242075,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039772295217479993,
      "loss": 4.6798,
      "mean_token_accuracy": 0.24438059628009795,
      "num_tokens": 77081379.0,
      "step": 33615
    },
    {
      "entropy": 5.032278347015381,
      "epoch": 3.229586935638809,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003976942158887671,
      "loss": 4.6748,
      "mean_token_accuracy": 0.252716463804245,
      "num_tokens": 77092220.0,
      "step": 33620
    },
    {
      "entropy": 5.14553165435791,
      "epoch": 3.230067243035543,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000397665476754139,
      "loss": 4.7048,
      "mean_token_accuracy": 0.24727760553359984,
      "num_tokens": 77102658.0,
      "step": 33625
    },
    {
      "entropy": 5.17089581489563,
      "epoch": 3.2305475504322767,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003976367347715828,
      "loss": 4.8242,
      "mean_token_accuracy": 0.2297576829791069,
      "num_tokens": 77114907.0,
      "step": 33630
    },
    {
      "entropy": 5.1287391662597654,
      "epoch": 3.2310278578290106,
      "grad_norm": 1.0,
      "learning_rate": 0.0003976079899417657,
      "loss": 4.7266,
      "mean_token_accuracy": 0.24637484699487686,
      "num_tokens": 77126879.0,
      "step": 33635
    },
    {
      "entropy": 5.218905115127564,
      "epoch": 3.2315081652257445,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003975792422653552,
      "loss": 4.7573,
      "mean_token_accuracy": 0.23678061962127686,
      "num_tokens": 77139428.0,
      "step": 33640
    },
    {
      "entropy": 5.116429805755615,
      "epoch": 3.2319884726224783,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003975504917430186,
      "loss": 4.7433,
      "mean_token_accuracy": 0.23789792209863664,
      "num_tokens": 77151409.0,
      "step": 33645
    },
    {
      "entropy": 5.169024658203125,
      "epoch": 3.2324687800192122,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003975217383754235,
      "loss": 4.7,
      "mean_token_accuracy": 0.2405163899064064,
      "num_tokens": 77164112.0,
      "step": 33650
    },
    {
      "entropy": 5.095850849151612,
      "epoch": 3.232949087415946,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003974929821632375,
      "loss": 4.6778,
      "mean_token_accuracy": 0.2501856192946434,
      "num_tokens": 77175350.0,
      "step": 33655
    },
    {
      "entropy": 5.054169321060181,
      "epoch": 3.23342939481268,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039746422310712814,
      "loss": 4.6952,
      "mean_token_accuracy": 0.24621298015117646,
      "num_tokens": 77187034.0,
      "step": 33660
    },
    {
      "entropy": 5.038189840316773,
      "epoch": 3.233909702209414,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003974354612077632,
      "loss": 4.627,
      "mean_token_accuracy": 0.24697300791740417,
      "num_tokens": 77199160.0,
      "step": 33665
    },
    {
      "entropy": 5.100728702545166,
      "epoch": 3.2343900096061478,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003974066964658104,
      "loss": 4.6464,
      "mean_token_accuracy": 0.2510615006089211,
      "num_tokens": 77211507.0,
      "step": 33670
    },
    {
      "entropy": 5.112456846237182,
      "epoch": 3.2348703170028816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039737792888193754,
      "loss": 4.6228,
      "mean_token_accuracy": 0.24240693747997283,
      "num_tokens": 77223418.0,
      "step": 33675
    },
    {
      "entropy": 5.075006628036499,
      "epoch": 3.235350624399616,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003973491584568126,
      "loss": 4.6849,
      "mean_token_accuracy": 0.24108828455209733,
      "num_tokens": 77235418.0,
      "step": 33680
    },
    {
      "entropy": 5.058162021636963,
      "epoch": 3.23583093179635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003973203851911035,
      "loss": 4.6921,
      "mean_token_accuracy": 0.23982137441635132,
      "num_tokens": 77247443.0,
      "step": 33685
    },
    {
      "entropy": 4.990799808502198,
      "epoch": 3.2363112391930837,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003972916090854782,
      "loss": 4.5686,
      "mean_token_accuracy": 0.25517975240945817,
      "num_tokens": 77259092.0,
      "step": 33690
    },
    {
      "entropy": 5.040581226348877,
      "epoch": 3.2367915465898176,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039726283014060497,
      "loss": 4.6359,
      "mean_token_accuracy": 0.2456675335764885,
      "num_tokens": 77270530.0,
      "step": 33695
    },
    {
      "entropy": 5.062804555892944,
      "epoch": 3.2372718539865515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003972340483571519,
      "loss": 4.5663,
      "mean_token_accuracy": 0.2542583286762238,
      "num_tokens": 77281553.0,
      "step": 33700
    },
    {
      "entropy": 4.98065414428711,
      "epoch": 3.2377521613832854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039720526373578704,
      "loss": 4.6393,
      "mean_token_accuracy": 0.2444194957613945,
      "num_tokens": 77293799.0,
      "step": 33705
    },
    {
      "entropy": 5.123990488052368,
      "epoch": 3.2382324687800192,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039717647627717894,
      "loss": 4.7123,
      "mean_token_accuracy": 0.2428125873208046,
      "num_tokens": 77304646.0,
      "step": 33710
    },
    {
      "entropy": 5.029402208328247,
      "epoch": 3.238712776176753,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003971476859819958,
      "loss": 4.6211,
      "mean_token_accuracy": 0.2503040701150894,
      "num_tokens": 77316270.0,
      "step": 33715
    },
    {
      "entropy": 5.072261905670166,
      "epoch": 3.239193083573487,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003971188928509062,
      "loss": 4.6825,
      "mean_token_accuracy": 0.2492876872420311,
      "num_tokens": 77327534.0,
      "step": 33720
    },
    {
      "entropy": 5.041183185577393,
      "epoch": 3.239673390970221,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003970900968845784,
      "loss": 4.4953,
      "mean_token_accuracy": 0.2652922585606575,
      "num_tokens": 77337628.0,
      "step": 33725
    },
    {
      "entropy": 5.098190021514893,
      "epoch": 3.2401536983669548,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039706129808368115,
      "loss": 4.6836,
      "mean_token_accuracy": 0.24807690382003783,
      "num_tokens": 77348777.0,
      "step": 33730
    },
    {
      "entropy": 5.104367256164551,
      "epoch": 3.2406340057636887,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003970324964488829,
      "loss": 4.6534,
      "mean_token_accuracy": 0.253289233148098,
      "num_tokens": 77359954.0,
      "step": 33735
    },
    {
      "entropy": 5.107568216323853,
      "epoch": 3.2411143131604225,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039700369198085255,
      "loss": 4.6868,
      "mean_token_accuracy": 0.24186179786920547,
      "num_tokens": 77370850.0,
      "step": 33740
    },
    {
      "entropy": 5.044596815109253,
      "epoch": 3.2415946205571564,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039697488468025876,
      "loss": 4.6631,
      "mean_token_accuracy": 0.24048286974430083,
      "num_tokens": 77381607.0,
      "step": 33745
    },
    {
      "entropy": 5.112248277664184,
      "epoch": 3.2420749279538903,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003969460745477703,
      "loss": 4.6778,
      "mean_token_accuracy": 0.2506404623389244,
      "num_tokens": 77393310.0,
      "step": 33750
    },
    {
      "entropy": 5.0687840461730955,
      "epoch": 3.2425552353506246,
      "grad_norm": 1.0,
      "learning_rate": 0.00039691726158405606,
      "loss": 4.614,
      "mean_token_accuracy": 0.24224845319986343,
      "num_tokens": 77403952.0,
      "step": 33755
    },
    {
      "entropy": 5.060126161575317,
      "epoch": 3.2430355427473585,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039688844578978516,
      "loss": 4.6345,
      "mean_token_accuracy": 0.23959431946277618,
      "num_tokens": 77416541.0,
      "step": 33760
    },
    {
      "entropy": 5.1295154094696045,
      "epoch": 3.2435158501440924,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003968596271656263,
      "loss": 4.7346,
      "mean_token_accuracy": 0.24326175451278687,
      "num_tokens": 77428156.0,
      "step": 33765
    },
    {
      "entropy": 4.996503591537476,
      "epoch": 3.2439961575408263,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00039683080571224885,
      "loss": 4.6382,
      "mean_token_accuracy": 0.252539786696434,
      "num_tokens": 77439750.0,
      "step": 33770
    },
    {
      "entropy": 5.148586130142212,
      "epoch": 3.24447646493756,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003968019814303219,
      "loss": 4.7384,
      "mean_token_accuracy": 0.23945121318101883,
      "num_tokens": 77454480.0,
      "step": 33775
    },
    {
      "entropy": 5.0404987812042235,
      "epoch": 3.244956772334294,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003967731543205145,
      "loss": 4.6175,
      "mean_token_accuracy": 0.2490919843316078,
      "num_tokens": 77466487.0,
      "step": 33780
    },
    {
      "entropy": 5.080452966690063,
      "epoch": 3.245437079731028,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039674432438349607,
      "loss": 4.7378,
      "mean_token_accuracy": 0.23947930783033372,
      "num_tokens": 77477766.0,
      "step": 33785
    },
    {
      "entropy": 5.123367738723755,
      "epoch": 3.245917387127762,
      "grad_norm": 1.015625,
      "learning_rate": 0.000396715491619936,
      "loss": 4.5846,
      "mean_token_accuracy": 0.25204969346523287,
      "num_tokens": 77488753.0,
      "step": 33790
    },
    {
      "entropy": 5.052779483795166,
      "epoch": 3.2463976945244957,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003966866560305036,
      "loss": 4.6819,
      "mean_token_accuracy": 0.24733250439167023,
      "num_tokens": 77499183.0,
      "step": 33795
    },
    {
      "entropy": 5.0042037010192875,
      "epoch": 3.2468780019212296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039665781761586837,
      "loss": 4.6216,
      "mean_token_accuracy": 0.24595211446285248,
      "num_tokens": 77510465.0,
      "step": 33800
    },
    {
      "entropy": 5.124184274673462,
      "epoch": 3.2473583093179634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003966289763766999,
      "loss": 4.6608,
      "mean_token_accuracy": 0.24941142052412033,
      "num_tokens": 77521675.0,
      "step": 33805
    },
    {
      "entropy": 5.127190399169922,
      "epoch": 3.2478386167146973,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003966001323136678,
      "loss": 4.7246,
      "mean_token_accuracy": 0.24434682577848435,
      "num_tokens": 77534281.0,
      "step": 33810
    },
    {
      "entropy": 5.167156124114991,
      "epoch": 3.248318924111431,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003965712854274416,
      "loss": 4.6872,
      "mean_token_accuracy": 0.24430643767118454,
      "num_tokens": 77546393.0,
      "step": 33815
    },
    {
      "entropy": 5.191148710250855,
      "epoch": 3.248799231508165,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003965424357186912,
      "loss": 4.6933,
      "mean_token_accuracy": 0.24179457724094391,
      "num_tokens": 77556813.0,
      "step": 33820
    },
    {
      "entropy": 5.034583663940429,
      "epoch": 3.249279538904899,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003965135831880864,
      "loss": 4.5978,
      "mean_token_accuracy": 0.24819955229759216,
      "num_tokens": 77567846.0,
      "step": 33825
    },
    {
      "entropy": 5.156908845901489,
      "epoch": 3.249759846301633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000396484727836297,
      "loss": 4.7057,
      "mean_token_accuracy": 0.24375976771116256,
      "num_tokens": 77579867.0,
      "step": 33830
    },
    {
      "entropy": 5.060035371780396,
      "epoch": 3.2502401536983667,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003964558696639928,
      "loss": 4.5661,
      "mean_token_accuracy": 0.2583190411329269,
      "num_tokens": 77589984.0,
      "step": 33835
    },
    {
      "entropy": 5.023146724700927,
      "epoch": 3.250720461095101,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003964270086718441,
      "loss": 4.6945,
      "mean_token_accuracy": 0.2534104257822037,
      "num_tokens": 77602171.0,
      "step": 33840
    },
    {
      "entropy": 5.141227054595947,
      "epoch": 3.251200768491835,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039639814486052083,
      "loss": 4.734,
      "mean_token_accuracy": 0.23792777806520463,
      "num_tokens": 77613972.0,
      "step": 33845
    },
    {
      "entropy": 5.07128791809082,
      "epoch": 3.251681075888569,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003963692782306931,
      "loss": 4.6111,
      "mean_token_accuracy": 0.254303015768528,
      "num_tokens": 77626486.0,
      "step": 33850
    },
    {
      "entropy": 5.131126785278321,
      "epoch": 3.2521613832853027,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003963404087830311,
      "loss": 4.7405,
      "mean_token_accuracy": 0.24900150746107103,
      "num_tokens": 77636368.0,
      "step": 33855
    },
    {
      "entropy": 5.060461235046387,
      "epoch": 3.2526416906820366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003963115365182051,
      "loss": 4.7006,
      "mean_token_accuracy": 0.24434809684753417,
      "num_tokens": 77649256.0,
      "step": 33860
    },
    {
      "entropy": 5.080546236038208,
      "epoch": 3.2531219980787704,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039628266143688554,
      "loss": 4.5985,
      "mean_token_accuracy": 0.25400634557008744,
      "num_tokens": 77659370.0,
      "step": 33865
    },
    {
      "entropy": 5.029153490066529,
      "epoch": 3.2536023054755043,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003962537835397426,
      "loss": 4.646,
      "mean_token_accuracy": 0.24811754673719405,
      "num_tokens": 77670447.0,
      "step": 33870
    },
    {
      "entropy": 5.115053987503051,
      "epoch": 3.254082612872238,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039622490282744684,
      "loss": 4.748,
      "mean_token_accuracy": 0.24478249847888947,
      "num_tokens": 77682833.0,
      "step": 33875
    },
    {
      "entropy": 5.143673467636108,
      "epoch": 3.254562920268972,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0003961960193006689,
      "loss": 4.7378,
      "mean_token_accuracy": 0.23581675589084625,
      "num_tokens": 77697244.0,
      "step": 33880
    },
    {
      "entropy": 5.133462858200073,
      "epoch": 3.255043227665706,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003961671329600792,
      "loss": 4.6353,
      "mean_token_accuracy": 0.24653532952070237,
      "num_tokens": 77708323.0,
      "step": 33885
    },
    {
      "entropy": 5.052926540374756,
      "epoch": 3.25552353506244,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003961382438063485,
      "loss": 4.6961,
      "mean_token_accuracy": 0.24364175647497177,
      "num_tokens": 77719655.0,
      "step": 33890
    },
    {
      "entropy": 5.0862438678741455,
      "epoch": 3.2560038424591737,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003961093518401475,
      "loss": 4.6283,
      "mean_token_accuracy": 0.25304655730724335,
      "num_tokens": 77730598.0,
      "step": 33895
    },
    {
      "entropy": 5.136554145812989,
      "epoch": 3.2564841498559076,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039608045706214696,
      "loss": 4.677,
      "mean_token_accuracy": 0.24666533321142198,
      "num_tokens": 77740937.0,
      "step": 33900
    },
    {
      "entropy": 5.0897321701049805,
      "epoch": 3.2569644572526415,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003960515594730177,
      "loss": 4.7149,
      "mean_token_accuracy": 0.24662487357854843,
      "num_tokens": 77753227.0,
      "step": 33905
    },
    {
      "entropy": 5.0907152652740475,
      "epoch": 3.2574447646493754,
      "grad_norm": 1.125,
      "learning_rate": 0.0003960226590734307,
      "loss": 4.6386,
      "mean_token_accuracy": 0.24216471761465072,
      "num_tokens": 77764149.0,
      "step": 33910
    },
    {
      "entropy": 5.04188494682312,
      "epoch": 3.2579250720461097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000395993755864057,
      "loss": 4.6363,
      "mean_token_accuracy": 0.25090264081954955,
      "num_tokens": 77774635.0,
      "step": 33915
    },
    {
      "entropy": 5.133968639373779,
      "epoch": 3.2584053794428436,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003959648498455674,
      "loss": 4.7296,
      "mean_token_accuracy": 0.2487366035580635,
      "num_tokens": 77787544.0,
      "step": 33920
    },
    {
      "entropy": 5.127537107467651,
      "epoch": 3.2588856868395775,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039593594101863333,
      "loss": 4.6171,
      "mean_token_accuracy": 0.25009467601776125,
      "num_tokens": 77798175.0,
      "step": 33925
    },
    {
      "entropy": 5.092445135116577,
      "epoch": 3.2593659942363113,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039590702938392577,
      "loss": 4.6834,
      "mean_token_accuracy": 0.24298848658800126,
      "num_tokens": 77810105.0,
      "step": 33930
    },
    {
      "entropy": 5.100037479400635,
      "epoch": 3.2598463016330452,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039587811494211594,
      "loss": 4.6814,
      "mean_token_accuracy": 0.24872162491083144,
      "num_tokens": 77821124.0,
      "step": 33935
    },
    {
      "entropy": 5.154888486862182,
      "epoch": 3.260326609029779,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039584919769387536,
      "loss": 4.7439,
      "mean_token_accuracy": 0.23998252153396607,
      "num_tokens": 77831476.0,
      "step": 33940
    },
    {
      "entropy": 5.04706039428711,
      "epoch": 3.260806916426513,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003958202776398751,
      "loss": 4.5769,
      "mean_token_accuracy": 0.2544890567660332,
      "num_tokens": 77842152.0,
      "step": 33945
    },
    {
      "entropy": 5.0330126762390135,
      "epoch": 3.261287223823247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003957913547807868,
      "loss": 4.6247,
      "mean_token_accuracy": 0.24567372351884842,
      "num_tokens": 77852794.0,
      "step": 33950
    },
    {
      "entropy": 5.04334282875061,
      "epoch": 3.2617675312199808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000395762429117282,
      "loss": 4.6187,
      "mean_token_accuracy": 0.24943804889917373,
      "num_tokens": 77864942.0,
      "step": 33955
    },
    {
      "entropy": 5.158883762359619,
      "epoch": 3.2622478386167146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003957335006500321,
      "loss": 4.7639,
      "mean_token_accuracy": 0.24067910015583038,
      "num_tokens": 77876254.0,
      "step": 33960
    },
    {
      "entropy": 5.092933416366577,
      "epoch": 3.2627281460134485,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039570456937970883,
      "loss": 4.6188,
      "mean_token_accuracy": 0.2482569172978401,
      "num_tokens": 77886575.0,
      "step": 33965
    },
    {
      "entropy": 5.1262078285217285,
      "epoch": 3.2632084534101824,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003956756353069839,
      "loss": 4.6788,
      "mean_token_accuracy": 0.24467212408781053,
      "num_tokens": 77898351.0,
      "step": 33970
    },
    {
      "entropy": 5.021429014205933,
      "epoch": 3.2636887608069163,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003956466984325291,
      "loss": 4.6748,
      "mean_token_accuracy": 0.24064320027828218,
      "num_tokens": 77910972.0,
      "step": 33975
    },
    {
      "entropy": 5.042398405075073,
      "epoch": 3.26416906820365,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00039561775875701616,
      "loss": 4.596,
      "mean_token_accuracy": 0.2577166274189949,
      "num_tokens": 77922606.0,
      "step": 33980
    },
    {
      "entropy": 5.099759197235107,
      "epoch": 3.264649375600384,
      "grad_norm": 1.03125,
      "learning_rate": 0.000395588816281117,
      "loss": 4.6688,
      "mean_token_accuracy": 0.24707345217466353,
      "num_tokens": 77933442.0,
      "step": 33985
    },
    {
      "entropy": 5.004439973831177,
      "epoch": 3.2651296829971184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003955598710055036,
      "loss": 4.5895,
      "mean_token_accuracy": 0.25432740449905394,
      "num_tokens": 77945912.0,
      "step": 33990
    },
    {
      "entropy": 5.11930775642395,
      "epoch": 3.2656099903938522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000395530922930848,
      "loss": 4.7102,
      "mean_token_accuracy": 0.24382829517126084,
      "num_tokens": 77957953.0,
      "step": 33995
    },
    {
      "entropy": 5.155955696105957,
      "epoch": 3.266090297790586,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003955019720578223,
      "loss": 4.7492,
      "mean_token_accuracy": 0.24757628738880158,
      "num_tokens": 77969998.0,
      "step": 34000
    },
    {
      "entropy": 5.158969163894653,
      "epoch": 3.26657060518732,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003954730183870987,
      "loss": 4.7043,
      "mean_token_accuracy": 0.2495545893907547,
      "num_tokens": 77981119.0,
      "step": 34005
    },
    {
      "entropy": 5.048513126373291,
      "epoch": 3.267050912584054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003954440619193491,
      "loss": 4.5876,
      "mean_token_accuracy": 0.2569925621151924,
      "num_tokens": 77992748.0,
      "step": 34010
    },
    {
      "entropy": 5.046076059341431,
      "epoch": 3.2675312199807878,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039541510265524626,
      "loss": 4.6861,
      "mean_token_accuracy": 0.24388480484485625,
      "num_tokens": 78004540.0,
      "step": 34015
    },
    {
      "entropy": 5.092257022857666,
      "epoch": 3.2680115273775217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003953861405954622,
      "loss": 4.6253,
      "mean_token_accuracy": 0.25197093933820724,
      "num_tokens": 78015320.0,
      "step": 34020
    },
    {
      "entropy": 5.060213279724121,
      "epoch": 3.2684918347742555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003953571757406694,
      "loss": 4.6035,
      "mean_token_accuracy": 0.2523085534572601,
      "num_tokens": 78027126.0,
      "step": 34025
    },
    {
      "entropy": 5.008608865737915,
      "epoch": 3.2689721421709894,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039532820809154044,
      "loss": 4.6174,
      "mean_token_accuracy": 0.25505351573228835,
      "num_tokens": 78039742.0,
      "step": 34030
    },
    {
      "entropy": 5.073780870437622,
      "epoch": 3.2694524495677233,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039529923764874774,
      "loss": 4.6557,
      "mean_token_accuracy": 0.24996946901082992,
      "num_tokens": 78050658.0,
      "step": 34035
    },
    {
      "entropy": 5.0041032314300535,
      "epoch": 3.269932756964457,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000395270264412964,
      "loss": 4.5701,
      "mean_token_accuracy": 0.2615212380886078,
      "num_tokens": 78062476.0,
      "step": 34040
    },
    {
      "entropy": 5.103834009170532,
      "epoch": 3.270413064361191,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00039524128838486184,
      "loss": 4.7386,
      "mean_token_accuracy": 0.24681733697652816,
      "num_tokens": 78076173.0,
      "step": 34045
    },
    {
      "entropy": 5.117926692962646,
      "epoch": 3.270893371757925,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000395212309565114,
      "loss": 4.6665,
      "mean_token_accuracy": 0.24898930937051772,
      "num_tokens": 78086505.0,
      "step": 34050
    },
    {
      "entropy": 5.09682149887085,
      "epoch": 3.271373679154659,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039518332795439326,
      "loss": 4.5873,
      "mean_token_accuracy": 0.25562510043382647,
      "num_tokens": 78098026.0,
      "step": 34055
    },
    {
      "entropy": 5.078846788406372,
      "epoch": 3.2718539865513927,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039515434355337254,
      "loss": 4.6967,
      "mean_token_accuracy": 0.24175934344530106,
      "num_tokens": 78109484.0,
      "step": 34060
    },
    {
      "entropy": 5.105457782745361,
      "epoch": 3.272334293948127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003951253563627248,
      "loss": 4.7382,
      "mean_token_accuracy": 0.2413875937461853,
      "num_tokens": 78120833.0,
      "step": 34065
    },
    {
      "entropy": 5.077240705490112,
      "epoch": 3.2728146013448605,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003950963663831229,
      "loss": 4.6338,
      "mean_token_accuracy": 0.24928263276815416,
      "num_tokens": 78131775.0,
      "step": 34070
    },
    {
      "entropy": 5.060640811920166,
      "epoch": 3.273294908741595,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039506737361524007,
      "loss": 4.6387,
      "mean_token_accuracy": 0.2518847584724426,
      "num_tokens": 78143540.0,
      "step": 34075
    },
    {
      "entropy": 5.1213969707489015,
      "epoch": 3.2737752161383287,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00039503837805974926,
      "loss": 4.6933,
      "mean_token_accuracy": 0.24318451285362244,
      "num_tokens": 78155048.0,
      "step": 34080
    },
    {
      "entropy": 5.054323625564575,
      "epoch": 3.2742555235350626,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039500937971732376,
      "loss": 4.6372,
      "mean_token_accuracy": 0.24411402344703675,
      "num_tokens": 78165274.0,
      "step": 34085
    },
    {
      "entropy": 5.083052301406861,
      "epoch": 3.2747358309317964,
      "grad_norm": 1.0,
      "learning_rate": 0.0003949803785886369,
      "loss": 4.6664,
      "mean_token_accuracy": 0.2483305051922798,
      "num_tokens": 78176211.0,
      "step": 34090
    },
    {
      "entropy": 5.103652620315552,
      "epoch": 3.2752161383285303,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039495137467436184,
      "loss": 4.6043,
      "mean_token_accuracy": 0.2515538588166237,
      "num_tokens": 78186691.0,
      "step": 34095
    },
    {
      "entropy": 5.140513801574707,
      "epoch": 3.275696445725264,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039492236797517206,
      "loss": 4.6984,
      "mean_token_accuracy": 0.243371319770813,
      "num_tokens": 78197871.0,
      "step": 34100
    },
    {
      "entropy": 5.065258598327636,
      "epoch": 3.276176753121998,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000394893358491741,
      "loss": 4.6146,
      "mean_token_accuracy": 0.25362198799848557,
      "num_tokens": 78209220.0,
      "step": 34105
    },
    {
      "entropy": 5.10217776298523,
      "epoch": 3.276657060518732,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00039486434622474216,
      "loss": 4.6487,
      "mean_token_accuracy": 0.24734475314617158,
      "num_tokens": 78222289.0,
      "step": 34110
    },
    {
      "entropy": 5.169832372665406,
      "epoch": 3.277137367915466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039483533117484916,
      "loss": 4.7936,
      "mean_token_accuracy": 0.23847814798355102,
      "num_tokens": 78233243.0,
      "step": 34115
    },
    {
      "entropy": 5.103730392456055,
      "epoch": 3.2776176753121997,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003948063133427356,
      "loss": 4.6505,
      "mean_token_accuracy": 0.25145943760871886,
      "num_tokens": 78243781.0,
      "step": 34120
    },
    {
      "entropy": 5.071635913848877,
      "epoch": 3.2780979827089336,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003947772927290751,
      "loss": 4.6254,
      "mean_token_accuracy": 0.24588890969753266,
      "num_tokens": 78255746.0,
      "step": 34125
    },
    {
      "entropy": 5.095411491394043,
      "epoch": 3.2785782901056675,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003947482693345416,
      "loss": 4.6236,
      "mean_token_accuracy": 0.25302644073963165,
      "num_tokens": 78266614.0,
      "step": 34130
    },
    {
      "entropy": 5.084534645080566,
      "epoch": 3.2790585975024014,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039471924315980894,
      "loss": 4.6397,
      "mean_token_accuracy": 0.25245085805654527,
      "num_tokens": 78277279.0,
      "step": 34135
    },
    {
      "entropy": 5.0850293159484865,
      "epoch": 3.2795389048991357,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039469021420555087,
      "loss": 4.6528,
      "mean_token_accuracy": 0.24800242185592652,
      "num_tokens": 78288443.0,
      "step": 34140
    },
    {
      "entropy": 5.1024572372436525,
      "epoch": 3.280019212295869,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00039466118247244143,
      "loss": 4.7409,
      "mean_token_accuracy": 0.24166578352451323,
      "num_tokens": 78301082.0,
      "step": 34145
    },
    {
      "entropy": 5.1272765636444095,
      "epoch": 3.2804995196926034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003946321479611547,
      "loss": 4.7179,
      "mean_token_accuracy": 0.24032539427280425,
      "num_tokens": 78312368.0,
      "step": 34150
    },
    {
      "entropy": 5.0540999412536625,
      "epoch": 3.2809798270893373,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039460311067236465,
      "loss": 4.603,
      "mean_token_accuracy": 0.2505288526415825,
      "num_tokens": 78323671.0,
      "step": 34155
    },
    {
      "entropy": 5.060015630722046,
      "epoch": 3.281460134486071,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039457407060674557,
      "loss": 4.6281,
      "mean_token_accuracy": 0.25390284806489943,
      "num_tokens": 78333337.0,
      "step": 34160
    },
    {
      "entropy": 5.17574405670166,
      "epoch": 3.281940441882805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039454502776497163,
      "loss": 4.7564,
      "mean_token_accuracy": 0.24137173742055892,
      "num_tokens": 78344805.0,
      "step": 34165
    },
    {
      "entropy": 5.172603178024292,
      "epoch": 3.282420749279539,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039451598214771706,
      "loss": 4.7276,
      "mean_token_accuracy": 0.24335715621709825,
      "num_tokens": 78355736.0,
      "step": 34170
    },
    {
      "entropy": 5.07543158531189,
      "epoch": 3.282901056676273,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003944869337556563,
      "loss": 4.6688,
      "mean_token_accuracy": 0.24938067346811293,
      "num_tokens": 78367545.0,
      "step": 34175
    },
    {
      "entropy": 5.121845865249634,
      "epoch": 3.2833813640730067,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003944578825894639,
      "loss": 4.6725,
      "mean_token_accuracy": 0.24798091351985932,
      "num_tokens": 78378984.0,
      "step": 34180
    },
    {
      "entropy": 5.079128122329712,
      "epoch": 3.2838616714697406,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039442882864981397,
      "loss": 4.6836,
      "mean_token_accuracy": 0.24647755175828934,
      "num_tokens": 78389816.0,
      "step": 34185
    },
    {
      "entropy": 5.136640882492065,
      "epoch": 3.2843419788664745,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039439977193738134,
      "loss": 4.6519,
      "mean_token_accuracy": 0.24981094896793365,
      "num_tokens": 78401092.0,
      "step": 34190
    },
    {
      "entropy": 5.158543586730957,
      "epoch": 3.2848222862632084,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039437071245284055,
      "loss": 4.7266,
      "mean_token_accuracy": 0.24542734622955323,
      "num_tokens": 78412663.0,
      "step": 34195
    },
    {
      "entropy": 5.19345121383667,
      "epoch": 3.2853025936599423,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003943416501968663,
      "loss": 4.741,
      "mean_token_accuracy": 0.2444005236029625,
      "num_tokens": 78423059.0,
      "step": 34200
    },
    {
      "entropy": 5.134626579284668,
      "epoch": 3.285782901056676,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039431258517013323,
      "loss": 4.7472,
      "mean_token_accuracy": 0.24204034060239793,
      "num_tokens": 78434392.0,
      "step": 34205
    },
    {
      "entropy": 5.078365421295166,
      "epoch": 3.28626320845341,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003942835173733163,
      "loss": 4.7029,
      "mean_token_accuracy": 0.24384837448596955,
      "num_tokens": 78446311.0,
      "step": 34210
    },
    {
      "entropy": 5.100257682800293,
      "epoch": 3.286743515850144,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003942544468070903,
      "loss": 4.7074,
      "mean_token_accuracy": 0.2426380544900894,
      "num_tokens": 78458949.0,
      "step": 34215
    },
    {
      "entropy": 5.200331592559815,
      "epoch": 3.287223823246878,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003942253734721301,
      "loss": 4.7789,
      "mean_token_accuracy": 0.2336071401834488,
      "num_tokens": 78471125.0,
      "step": 34220
    },
    {
      "entropy": 5.037768411636352,
      "epoch": 3.287704130643612,
      "grad_norm": 1.0,
      "learning_rate": 0.00039419629736911076,
      "loss": 4.6212,
      "mean_token_accuracy": 0.25339575558900834,
      "num_tokens": 78481712.0,
      "step": 34225
    },
    {
      "entropy": 5.0764946937561035,
      "epoch": 3.288184438040346,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00039416721849870736,
      "loss": 4.677,
      "mean_token_accuracy": 0.24566420763731003,
      "num_tokens": 78494502.0,
      "step": 34230
    },
    {
      "entropy": 5.203457069396973,
      "epoch": 3.28866474543708,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003941381368615951,
      "loss": 4.7662,
      "mean_token_accuracy": 0.23994168490171433,
      "num_tokens": 78504728.0,
      "step": 34235
    },
    {
      "entropy": 5.189733457565308,
      "epoch": 3.2891450528338138,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003941090524584489,
      "loss": 4.7218,
      "mean_token_accuracy": 0.2513398200273514,
      "num_tokens": 78515765.0,
      "step": 34240
    },
    {
      "entropy": 5.039652681350708,
      "epoch": 3.2896253602305476,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003940799652899442,
      "loss": 4.5632,
      "mean_token_accuracy": 0.25343370586633684,
      "num_tokens": 78527519.0,
      "step": 34245
    },
    {
      "entropy": 5.065464496612549,
      "epoch": 3.2901056676272815,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003940508753567564,
      "loss": 4.6745,
      "mean_token_accuracy": 0.243637415766716,
      "num_tokens": 78539188.0,
      "step": 34250
    },
    {
      "entropy": 5.147671842575074,
      "epoch": 3.2905859750240154,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039402178265956074,
      "loss": 4.7108,
      "mean_token_accuracy": 0.23797611147165298,
      "num_tokens": 78551131.0,
      "step": 34255
    },
    {
      "entropy": 5.143290758132935,
      "epoch": 3.2910662824207493,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003939926871990328,
      "loss": 4.5933,
      "mean_token_accuracy": 0.25632441639900205,
      "num_tokens": 78561760.0,
      "step": 34260
    },
    {
      "entropy": 5.095315074920654,
      "epoch": 3.291546589817483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003939635889758478,
      "loss": 4.7209,
      "mean_token_accuracy": 0.24466701447963715,
      "num_tokens": 78573162.0,
      "step": 34265
    },
    {
      "entropy": 5.0765269756317135,
      "epoch": 3.292026897214217,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039393448799068164,
      "loss": 4.6579,
      "mean_token_accuracy": 0.24952416718006135,
      "num_tokens": 78584453.0,
      "step": 34270
    },
    {
      "entropy": 5.134502315521241,
      "epoch": 3.292507204610951,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003939053842442098,
      "loss": 4.7191,
      "mean_token_accuracy": 0.24012247622013091,
      "num_tokens": 78596411.0,
      "step": 34275
    },
    {
      "entropy": 5.092968130111695,
      "epoch": 3.292987512007685,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039387627773710803,
      "loss": 4.6154,
      "mean_token_accuracy": 0.2503552183508873,
      "num_tokens": 78608320.0,
      "step": 34280
    },
    {
      "entropy": 5.044772100448609,
      "epoch": 3.2934678194044187,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039384716847005216,
      "loss": 4.6264,
      "mean_token_accuracy": 0.24845608621835708,
      "num_tokens": 78620153.0,
      "step": 34285
    },
    {
      "entropy": 5.12590274810791,
      "epoch": 3.2939481268011526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039381805644371774,
      "loss": 4.7118,
      "mean_token_accuracy": 0.2442764326930046,
      "num_tokens": 78630858.0,
      "step": 34290
    },
    {
      "entropy": 5.15791015625,
      "epoch": 3.2944284341978864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000393788941658781,
      "loss": 4.7281,
      "mean_token_accuracy": 0.2441214695572853,
      "num_tokens": 78642497.0,
      "step": 34295
    },
    {
      "entropy": 5.122145938873291,
      "epoch": 3.2949087415946208,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039375982411591774,
      "loss": 4.6873,
      "mean_token_accuracy": 0.24561283886432647,
      "num_tokens": 78653702.0,
      "step": 34300
    },
    {
      "entropy": 5.0443144798278805,
      "epoch": 3.2953890489913547,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039373070381580404,
      "loss": 4.6572,
      "mean_token_accuracy": 0.25179167538881303,
      "num_tokens": 78664848.0,
      "step": 34305
    },
    {
      "entropy": 5.167015743255615,
      "epoch": 3.2958693563880885,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003937015807591159,
      "loss": 4.7773,
      "mean_token_accuracy": 0.23826726227998735,
      "num_tokens": 78676662.0,
      "step": 34310
    },
    {
      "entropy": 5.087010383605957,
      "epoch": 3.2963496637848224,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039367245494652963,
      "loss": 4.6269,
      "mean_token_accuracy": 0.24948279857635497,
      "num_tokens": 78687526.0,
      "step": 34315
    },
    {
      "entropy": 5.0794435977935795,
      "epoch": 3.2968299711815563,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039364332637872125,
      "loss": 4.6472,
      "mean_token_accuracy": 0.2535652741789818,
      "num_tokens": 78698414.0,
      "step": 34320
    },
    {
      "entropy": 5.0818322658538815,
      "epoch": 3.29731027857829,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039361419505636714,
      "loss": 4.5938,
      "mean_token_accuracy": 0.24699832648038864,
      "num_tokens": 78710813.0,
      "step": 34325
    },
    {
      "entropy": 5.137038993835449,
      "epoch": 3.297790585975024,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039358506098014363,
      "loss": 4.6936,
      "mean_token_accuracy": 0.24105131328105928,
      "num_tokens": 78721463.0,
      "step": 34330
    },
    {
      "entropy": 5.124330186843872,
      "epoch": 3.298270893371758,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039355592415072716,
      "loss": 4.681,
      "mean_token_accuracy": 0.2496887966990471,
      "num_tokens": 78733288.0,
      "step": 34335
    },
    {
      "entropy": 5.074912881851196,
      "epoch": 3.298751200768492,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039352678456879415,
      "loss": 4.6229,
      "mean_token_accuracy": 0.2536503180861473,
      "num_tokens": 78743921.0,
      "step": 34340
    },
    {
      "entropy": 5.129919624328613,
      "epoch": 3.2992315081652257,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003934976422350212,
      "loss": 4.7238,
      "mean_token_accuracy": 0.2434869959950447,
      "num_tokens": 78756294.0,
      "step": 34345
    },
    {
      "entropy": 5.102303314208984,
      "epoch": 3.2997118155619596,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003934684971500848,
      "loss": 4.6661,
      "mean_token_accuracy": 0.24669007062911988,
      "num_tokens": 78769061.0,
      "step": 34350
    },
    {
      "entropy": 5.069124698638916,
      "epoch": 3.3001921229586935,
      "grad_norm": 1.515625,
      "learning_rate": 0.00039343934931466165,
      "loss": 4.6029,
      "mean_token_accuracy": 0.2540361970663071,
      "num_tokens": 78781706.0,
      "step": 34355
    },
    {
      "entropy": 5.077430582046508,
      "epoch": 3.3006724303554273,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039341019872942855,
      "loss": 4.7715,
      "mean_token_accuracy": 0.24265347868204118,
      "num_tokens": 78793411.0,
      "step": 34360
    },
    {
      "entropy": 5.072257232666016,
      "epoch": 3.3011527377521612,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039338104539506227,
      "loss": 4.6426,
      "mean_token_accuracy": 0.2540634900331497,
      "num_tokens": 78805202.0,
      "step": 34365
    },
    {
      "entropy": 5.134520578384399,
      "epoch": 3.301633045148895,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003933518893122396,
      "loss": 4.6814,
      "mean_token_accuracy": 0.2515650749206543,
      "num_tokens": 78816927.0,
      "step": 34370
    },
    {
      "entropy": 5.070262145996094,
      "epoch": 3.3021133525456294,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003933227304816375,
      "loss": 4.6437,
      "mean_token_accuracy": 0.24806999415159225,
      "num_tokens": 78828880.0,
      "step": 34375
    },
    {
      "entropy": 5.1943236827850345,
      "epoch": 3.302593659942363,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000393293568903933,
      "loss": 4.7248,
      "mean_token_accuracy": 0.2429069310426712,
      "num_tokens": 78840641.0,
      "step": 34380
    },
    {
      "entropy": 5.052430582046509,
      "epoch": 3.303073967339097,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003932644045798029,
      "loss": 4.663,
      "mean_token_accuracy": 0.249713434278965,
      "num_tokens": 78852601.0,
      "step": 34385
    },
    {
      "entropy": 5.097179317474366,
      "epoch": 3.303554274735831,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003932352375099247,
      "loss": 4.7244,
      "mean_token_accuracy": 0.25118002742528917,
      "num_tokens": 78864757.0,
      "step": 34390
    },
    {
      "entropy": 5.055942821502685,
      "epoch": 3.304034582132565,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003932060676949753,
      "loss": 4.6343,
      "mean_token_accuracy": 0.25395305901765824,
      "num_tokens": 78875771.0,
      "step": 34395
    },
    {
      "entropy": 5.130750131607056,
      "epoch": 3.304514889529299,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003931768951356319,
      "loss": 4.7285,
      "mean_token_accuracy": 0.2386087417602539,
      "num_tokens": 78887101.0,
      "step": 34400
    },
    {
      "entropy": 5.169404029846191,
      "epoch": 3.3049951969260327,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003931477198325721,
      "loss": 4.7021,
      "mean_token_accuracy": 0.23902850449085236,
      "num_tokens": 78898874.0,
      "step": 34405
    },
    {
      "entropy": 5.176192283630371,
      "epoch": 3.3054755043227666,
      "grad_norm": 1.0,
      "learning_rate": 0.000393118541786473,
      "loss": 4.6685,
      "mean_token_accuracy": 0.24163675755262376,
      "num_tokens": 78908820.0,
      "step": 34410
    },
    {
      "entropy": 5.0358054637908936,
      "epoch": 3.3059558117195005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039308936099801203,
      "loss": 4.6393,
      "mean_token_accuracy": 0.2500953793525696,
      "num_tokens": 78920055.0,
      "step": 34415
    },
    {
      "entropy": 5.042429208755493,
      "epoch": 3.3064361191162344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003930601774678669,
      "loss": 4.576,
      "mean_token_accuracy": 0.2499839708209038,
      "num_tokens": 78930846.0,
      "step": 34420
    },
    {
      "entropy": 5.074540376663208,
      "epoch": 3.3069164265129682,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039303099119671487,
      "loss": 4.6376,
      "mean_token_accuracy": 0.2543348327279091,
      "num_tokens": 78943330.0,
      "step": 34425
    },
    {
      "entropy": 5.124925518035889,
      "epoch": 3.307396733909702,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039300180218523374,
      "loss": 4.7454,
      "mean_token_accuracy": 0.23780534714460372,
      "num_tokens": 78955216.0,
      "step": 34430
    },
    {
      "entropy": 5.1640490055084225,
      "epoch": 3.307877041306436,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003929726104341013,
      "loss": 4.747,
      "mean_token_accuracy": 0.24710142314434053,
      "num_tokens": 78965397.0,
      "step": 34435
    },
    {
      "entropy": 5.09324369430542,
      "epoch": 3.30835734870317,
      "grad_norm": 1.0,
      "learning_rate": 0.00039294341594399494,
      "loss": 4.6054,
      "mean_token_accuracy": 0.25135989785194396,
      "num_tokens": 78976649.0,
      "step": 34440
    },
    {
      "entropy": 5.139302921295166,
      "epoch": 3.3088376560999038,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039291421871559274,
      "loss": 4.722,
      "mean_token_accuracy": 0.23790765851736068,
      "num_tokens": 78988060.0,
      "step": 34445
    },
    {
      "entropy": 5.117081260681152,
      "epoch": 3.309317963496638,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039288501874957263,
      "loss": 4.7329,
      "mean_token_accuracy": 0.23985478281974792,
      "num_tokens": 78999376.0,
      "step": 34450
    },
    {
      "entropy": 5.100318098068238,
      "epoch": 3.3097982708933715,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003928558160466123,
      "loss": 4.5609,
      "mean_token_accuracy": 0.2537337005138397,
      "num_tokens": 79011043.0,
      "step": 34455
    },
    {
      "entropy": 5.093052577972412,
      "epoch": 3.310278578290106,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003928266106073899,
      "loss": 4.7734,
      "mean_token_accuracy": 0.24084100723266602,
      "num_tokens": 79023178.0,
      "step": 34460
    },
    {
      "entropy": 5.137637901306152,
      "epoch": 3.3107588856868397,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039279740243258353,
      "loss": 4.678,
      "mean_token_accuracy": 0.24552266746759416,
      "num_tokens": 79034698.0,
      "step": 34465
    },
    {
      "entropy": 5.139394807815552,
      "epoch": 3.3112391930835736,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003927681915228712,
      "loss": 4.7014,
      "mean_token_accuracy": 0.2415623188018799,
      "num_tokens": 79045529.0,
      "step": 34470
    },
    {
      "entropy": 5.0824305534362795,
      "epoch": 3.3117195004803075,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003927389778789312,
      "loss": 4.6828,
      "mean_token_accuracy": 0.24388751089572908,
      "num_tokens": 79057932.0,
      "step": 34475
    },
    {
      "entropy": 5.045950222015381,
      "epoch": 3.3121998078770414,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003927097615014418,
      "loss": 4.6062,
      "mean_token_accuracy": 0.25504413396120074,
      "num_tokens": 79069486.0,
      "step": 34480
    },
    {
      "entropy": 5.108188962936401,
      "epoch": 3.3126801152737753,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003926805423910811,
      "loss": 4.6961,
      "mean_token_accuracy": 0.24886075109243394,
      "num_tokens": 79080751.0,
      "step": 34485
    },
    {
      "entropy": 5.070665454864502,
      "epoch": 3.313160422670509,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003926513205485278,
      "loss": 4.6102,
      "mean_token_accuracy": 0.2474851056933403,
      "num_tokens": 79092000.0,
      "step": 34490
    },
    {
      "entropy": 5.064667415618897,
      "epoch": 3.313640730067243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003926220959744602,
      "loss": 4.6056,
      "mean_token_accuracy": 0.24921095818281175,
      "num_tokens": 79102588.0,
      "step": 34495
    },
    {
      "entropy": 5.073616027832031,
      "epoch": 3.314121037463977,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003925928686695567,
      "loss": 4.6862,
      "mean_token_accuracy": 0.2445344015955925,
      "num_tokens": 79114859.0,
      "step": 34500
    },
    {
      "entropy": 5.091532897949219,
      "epoch": 3.314601344860711,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000392563638634496,
      "loss": 4.6472,
      "mean_token_accuracy": 0.2518300101161003,
      "num_tokens": 79126529.0,
      "step": 34505
    },
    {
      "entropy": 5.1970658779144285,
      "epoch": 3.3150816522574447,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003925344058699567,
      "loss": 4.7233,
      "mean_token_accuracy": 0.24199410527944565,
      "num_tokens": 79139120.0,
      "step": 34510
    },
    {
      "entropy": 5.0975243091583256,
      "epoch": 3.3155619596541785,
      "grad_norm": 1.0,
      "learning_rate": 0.0003925051703766175,
      "loss": 4.6647,
      "mean_token_accuracy": 0.23965206146240234,
      "num_tokens": 79151169.0,
      "step": 34515
    },
    {
      "entropy": 5.044404935836792,
      "epoch": 3.3160422670509124,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003924759321551572,
      "loss": 4.6567,
      "mean_token_accuracy": 0.24770238548517226,
      "num_tokens": 79162016.0,
      "step": 34520
    },
    {
      "entropy": 5.03289909362793,
      "epoch": 3.3165225744476463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003924466912062546,
      "loss": 4.6132,
      "mean_token_accuracy": 0.24861456453800201,
      "num_tokens": 79174195.0,
      "step": 34525
    },
    {
      "entropy": 5.25216817855835,
      "epoch": 3.31700288184438,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003924174475305885,
      "loss": 4.7839,
      "mean_token_accuracy": 0.24134636670351028,
      "num_tokens": 79186160.0,
      "step": 34530
    },
    {
      "entropy": 5.110486745834351,
      "epoch": 3.3174831892411145,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039238820112883795,
      "loss": 4.6624,
      "mean_token_accuracy": 0.24025769084692,
      "num_tokens": 79197884.0,
      "step": 34535
    },
    {
      "entropy": 5.023403215408325,
      "epoch": 3.3179634966378484,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003923589520016819,
      "loss": 4.6294,
      "mean_token_accuracy": 0.2547737643122673,
      "num_tokens": 79210799.0,
      "step": 34540
    },
    {
      "entropy": 5.066167974472046,
      "epoch": 3.3184438040345823,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039232970014979965,
      "loss": 4.695,
      "mean_token_accuracy": 0.24728138893842697,
      "num_tokens": 79223092.0,
      "step": 34545
    },
    {
      "entropy": 5.095422697067261,
      "epoch": 3.318924111431316,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003923004455738699,
      "loss": 4.6477,
      "mean_token_accuracy": 0.2521927371621132,
      "num_tokens": 79234801.0,
      "step": 34550
    },
    {
      "entropy": 5.1220087051391605,
      "epoch": 3.31940441882805,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039227118827457234,
      "loss": 4.7232,
      "mean_token_accuracy": 0.2435407191514969,
      "num_tokens": 79245763.0,
      "step": 34555
    },
    {
      "entropy": 5.129978179931641,
      "epoch": 3.319884726224784,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039224192825258584,
      "loss": 4.7041,
      "mean_token_accuracy": 0.24303609281778335,
      "num_tokens": 79256113.0,
      "step": 34560
    },
    {
      "entropy": 5.165873241424561,
      "epoch": 3.320365033621518,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039221266550859004,
      "loss": 4.734,
      "mean_token_accuracy": 0.2446964919567108,
      "num_tokens": 79269244.0,
      "step": 34565
    },
    {
      "entropy": 5.130612182617187,
      "epoch": 3.3208453410182517,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039218340004326414,
      "loss": 4.6614,
      "mean_token_accuracy": 0.25059044510126116,
      "num_tokens": 79279861.0,
      "step": 34570
    },
    {
      "entropy": 5.109393072128296,
      "epoch": 3.3213256484149856,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003921541318572876,
      "loss": 4.6766,
      "mean_token_accuracy": 0.24901448488235473,
      "num_tokens": 79291374.0,
      "step": 34575
    },
    {
      "entropy": 5.097543382644654,
      "epoch": 3.3218059558117194,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039212486095134005,
      "loss": 4.6173,
      "mean_token_accuracy": 0.25035566836595535,
      "num_tokens": 79301672.0,
      "step": 34580
    },
    {
      "entropy": 5.064320659637451,
      "epoch": 3.3222862632084533,
      "grad_norm": 1.0,
      "learning_rate": 0.0003920955873261011,
      "loss": 4.6051,
      "mean_token_accuracy": 0.2509954750537872,
      "num_tokens": 79313131.0,
      "step": 34585
    },
    {
      "entropy": 5.1141729831695555,
      "epoch": 3.322766570605187,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003920663109822502,
      "loss": 4.7056,
      "mean_token_accuracy": 0.24250114262104033,
      "num_tokens": 79324117.0,
      "step": 34590
    },
    {
      "entropy": 5.118050575256348,
      "epoch": 3.323246878001921,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00039203703192046717,
      "loss": 4.6396,
      "mean_token_accuracy": 0.2549441158771515,
      "num_tokens": 79335724.0,
      "step": 34595
    },
    {
      "entropy": 5.141208553314209,
      "epoch": 3.323727185398655,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003920077501414318,
      "loss": 4.7422,
      "mean_token_accuracy": 0.24125732332468033,
      "num_tokens": 79347984.0,
      "step": 34600
    },
    {
      "entropy": 5.14498553276062,
      "epoch": 3.324207492795389,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039197846564582395,
      "loss": 4.7488,
      "mean_token_accuracy": 0.24562108367681504,
      "num_tokens": 79360752.0,
      "step": 34605
    },
    {
      "entropy": 5.208582925796509,
      "epoch": 3.324687800192123,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039194917843432347,
      "loss": 4.8511,
      "mean_token_accuracy": 0.23018649220466614,
      "num_tokens": 79374525.0,
      "step": 34610
    },
    {
      "entropy": 5.0313441276550295,
      "epoch": 3.325168107588857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003919198885076103,
      "loss": 4.5699,
      "mean_token_accuracy": 0.2565922960639,
      "num_tokens": 79385014.0,
      "step": 34615
    },
    {
      "entropy": 5.044430446624756,
      "epoch": 3.325648414985591,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039189059586636465,
      "loss": 4.5911,
      "mean_token_accuracy": 0.2553503572940826,
      "num_tokens": 79396027.0,
      "step": 34620
    },
    {
      "entropy": 5.117130374908447,
      "epoch": 3.326128722382325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003918613005112663,
      "loss": 4.6541,
      "mean_token_accuracy": 0.24326271265745164,
      "num_tokens": 79407080.0,
      "step": 34625
    },
    {
      "entropy": 5.115348768234253,
      "epoch": 3.3266090297790587,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003918320024429956,
      "loss": 4.6402,
      "mean_token_accuracy": 0.2540326327085495,
      "num_tokens": 79416986.0,
      "step": 34630
    },
    {
      "entropy": 5.071494770050049,
      "epoch": 3.3270893371757926,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003918027016622328,
      "loss": 4.5946,
      "mean_token_accuracy": 0.2537030458450317,
      "num_tokens": 79428088.0,
      "step": 34635
    },
    {
      "entropy": 5.089864778518677,
      "epoch": 3.3275696445725265,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003917733981696579,
      "loss": 4.7087,
      "mean_token_accuracy": 0.24739547520875932,
      "num_tokens": 79439827.0,
      "step": 34640
    },
    {
      "entropy": 5.1801097869873045,
      "epoch": 3.3280499519692603,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003917440919659516,
      "loss": 4.757,
      "mean_token_accuracy": 0.2360815867781639,
      "num_tokens": 79451994.0,
      "step": 34645
    },
    {
      "entropy": 5.097568559646606,
      "epoch": 3.3285302593659942,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003917147830517941,
      "loss": 4.6097,
      "mean_token_accuracy": 0.25168000906705856,
      "num_tokens": 79464186.0,
      "step": 34650
    },
    {
      "entropy": 5.114072847366333,
      "epoch": 3.329010566762728,
      "grad_norm": 1.0,
      "learning_rate": 0.0003916854714278659,
      "loss": 4.7202,
      "mean_token_accuracy": 0.2421091765165329,
      "num_tokens": 79475794.0,
      "step": 34655
    },
    {
      "entropy": 5.130643224716186,
      "epoch": 3.329490874159462,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003916561570948476,
      "loss": 4.6496,
      "mean_token_accuracy": 0.2513890787959099,
      "num_tokens": 79486517.0,
      "step": 34660
    },
    {
      "entropy": 5.127585983276367,
      "epoch": 3.329971181556196,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003916268400534197,
      "loss": 4.7485,
      "mean_token_accuracy": 0.2374435231089592,
      "num_tokens": 79498150.0,
      "step": 34665
    },
    {
      "entropy": 5.112618064880371,
      "epoch": 3.3304514889529298,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003915975203042628,
      "loss": 4.7106,
      "mean_token_accuracy": 0.24600221067667008,
      "num_tokens": 79508131.0,
      "step": 34670
    },
    {
      "entropy": 5.160793352127075,
      "epoch": 3.3309317963496636,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039156819784805783,
      "loss": 4.7697,
      "mean_token_accuracy": 0.24028586149215697,
      "num_tokens": 79519128.0,
      "step": 34675
    },
    {
      "entropy": 5.09904260635376,
      "epoch": 3.3314121037463975,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003915388726854854,
      "loss": 4.6557,
      "mean_token_accuracy": 0.25198905169963837,
      "num_tokens": 79530202.0,
      "step": 34680
    },
    {
      "entropy": 5.055880928039551,
      "epoch": 3.331892411143132,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039150954481722634,
      "loss": 4.6574,
      "mean_token_accuracy": 0.24316917955875397,
      "num_tokens": 79542130.0,
      "step": 34685
    },
    {
      "entropy": 5.110525131225586,
      "epoch": 3.3323727185398653,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003914802142439617,
      "loss": 4.6692,
      "mean_token_accuracy": 0.24698539972305297,
      "num_tokens": 79554311.0,
      "step": 34690
    },
    {
      "entropy": 5.113149118423462,
      "epoch": 3.3328530259365996,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003914508809663723,
      "loss": 4.6132,
      "mean_token_accuracy": 0.24985936284065247,
      "num_tokens": 79566092.0,
      "step": 34695
    },
    {
      "entropy": 5.079854297637939,
      "epoch": 3.3333333333333335,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00039142154498513913,
      "loss": 4.6527,
      "mean_token_accuracy": 0.2498472139239311,
      "num_tokens": 79578157.0,
      "step": 34700
    },
    {
      "entropy": 5.085790300369263,
      "epoch": 3.3338136407300674,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039139220630094357,
      "loss": 4.6357,
      "mean_token_accuracy": 0.2495850905776024,
      "num_tokens": 79591414.0,
      "step": 34705
    },
    {
      "entropy": 5.1140495300292965,
      "epoch": 3.3342939481268012,
      "grad_norm": 1.0,
      "learning_rate": 0.00039136286491446657,
      "loss": 4.616,
      "mean_token_accuracy": 0.2529078394174576,
      "num_tokens": 79603671.0,
      "step": 34710
    },
    {
      "entropy": 5.128128433227539,
      "epoch": 3.334774255523535,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00039133352082638923,
      "loss": 4.7396,
      "mean_token_accuracy": 0.23857685029506684,
      "num_tokens": 79616014.0,
      "step": 34715
    },
    {
      "entropy": 5.127895832061768,
      "epoch": 3.335254562920269,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00039130417403739315,
      "loss": 4.7075,
      "mean_token_accuracy": 0.2471451312303543,
      "num_tokens": 79628002.0,
      "step": 34720
    },
    {
      "entropy": 5.107656955718994,
      "epoch": 3.335734870317003,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003912748245481594,
      "loss": 4.6864,
      "mean_token_accuracy": 0.24840205013751984,
      "num_tokens": 79638927.0,
      "step": 34725
    },
    {
      "entropy": 5.183277273178101,
      "epoch": 3.3362151777137368,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039124547235936947,
      "loss": 4.7826,
      "mean_token_accuracy": 0.23500719666481018,
      "num_tokens": 79651797.0,
      "step": 34730
    },
    {
      "entropy": 5.22091965675354,
      "epoch": 3.3366954851104706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039121611747170495,
      "loss": 4.7283,
      "mean_token_accuracy": 0.24099528789520264,
      "num_tokens": 79662238.0,
      "step": 34735
    },
    {
      "entropy": 5.120099067687988,
      "epoch": 3.3371757925072045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039118675988584724,
      "loss": 4.6789,
      "mean_token_accuracy": 0.24922218769788743,
      "num_tokens": 79672868.0,
      "step": 34740
    },
    {
      "entropy": 5.069494581222534,
      "epoch": 3.3376560999039384,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003911573996024779,
      "loss": 4.704,
      "mean_token_accuracy": 0.24478663206100465,
      "num_tokens": 79685242.0,
      "step": 34745
    },
    {
      "entropy": 5.118858575820923,
      "epoch": 3.3381364073006723,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003911280366222787,
      "loss": 4.7007,
      "mean_token_accuracy": 0.24606858789920807,
      "num_tokens": 79697172.0,
      "step": 34750
    },
    {
      "entropy": 5.136493778228759,
      "epoch": 3.338616714697406,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039109867094593134,
      "loss": 4.6876,
      "mean_token_accuracy": 0.2467506095767021,
      "num_tokens": 79709261.0,
      "step": 34755
    },
    {
      "entropy": 5.093030834197998,
      "epoch": 3.3390970220941405,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003910693025741175,
      "loss": 4.6322,
      "mean_token_accuracy": 0.24845488369464874,
      "num_tokens": 79719990.0,
      "step": 34760
    },
    {
      "entropy": 5.112219905853271,
      "epoch": 3.339577329490874,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039103993150751916,
      "loss": 4.6832,
      "mean_token_accuracy": 0.25032223463058473,
      "num_tokens": 79732835.0,
      "step": 34765
    },
    {
      "entropy": 5.168570804595947,
      "epoch": 3.3400576368876083,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039101055774681825,
      "loss": 4.7206,
      "mean_token_accuracy": 0.24251952916383743,
      "num_tokens": 79743752.0,
      "step": 34770
    },
    {
      "entropy": 5.163613319396973,
      "epoch": 3.340537944284342,
      "grad_norm": 1.0,
      "learning_rate": 0.0003909811812926966,
      "loss": 4.7538,
      "mean_token_accuracy": 0.2450694590806961,
      "num_tokens": 79756209.0,
      "step": 34775
    },
    {
      "entropy": 5.0656548023223875,
      "epoch": 3.341018251681076,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003909518021458363,
      "loss": 4.63,
      "mean_token_accuracy": 0.253447425365448,
      "num_tokens": 79766806.0,
      "step": 34780
    },
    {
      "entropy": 5.072398138046265,
      "epoch": 3.34149855907781,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003909224203069195,
      "loss": 4.6965,
      "mean_token_accuracy": 0.24773208945989608,
      "num_tokens": 79778803.0,
      "step": 34785
    },
    {
      "entropy": 5.0793109893798825,
      "epoch": 3.341978866474544,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003908930357766283,
      "loss": 4.6359,
      "mean_token_accuracy": 0.25117896497249603,
      "num_tokens": 79789687.0,
      "step": 34790
    },
    {
      "entropy": 5.138288927078247,
      "epoch": 3.3424591738712777,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003908636485556449,
      "loss": 4.6969,
      "mean_token_accuracy": 0.24582888484001159,
      "num_tokens": 79800459.0,
      "step": 34795
    },
    {
      "entropy": 5.051034545898437,
      "epoch": 3.3429394812680115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039083425864465165,
      "loss": 4.6024,
      "mean_token_accuracy": 0.25302841514348984,
      "num_tokens": 79811641.0,
      "step": 34800
    },
    {
      "entropy": 5.108280658721924,
      "epoch": 3.3434197886647454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003908048660443309,
      "loss": 4.7255,
      "mean_token_accuracy": 0.24530348628759385,
      "num_tokens": 79823276.0,
      "step": 34805
    },
    {
      "entropy": 5.118523740768433,
      "epoch": 3.3439000960614793,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003907754707553651,
      "loss": 4.6552,
      "mean_token_accuracy": 0.2540632888674736,
      "num_tokens": 79834694.0,
      "step": 34810
    },
    {
      "entropy": 5.205039358139038,
      "epoch": 3.344380403458213,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003907460727784365,
      "loss": 4.7601,
      "mean_token_accuracy": 0.2406563326716423,
      "num_tokens": 79846206.0,
      "step": 34815
    },
    {
      "entropy": 5.1722818374633786,
      "epoch": 3.344860710854947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039071667211422787,
      "loss": 4.7481,
      "mean_token_accuracy": 0.2393363893032074,
      "num_tokens": 79857273.0,
      "step": 34820
    },
    {
      "entropy": 5.11325216293335,
      "epoch": 3.345341018251681,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003906872687634217,
      "loss": 4.6428,
      "mean_token_accuracy": 0.250268018245697,
      "num_tokens": 79868799.0,
      "step": 34825
    },
    {
      "entropy": 5.033809232711792,
      "epoch": 3.345821325648415,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039065786272670066,
      "loss": 4.6146,
      "mean_token_accuracy": 0.2602176159620285,
      "num_tokens": 79879142.0,
      "step": 34830
    },
    {
      "entropy": 5.019201755523682,
      "epoch": 3.3463016330451487,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003906284540047475,
      "loss": 4.6101,
      "mean_token_accuracy": 0.2555994883179665,
      "num_tokens": 79890455.0,
      "step": 34835
    },
    {
      "entropy": 5.060307598114013,
      "epoch": 3.3467819404418826,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039059904259824507,
      "loss": 4.625,
      "mean_token_accuracy": 0.2538018822669983,
      "num_tokens": 79901198.0,
      "step": 34840
    },
    {
      "entropy": 5.070934295654297,
      "epoch": 3.347262247838617,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000390569628507876,
      "loss": 4.626,
      "mean_token_accuracy": 0.25019067376852033,
      "num_tokens": 79912567.0,
      "step": 34845
    },
    {
      "entropy": 5.127518558502198,
      "epoch": 3.347742555235351,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039054021173432336,
      "loss": 4.7666,
      "mean_token_accuracy": 0.2466248467564583,
      "num_tokens": 79925049.0,
      "step": 34850
    },
    {
      "entropy": 5.1655388355255125,
      "epoch": 3.3482228626320847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003905107922782701,
      "loss": 4.7379,
      "mean_token_accuracy": 0.25238690674304964,
      "num_tokens": 79936746.0,
      "step": 34855
    },
    {
      "entropy": 5.064845561981201,
      "epoch": 3.3487031700288186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003904813701403993,
      "loss": 4.5882,
      "mean_token_accuracy": 0.2518634393811226,
      "num_tokens": 79947022.0,
      "step": 34860
    },
    {
      "entropy": 5.057610368728637,
      "epoch": 3.3491834774255524,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039045194532139396,
      "loss": 4.6528,
      "mean_token_accuracy": 0.2478608176112175,
      "num_tokens": 79958085.0,
      "step": 34865
    },
    {
      "entropy": 5.144690656661988,
      "epoch": 3.3496637848222863,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003904225178219372,
      "loss": 4.7042,
      "mean_token_accuracy": 0.24799249917268754,
      "num_tokens": 79970049.0,
      "step": 34870
    },
    {
      "entropy": 5.062935638427734,
      "epoch": 3.35014409221902,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00039039308764271237,
      "loss": 4.6503,
      "mean_token_accuracy": 0.24749303460121155,
      "num_tokens": 79983486.0,
      "step": 34875
    },
    {
      "entropy": 5.010172414779663,
      "epoch": 3.350624399615754,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003903636547844026,
      "loss": 4.5564,
      "mean_token_accuracy": 0.2540229707956314,
      "num_tokens": 79994955.0,
      "step": 34880
    },
    {
      "entropy": 5.132677221298218,
      "epoch": 3.351104707012488,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039033421924769145,
      "loss": 4.6835,
      "mean_token_accuracy": 0.2405211955308914,
      "num_tokens": 80005472.0,
      "step": 34885
    },
    {
      "entropy": 5.084688758850097,
      "epoch": 3.351585014409222,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039030478103326216,
      "loss": 4.5513,
      "mean_token_accuracy": 0.2647500365972519,
      "num_tokens": 80017081.0,
      "step": 34890
    },
    {
      "entropy": 5.044216108322144,
      "epoch": 3.3520653218059557,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039027534014179823,
      "loss": 4.6434,
      "mean_token_accuracy": 0.246346352994442,
      "num_tokens": 80027533.0,
      "step": 34895
    },
    {
      "entropy": 5.075244903564453,
      "epoch": 3.3525456292026896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003902458965739832,
      "loss": 4.6408,
      "mean_token_accuracy": 0.253554967045784,
      "num_tokens": 80038311.0,
      "step": 34900
    },
    {
      "entropy": 5.095725631713867,
      "epoch": 3.3530259365994235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003902164503305006,
      "loss": 4.6439,
      "mean_token_accuracy": 0.25506806671619414,
      "num_tokens": 80049102.0,
      "step": 34905
    },
    {
      "entropy": 5.184767055511474,
      "epoch": 3.3535062439961574,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003901870014120343,
      "loss": 4.786,
      "mean_token_accuracy": 0.23540452718734742,
      "num_tokens": 80060797.0,
      "step": 34910
    },
    {
      "entropy": 5.024787282943725,
      "epoch": 3.3539865513928913,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003901575498192678,
      "loss": 4.4625,
      "mean_token_accuracy": 0.2621771663427353,
      "num_tokens": 80071473.0,
      "step": 34915
    },
    {
      "entropy": 5.038287353515625,
      "epoch": 3.3544668587896256,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003901280955528849,
      "loss": 4.6229,
      "mean_token_accuracy": 0.2504727840423584,
      "num_tokens": 80083245.0,
      "step": 34920
    },
    {
      "entropy": 5.065965032577514,
      "epoch": 3.3549471661863595,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003900986386135695,
      "loss": 4.6926,
      "mean_token_accuracy": 0.2423554465174675,
      "num_tokens": 80095471.0,
      "step": 34925
    },
    {
      "entropy": 5.078068208694458,
      "epoch": 3.3554274735830933,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039006917900200543,
      "loss": 4.6197,
      "mean_token_accuracy": 0.24859418869018554,
      "num_tokens": 80106640.0,
      "step": 34930
    },
    {
      "entropy": 5.008115625381469,
      "epoch": 3.3559077809798272,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039003971671887675,
      "loss": 4.5977,
      "mean_token_accuracy": 0.2565582200884819,
      "num_tokens": 80119358.0,
      "step": 34935
    },
    {
      "entropy": 5.022244119644165,
      "epoch": 3.356388088376561,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003900102517648674,
      "loss": 4.6291,
      "mean_token_accuracy": 0.24355473667383193,
      "num_tokens": 80130383.0,
      "step": 34940
    },
    {
      "entropy": 5.090456962585449,
      "epoch": 3.356868395773295,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003899807841406617,
      "loss": 4.6565,
      "mean_token_accuracy": 0.24953960627317429,
      "num_tokens": 80142185.0,
      "step": 34945
    },
    {
      "entropy": 4.9857964515686035,
      "epoch": 3.357348703170029,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003899513138469434,
      "loss": 4.6304,
      "mean_token_accuracy": 0.24399228096008302,
      "num_tokens": 80154244.0,
      "step": 34950
    },
    {
      "entropy": 5.1461066722869875,
      "epoch": 3.3578290105667628,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000389921840884397,
      "loss": 4.6795,
      "mean_token_accuracy": 0.2458556815981865,
      "num_tokens": 80165061.0,
      "step": 34955
    },
    {
      "entropy": 5.05560622215271,
      "epoch": 3.3583093179634966,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038989236525370676,
      "loss": 4.6333,
      "mean_token_accuracy": 0.2534294754266739,
      "num_tokens": 80177017.0,
      "step": 34960
    },
    {
      "entropy": 5.091989469528198,
      "epoch": 3.3587896253602305,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003898628869555569,
      "loss": 4.6515,
      "mean_token_accuracy": 0.2489737465977669,
      "num_tokens": 80189067.0,
      "step": 34965
    },
    {
      "entropy": 5.095130205154419,
      "epoch": 3.3592699327569644,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038983340599063187,
      "loss": 4.6877,
      "mean_token_accuracy": 0.24312635362148285,
      "num_tokens": 80200069.0,
      "step": 34970
    },
    {
      "entropy": 5.062352418899536,
      "epoch": 3.3597502401536983,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003898039223596162,
      "loss": 4.6518,
      "mean_token_accuracy": 0.24744091928005219,
      "num_tokens": 80212343.0,
      "step": 34975
    },
    {
      "entropy": 5.068124437332154,
      "epoch": 3.360230547550432,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003897744360631943,
      "loss": 4.685,
      "mean_token_accuracy": 0.24397334307432175,
      "num_tokens": 80222914.0,
      "step": 34980
    },
    {
      "entropy": 5.0670675277709964,
      "epoch": 3.360710854947166,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038974494710205084,
      "loss": 4.6002,
      "mean_token_accuracy": 0.25926837176084516,
      "num_tokens": 80235344.0,
      "step": 34985
    },
    {
      "entropy": 5.079458713531494,
      "epoch": 3.3611911623439,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038971545547687036,
      "loss": 4.6492,
      "mean_token_accuracy": 0.2490754172205925,
      "num_tokens": 80246883.0,
      "step": 34990
    },
    {
      "entropy": 5.108183240890503,
      "epoch": 3.3616714697406342,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038968596118833766,
      "loss": 4.7011,
      "mean_token_accuracy": 0.2471543461084366,
      "num_tokens": 80258387.0,
      "step": 34995
    },
    {
      "entropy": 5.096744632720947,
      "epoch": 3.3621517771373677,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038965646423713744,
      "loss": 4.6967,
      "mean_token_accuracy": 0.2425475761294365,
      "num_tokens": 80269687.0,
      "step": 35000
    },
    {
      "entropy": 5.115563154220581,
      "epoch": 3.362632084534102,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038962696462395473,
      "loss": 4.7097,
      "mean_token_accuracy": 0.23883591890335082,
      "num_tokens": 80280639.0,
      "step": 35005
    },
    {
      "entropy": 5.169279146194458,
      "epoch": 3.363112391930836,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003895974623494742,
      "loss": 4.7596,
      "mean_token_accuracy": 0.2324692815542221,
      "num_tokens": 80292565.0,
      "step": 35010
    },
    {
      "entropy": 5.184310245513916,
      "epoch": 3.3635926993275698,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038956795741438085,
      "loss": 4.7014,
      "mean_token_accuracy": 0.2504858300089836,
      "num_tokens": 80302599.0,
      "step": 35015
    },
    {
      "entropy": 5.129266738891602,
      "epoch": 3.3640730067243036,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038953844981935975,
      "loss": 4.7259,
      "mean_token_accuracy": 0.23678299635648728,
      "num_tokens": 80315050.0,
      "step": 35020
    },
    {
      "entropy": 5.158787488937378,
      "epoch": 3.3645533141210375,
      "grad_norm": 1.0,
      "learning_rate": 0.00038950893956509597,
      "loss": 4.7485,
      "mean_token_accuracy": 0.24751344323158264,
      "num_tokens": 80327935.0,
      "step": 35025
    },
    {
      "entropy": 5.146347141265869,
      "epoch": 3.3650336215177714,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003894794266522746,
      "loss": 4.6557,
      "mean_token_accuracy": 0.24658633172512054,
      "num_tokens": 80338376.0,
      "step": 35030
    },
    {
      "entropy": 5.065600156784058,
      "epoch": 3.3655139289145053,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038944991108158094,
      "loss": 4.7144,
      "mean_token_accuracy": 0.24468704909086228,
      "num_tokens": 80350294.0,
      "step": 35035
    },
    {
      "entropy": 5.079691219329834,
      "epoch": 3.365994236311239,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003894203928537001,
      "loss": 4.5526,
      "mean_token_accuracy": 0.2550753951072693,
      "num_tokens": 80361893.0,
      "step": 35040
    },
    {
      "entropy": 5.101787757873535,
      "epoch": 3.366474543707973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038939087196931754,
      "loss": 4.6892,
      "mean_token_accuracy": 0.2465742826461792,
      "num_tokens": 80374370.0,
      "step": 35045
    },
    {
      "entropy": 5.106787443161011,
      "epoch": 3.366954851104707,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038936134842911863,
      "loss": 4.6206,
      "mean_token_accuracy": 0.24869674444198608,
      "num_tokens": 80385523.0,
      "step": 35050
    },
    {
      "entropy": 5.069877099990845,
      "epoch": 3.367435158501441,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003893318222337888,
      "loss": 4.6383,
      "mean_token_accuracy": 0.2504971966147423,
      "num_tokens": 80397471.0,
      "step": 35055
    },
    {
      "entropy": 5.051280212402344,
      "epoch": 3.3679154658981747,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038930229338401354,
      "loss": 4.5291,
      "mean_token_accuracy": 0.26572591364383696,
      "num_tokens": 80407742.0,
      "step": 35060
    },
    {
      "entropy": 5.114731311798096,
      "epoch": 3.3683957732949086,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003892727618804783,
      "loss": 4.6944,
      "mean_token_accuracy": 0.2446904808282852,
      "num_tokens": 80419165.0,
      "step": 35065
    },
    {
      "entropy": 5.090797758102417,
      "epoch": 3.3688760806916425,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000389243227723869,
      "loss": 4.7241,
      "mean_token_accuracy": 0.24841180443763733,
      "num_tokens": 80430789.0,
      "step": 35070
    },
    {
      "entropy": 5.033324813842773,
      "epoch": 3.3693563880883763,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003892136909148711,
      "loss": 4.6478,
      "mean_token_accuracy": 0.25277635604143145,
      "num_tokens": 80441797.0,
      "step": 35075
    },
    {
      "entropy": 5.075445938110351,
      "epoch": 3.3698366954851107,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003891841514541706,
      "loss": 4.6614,
      "mean_token_accuracy": 0.25142702460289,
      "num_tokens": 80452655.0,
      "step": 35080
    },
    {
      "entropy": 5.108911228179932,
      "epoch": 3.3703170028818445,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000389154609342453,
      "loss": 4.6122,
      "mean_token_accuracy": 0.24996165484189986,
      "num_tokens": 80463348.0,
      "step": 35085
    },
    {
      "entropy": 5.139072561264038,
      "epoch": 3.3707973102785784,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003891250645804044,
      "loss": 4.7331,
      "mean_token_accuracy": 0.24166405349969863,
      "num_tokens": 80475815.0,
      "step": 35090
    },
    {
      "entropy": 5.113673353195191,
      "epoch": 3.3712776176753123,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038909551716871074,
      "loss": 4.6392,
      "mean_token_accuracy": 0.2532001256942749,
      "num_tokens": 80485854.0,
      "step": 35095
    },
    {
      "entropy": 5.140195703506469,
      "epoch": 3.371757925072046,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003890659671080579,
      "loss": 4.6463,
      "mean_token_accuracy": 0.2520283102989197,
      "num_tokens": 80498032.0,
      "step": 35100
    },
    {
      "entropy": 5.055180263519287,
      "epoch": 3.37223823246878,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000389036414399132,
      "loss": 4.6332,
      "mean_token_accuracy": 0.24695904105901717,
      "num_tokens": 80509078.0,
      "step": 35105
    },
    {
      "entropy": 5.028628301620484,
      "epoch": 3.372718539865514,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003890068590426191,
      "loss": 4.5355,
      "mean_token_accuracy": 0.26004105508327485,
      "num_tokens": 80519852.0,
      "step": 35110
    },
    {
      "entropy": 5.055425691604614,
      "epoch": 3.373198847262248,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003889773010392056,
      "loss": 4.6817,
      "mean_token_accuracy": 0.2501548156142235,
      "num_tokens": 80531876.0,
      "step": 35115
    },
    {
      "entropy": 5.082882356643677,
      "epoch": 3.3736791546589817,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038894774038957756,
      "loss": 4.6677,
      "mean_token_accuracy": 0.2502983421087265,
      "num_tokens": 80542864.0,
      "step": 35120
    },
    {
      "entropy": 5.0937323570251465,
      "epoch": 3.3741594620557156,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038891817709442135,
      "loss": 4.6146,
      "mean_token_accuracy": 0.2543604806065559,
      "num_tokens": 80553274.0,
      "step": 35125
    },
    {
      "entropy": 5.021528053283691,
      "epoch": 3.3746397694524495,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038888861115442334,
      "loss": 4.6585,
      "mean_token_accuracy": 0.2447609096765518,
      "num_tokens": 80565397.0,
      "step": 35130
    },
    {
      "entropy": 5.102515602111817,
      "epoch": 3.3751200768491834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003888590425702699,
      "loss": 4.6441,
      "mean_token_accuracy": 0.24912097603082656,
      "num_tokens": 80575683.0,
      "step": 35135
    },
    {
      "entropy": 5.1276966571807865,
      "epoch": 3.3756003842459172,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003888294713426477,
      "loss": 4.6834,
      "mean_token_accuracy": 0.2526715859770775,
      "num_tokens": 80587398.0,
      "step": 35140
    },
    {
      "entropy": 5.077829790115357,
      "epoch": 3.376080691642651,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00038879989747224317,
      "loss": 4.5845,
      "mean_token_accuracy": 0.2578653395175934,
      "num_tokens": 80598670.0,
      "step": 35145
    },
    {
      "entropy": 5.138495063781738,
      "epoch": 3.376560999039385,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003887703209597428,
      "loss": 4.7475,
      "mean_token_accuracy": 0.2394913464784622,
      "num_tokens": 80610192.0,
      "step": 35150
    },
    {
      "entropy": 5.110005235671997,
      "epoch": 3.3770413064361193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003887407418058335,
      "loss": 4.6165,
      "mean_token_accuracy": 0.2506800964474678,
      "num_tokens": 80621736.0,
      "step": 35155
    },
    {
      "entropy": 5.098136234283447,
      "epoch": 3.377521613832853,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038871116001120196,
      "loss": 4.6808,
      "mean_token_accuracy": 0.24695847928524017,
      "num_tokens": 80633172.0,
      "step": 35160
    },
    {
      "entropy": 5.017114782333374,
      "epoch": 3.378001921229587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003886815755765348,
      "loss": 4.635,
      "mean_token_accuracy": 0.24263230562210084,
      "num_tokens": 80643744.0,
      "step": 35165
    },
    {
      "entropy": 5.034330081939697,
      "epoch": 3.378482228626321,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003886519885025191,
      "loss": 4.664,
      "mean_token_accuracy": 0.2487565502524376,
      "num_tokens": 80654960.0,
      "step": 35170
    },
    {
      "entropy": 5.1546399116516115,
      "epoch": 3.378962536023055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038862239878984173,
      "loss": 4.6794,
      "mean_token_accuracy": 0.24679398834705352,
      "num_tokens": 80667119.0,
      "step": 35175
    },
    {
      "entropy": 5.06045708656311,
      "epoch": 3.3794428434197887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003885928064391897,
      "loss": 4.6387,
      "mean_token_accuracy": 0.24800171554088593,
      "num_tokens": 80677975.0,
      "step": 35180
    },
    {
      "entropy": 5.192850589752197,
      "epoch": 3.3799231508165226,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003885632114512499,
      "loss": 4.802,
      "mean_token_accuracy": 0.23716465383768082,
      "num_tokens": 80690033.0,
      "step": 35185
    },
    {
      "entropy": 5.072188472747802,
      "epoch": 3.3804034582132565,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038853361382670956,
      "loss": 4.5989,
      "mean_token_accuracy": 0.2574413985013962,
      "num_tokens": 80700299.0,
      "step": 35190
    },
    {
      "entropy": 5.148026180267334,
      "epoch": 3.3808837656099904,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038850401356625583,
      "loss": 4.7626,
      "mean_token_accuracy": 0.24075950533151627,
      "num_tokens": 80713054.0,
      "step": 35195
    },
    {
      "entropy": 5.093982553482055,
      "epoch": 3.3813640730067243,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003884744106705759,
      "loss": 4.6951,
      "mean_token_accuracy": 0.24300110042095185,
      "num_tokens": 80724483.0,
      "step": 35200
    },
    {
      "entropy": 5.1560145854949955,
      "epoch": 3.381844380403458,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038844480514035727,
      "loss": 4.6826,
      "mean_token_accuracy": 0.2450705423951149,
      "num_tokens": 80736559.0,
      "step": 35205
    },
    {
      "entropy": 5.110504961013794,
      "epoch": 3.382324687800192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000388415196976287,
      "loss": 4.6671,
      "mean_token_accuracy": 0.23996759802103043,
      "num_tokens": 80747938.0,
      "step": 35210
    },
    {
      "entropy": 5.077985715866089,
      "epoch": 3.382804995196926,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003883855861790526,
      "loss": 4.6978,
      "mean_token_accuracy": 0.2472958445549011,
      "num_tokens": 80759347.0,
      "step": 35215
    },
    {
      "entropy": 5.124229049682617,
      "epoch": 3.38328530259366,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003883559727493417,
      "loss": 4.6109,
      "mean_token_accuracy": 0.24816007763147355,
      "num_tokens": 80771055.0,
      "step": 35220
    },
    {
      "entropy": 5.100525808334351,
      "epoch": 3.3837656099903937,
      "grad_norm": 1.0,
      "learning_rate": 0.0003883263566878416,
      "loss": 4.6651,
      "mean_token_accuracy": 0.2441679358482361,
      "num_tokens": 80783000.0,
      "step": 35225
    },
    {
      "entropy": 5.112669658660889,
      "epoch": 3.384245917387128,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038829673799524006,
      "loss": 4.6698,
      "mean_token_accuracy": 0.2453371375799179,
      "num_tokens": 80794435.0,
      "step": 35230
    },
    {
      "entropy": 5.021334886550903,
      "epoch": 3.3847262247838614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038826711667222464,
      "loss": 4.504,
      "mean_token_accuracy": 0.2579538285732269,
      "num_tokens": 80805958.0,
      "step": 35235
    },
    {
      "entropy": 5.1914793968200685,
      "epoch": 3.3852065321805958,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038823749271948315,
      "loss": 4.7222,
      "mean_token_accuracy": 0.23750917315483094,
      "num_tokens": 80816799.0,
      "step": 35240
    },
    {
      "entropy": 5.099913024902344,
      "epoch": 3.3856868395773296,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038820786613770334,
      "loss": 4.6502,
      "mean_token_accuracy": 0.25630719512701033,
      "num_tokens": 80828481.0,
      "step": 35245
    },
    {
      "entropy": 5.069317102432251,
      "epoch": 3.3861671469740635,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038817823692757303,
      "loss": 4.6583,
      "mean_token_accuracy": 0.24924203604459763,
      "num_tokens": 80841155.0,
      "step": 35250
    },
    {
      "entropy": 5.111768674850464,
      "epoch": 3.3866474543707974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038814860508978004,
      "loss": 4.6352,
      "mean_token_accuracy": 0.2505061388015747,
      "num_tokens": 80852058.0,
      "step": 35255
    },
    {
      "entropy": 5.0614642143249515,
      "epoch": 3.3871277617675313,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003881189706250125,
      "loss": 4.5729,
      "mean_token_accuracy": 0.2595520317554474,
      "num_tokens": 80863565.0,
      "step": 35260
    },
    {
      "entropy": 5.046646785736084,
      "epoch": 3.387608069164265,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038808933353395836,
      "loss": 4.6068,
      "mean_token_accuracy": 0.25126185417175295,
      "num_tokens": 80876338.0,
      "step": 35265
    },
    {
      "entropy": 5.10563702583313,
      "epoch": 3.388088376560999,
      "grad_norm": 1.0,
      "learning_rate": 0.00038805969381730564,
      "loss": 4.6963,
      "mean_token_accuracy": 0.24201047271490098,
      "num_tokens": 80887276.0,
      "step": 35270
    },
    {
      "entropy": 5.107291603088379,
      "epoch": 3.388568683957733,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038803005147574265,
      "loss": 4.6184,
      "mean_token_accuracy": 0.2517374694347382,
      "num_tokens": 80897894.0,
      "step": 35275
    },
    {
      "entropy": 5.151433944702148,
      "epoch": 3.389048991354467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003880004065099575,
      "loss": 4.7431,
      "mean_token_accuracy": 0.24191664010286332,
      "num_tokens": 80910223.0,
      "step": 35280
    },
    {
      "entropy": 4.99162073135376,
      "epoch": 3.3895292987512007,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003879707589206383,
      "loss": 4.5252,
      "mean_token_accuracy": 0.2603502064943314,
      "num_tokens": 80921457.0,
      "step": 35285
    },
    {
      "entropy": 5.108259773254394,
      "epoch": 3.3900096061479346,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003879411087084736,
      "loss": 4.6383,
      "mean_token_accuracy": 0.25257123708724977,
      "num_tokens": 80932051.0,
      "step": 35290
    },
    {
      "entropy": 5.000164937973023,
      "epoch": 3.3904899135446684,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038791145587415186,
      "loss": 4.6513,
      "mean_token_accuracy": 0.24614595621824265,
      "num_tokens": 80943434.0,
      "step": 35295
    },
    {
      "entropy": 5.0758874893188475,
      "epoch": 3.3909702209414023,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038788180041836117,
      "loss": 4.6853,
      "mean_token_accuracy": 0.24621425122022628,
      "num_tokens": 80955260.0,
      "step": 35300
    },
    {
      "entropy": 5.182381105422974,
      "epoch": 3.3914505283381366,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038785214234179037,
      "loss": 4.678,
      "mean_token_accuracy": 0.2495055764913559,
      "num_tokens": 80966848.0,
      "step": 35305
    },
    {
      "entropy": 5.152253818511963,
      "epoch": 3.39193083573487,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038782248164512804,
      "loss": 4.6624,
      "mean_token_accuracy": 0.2426937684416771,
      "num_tokens": 80979051.0,
      "step": 35310
    },
    {
      "entropy": 5.0261541366577145,
      "epoch": 3.3924111431316044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038779281832906253,
      "loss": 4.6182,
      "mean_token_accuracy": 0.26069941371679306,
      "num_tokens": 80989879.0,
      "step": 35315
    },
    {
      "entropy": 5.012664842605591,
      "epoch": 3.3928914505283383,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038776315239428275,
      "loss": 4.6117,
      "mean_token_accuracy": 0.2527031511068344,
      "num_tokens": 81000904.0,
      "step": 35320
    },
    {
      "entropy": 5.117339134216309,
      "epoch": 3.393371757925072,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00038773348384147743,
      "loss": 4.6989,
      "mean_token_accuracy": 0.24026124626398088,
      "num_tokens": 81013637.0,
      "step": 35325
    },
    {
      "entropy": 5.206608533859253,
      "epoch": 3.393852065321806,
      "grad_norm": 1.0,
      "learning_rate": 0.0003877038126713354,
      "loss": 4.7363,
      "mean_token_accuracy": 0.2443919599056244,
      "num_tokens": 81025157.0,
      "step": 35330
    },
    {
      "entropy": 5.11647400856018,
      "epoch": 3.39433237271854,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038767413888454537,
      "loss": 4.6909,
      "mean_token_accuracy": 0.2504799500107765,
      "num_tokens": 81036402.0,
      "step": 35335
    },
    {
      "entropy": 4.981196737289428,
      "epoch": 3.394812680115274,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038764446248179665,
      "loss": 4.5736,
      "mean_token_accuracy": 0.2514304146170616,
      "num_tokens": 81048241.0,
      "step": 35340
    },
    {
      "entropy": 5.142996120452881,
      "epoch": 3.3952929875120077,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003876147834637778,
      "loss": 4.7113,
      "mean_token_accuracy": 0.2507734790444374,
      "num_tokens": 81059590.0,
      "step": 35345
    },
    {
      "entropy": 5.142266130447387,
      "epoch": 3.3957732949087416,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038758510183117806,
      "loss": 4.6908,
      "mean_token_accuracy": 0.24457307755947114,
      "num_tokens": 81070516.0,
      "step": 35350
    },
    {
      "entropy": 5.112433052062988,
      "epoch": 3.3962536023054755,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003875554175846866,
      "loss": 4.7115,
      "mean_token_accuracy": 0.24098230749368668,
      "num_tokens": 81081141.0,
      "step": 35355
    },
    {
      "entropy": 5.12908821105957,
      "epoch": 3.3967339097022093,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038752573072499267,
      "loss": 4.6814,
      "mean_token_accuracy": 0.24744635820388794,
      "num_tokens": 81091785.0,
      "step": 35360
    },
    {
      "entropy": 5.105901622772217,
      "epoch": 3.3972142170989432,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038749604125278524,
      "loss": 4.6165,
      "mean_token_accuracy": 0.24515119642019273,
      "num_tokens": 81102500.0,
      "step": 35365
    },
    {
      "entropy": 5.101875162124633,
      "epoch": 3.397694524495677,
      "grad_norm": 1.0,
      "learning_rate": 0.0003874663491687539,
      "loss": 4.7004,
      "mean_token_accuracy": 0.23888498693704605,
      "num_tokens": 81113553.0,
      "step": 35370
    },
    {
      "entropy": 5.022234153747559,
      "epoch": 3.398174831892411,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038743665447358785,
      "loss": 4.6161,
      "mean_token_accuracy": 0.2531415969133377,
      "num_tokens": 81124843.0,
      "step": 35375
    },
    {
      "entropy": 5.125821113586426,
      "epoch": 3.398655139289145,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003874069571679766,
      "loss": 4.7434,
      "mean_token_accuracy": 0.2475829988718033,
      "num_tokens": 81136716.0,
      "step": 35380
    },
    {
      "entropy": 5.144448471069336,
      "epoch": 3.3991354466858787,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038737725725260946,
      "loss": 4.7601,
      "mean_token_accuracy": 0.246502348780632,
      "num_tokens": 81148143.0,
      "step": 35385
    },
    {
      "entropy": 5.137043285369873,
      "epoch": 3.399615754082613,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038734755472817617,
      "loss": 4.6312,
      "mean_token_accuracy": 0.2521165326237679,
      "num_tokens": 81159632.0,
      "step": 35390
    },
    {
      "entropy": 5.107404994964599,
      "epoch": 3.400096061479347,
      "grad_norm": 1.0,
      "learning_rate": 0.00038731784959536626,
      "loss": 4.6817,
      "mean_token_accuracy": 0.24378161430358886,
      "num_tokens": 81171190.0,
      "step": 35395
    },
    {
      "entropy": 5.027223777770996,
      "epoch": 3.400576368876081,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038728814185486944,
      "loss": 4.5426,
      "mean_token_accuracy": 0.25690784603357314,
      "num_tokens": 81181490.0,
      "step": 35400
    },
    {
      "entropy": 5.00314621925354,
      "epoch": 3.4010566762728147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003872584315073753,
      "loss": 4.5791,
      "mean_token_accuracy": 0.2568309888243675,
      "num_tokens": 81191956.0,
      "step": 35405
    },
    {
      "entropy": 5.090681076049805,
      "epoch": 3.4015369836695486,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003872287185535738,
      "loss": 4.665,
      "mean_token_accuracy": 0.25309066772460936,
      "num_tokens": 81204115.0,
      "step": 35410
    },
    {
      "entropy": 4.97905478477478,
      "epoch": 3.4020172910662825,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038719900299415475,
      "loss": 4.5243,
      "mean_token_accuracy": 0.25820490121841433,
      "num_tokens": 81215595.0,
      "step": 35415
    },
    {
      "entropy": 5.011223077774048,
      "epoch": 3.4024975984630164,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003871692848298079,
      "loss": 4.5956,
      "mean_token_accuracy": 0.2539547488093376,
      "num_tokens": 81227858.0,
      "step": 35420
    },
    {
      "entropy": 5.092417764663696,
      "epoch": 3.4029779058597502,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038713956406122334,
      "loss": 4.6706,
      "mean_token_accuracy": 0.25440116226673126,
      "num_tokens": 81239771.0,
      "step": 35425
    },
    {
      "entropy": 5.191079711914062,
      "epoch": 3.403458213256484,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038710984068909116,
      "loss": 4.7251,
      "mean_token_accuracy": 0.24493889659643173,
      "num_tokens": 81251342.0,
      "step": 35430
    },
    {
      "entropy": 5.115623092651367,
      "epoch": 3.403938520653218,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003870801147141014,
      "loss": 4.7036,
      "mean_token_accuracy": 0.25103136748075483,
      "num_tokens": 81262958.0,
      "step": 35435
    },
    {
      "entropy": 5.053038644790649,
      "epoch": 3.404418828049952,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003870503861369441,
      "loss": 4.6385,
      "mean_token_accuracy": 0.24963057041168213,
      "num_tokens": 81274629.0,
      "step": 35440
    },
    {
      "entropy": 5.111803913116455,
      "epoch": 3.4048991354466858,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038702065495830956,
      "loss": 4.6654,
      "mean_token_accuracy": 0.2513851597905159,
      "num_tokens": 81285233.0,
      "step": 35445
    },
    {
      "entropy": 5.1245402812957765,
      "epoch": 3.4053794428434196,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003869909211788881,
      "loss": 4.6961,
      "mean_token_accuracy": 0.24634861946105957,
      "num_tokens": 81297117.0,
      "step": 35450
    },
    {
      "entropy": 5.076699686050415,
      "epoch": 3.4058597502401535,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038696118479936994,
      "loss": 4.6582,
      "mean_token_accuracy": 0.2465658962726593,
      "num_tokens": 81307469.0,
      "step": 35455
    },
    {
      "entropy": 5.120177745819092,
      "epoch": 3.4063400576368874,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038693144582044553,
      "loss": 4.6423,
      "mean_token_accuracy": 0.24447673410177231,
      "num_tokens": 81319873.0,
      "step": 35460
    },
    {
      "entropy": 5.1550414085388185,
      "epoch": 3.4068203650336217,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038690170424280534,
      "loss": 4.7113,
      "mean_token_accuracy": 0.24020100086927415,
      "num_tokens": 81331044.0,
      "step": 35465
    },
    {
      "entropy": 5.084061622619629,
      "epoch": 3.4073006724303556,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003868719600671399,
      "loss": 4.7276,
      "mean_token_accuracy": 0.2487634614109993,
      "num_tokens": 81342070.0,
      "step": 35470
    },
    {
      "entropy": 5.080636405944825,
      "epoch": 3.4077809798270895,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038684221329413965,
      "loss": 4.7165,
      "mean_token_accuracy": 0.23651882112026215,
      "num_tokens": 81353547.0,
      "step": 35475
    },
    {
      "entropy": 5.1688196659088135,
      "epoch": 3.4082612872238234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003868124639244954,
      "loss": 4.7141,
      "mean_token_accuracy": 0.24284528493881224,
      "num_tokens": 81364280.0,
      "step": 35480
    },
    {
      "entropy": 5.0081658363342285,
      "epoch": 3.4087415946205573,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038678271195889766,
      "loss": 4.589,
      "mean_token_accuracy": 0.25972223579883574,
      "num_tokens": 81375923.0,
      "step": 35485
    },
    {
      "entropy": 5.039915657043457,
      "epoch": 3.409221902017291,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038675295739803734,
      "loss": 4.6257,
      "mean_token_accuracy": 0.25159106999635694,
      "num_tokens": 81387288.0,
      "step": 35490
    },
    {
      "entropy": 5.039334726333618,
      "epoch": 3.409702209414025,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038672320024260516,
      "loss": 4.6257,
      "mean_token_accuracy": 0.24970414340496064,
      "num_tokens": 81399228.0,
      "step": 35495
    },
    {
      "entropy": 5.131591844558716,
      "epoch": 3.410182516810759,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038669344049329204,
      "loss": 4.6191,
      "mean_token_accuracy": 0.2588564336299896,
      "num_tokens": 81410353.0,
      "step": 35500
    },
    {
      "entropy": 5.054448652267456,
      "epoch": 3.410662824207493,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038666367815078887,
      "loss": 4.6956,
      "mean_token_accuracy": 0.24370137602090836,
      "num_tokens": 81422344.0,
      "step": 35505
    },
    {
      "entropy": 5.167194843292236,
      "epoch": 3.4111431316042267,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003866339132157867,
      "loss": 4.7437,
      "mean_token_accuracy": 0.2415718913078308,
      "num_tokens": 81434132.0,
      "step": 35510
    },
    {
      "entropy": 5.111783838272094,
      "epoch": 3.4116234390009605,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003866041456889764,
      "loss": 4.6869,
      "mean_token_accuracy": 0.24979924261569977,
      "num_tokens": 81445735.0,
      "step": 35515
    },
    {
      "entropy": 4.985712671279908,
      "epoch": 3.4121037463976944,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038657437557104946,
      "loss": 4.5841,
      "mean_token_accuracy": 0.25548832714557645,
      "num_tokens": 81457513.0,
      "step": 35520
    },
    {
      "entropy": 5.10706729888916,
      "epoch": 3.4125840537944283,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003865446028626967,
      "loss": 4.6998,
      "mean_token_accuracy": 0.24847433418035508,
      "num_tokens": 81469412.0,
      "step": 35525
    },
    {
      "entropy": 5.121689987182617,
      "epoch": 3.413064361191162,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038651482756460947,
      "loss": 4.7776,
      "mean_token_accuracy": 0.24177989065647126,
      "num_tokens": 81482445.0,
      "step": 35530
    },
    {
      "entropy": 5.074722671508789,
      "epoch": 3.413544668587896,
      "grad_norm": 1.0,
      "learning_rate": 0.00038648504967747914,
      "loss": 4.5691,
      "mean_token_accuracy": 0.25709930807352066,
      "num_tokens": 81493073.0,
      "step": 35535
    },
    {
      "entropy": 5.073358488082886,
      "epoch": 3.4140249759846304,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038645526920199697,
      "loss": 4.6048,
      "mean_token_accuracy": 0.2549064576625824,
      "num_tokens": 81503860.0,
      "step": 35540
    },
    {
      "entropy": 4.989995050430298,
      "epoch": 3.414505283381364,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003864254861388544,
      "loss": 4.5117,
      "mean_token_accuracy": 0.2607329323887825,
      "num_tokens": 81514754.0,
      "step": 35545
    },
    {
      "entropy": 5.057687664031983,
      "epoch": 3.414985590778098,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038639570048874295,
      "loss": 4.6067,
      "mean_token_accuracy": 0.25016269236803057,
      "num_tokens": 81526163.0,
      "step": 35550
    },
    {
      "entropy": 5.219683694839477,
      "epoch": 3.415465898174832,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038636591225235407,
      "loss": 4.7995,
      "mean_token_accuracy": 0.23469086736440659,
      "num_tokens": 81538316.0,
      "step": 35555
    },
    {
      "entropy": 5.074326276779175,
      "epoch": 3.415946205571566,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003863361214303794,
      "loss": 4.5895,
      "mean_token_accuracy": 0.2537877008318901,
      "num_tokens": 81549441.0,
      "step": 35560
    },
    {
      "entropy": 5.12176284790039,
      "epoch": 3.4164265129683,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003863063280235106,
      "loss": 4.6942,
      "mean_token_accuracy": 0.2510987237095833,
      "num_tokens": 81559066.0,
      "step": 35565
    },
    {
      "entropy": 5.094378042221069,
      "epoch": 3.4169068203650337,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038627653203243933,
      "loss": 4.7083,
      "mean_token_accuracy": 0.24959832280874253,
      "num_tokens": 81570985.0,
      "step": 35570
    },
    {
      "entropy": 5.0879370212554935,
      "epoch": 3.4173871277617676,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003862467334578574,
      "loss": 4.6086,
      "mean_token_accuracy": 0.25667123347520826,
      "num_tokens": 81582236.0,
      "step": 35575
    },
    {
      "entropy": 4.99848952293396,
      "epoch": 3.4178674351585014,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038621693230045677,
      "loss": 4.61,
      "mean_token_accuracy": 0.25671776533126833,
      "num_tokens": 81594239.0,
      "step": 35580
    },
    {
      "entropy": 5.090468215942383,
      "epoch": 3.4183477425552353,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003861871285609291,
      "loss": 4.6033,
      "mean_token_accuracy": 0.24922395646572112,
      "num_tokens": 81606877.0,
      "step": 35585
    },
    {
      "entropy": 5.126850414276123,
      "epoch": 3.418828049951969,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003861573222399665,
      "loss": 4.6646,
      "mean_token_accuracy": 0.24645139425992965,
      "num_tokens": 81618249.0,
      "step": 35590
    },
    {
      "entropy": 5.008663320541382,
      "epoch": 3.419308357348703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000386127513338261,
      "loss": 4.5492,
      "mean_token_accuracy": 0.2618724197149277,
      "num_tokens": 81629358.0,
      "step": 35595
    },
    {
      "entropy": 5.0673693180084225,
      "epoch": 3.419788664745437,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003860977018565046,
      "loss": 4.6061,
      "mean_token_accuracy": 0.24794525504112244,
      "num_tokens": 81640133.0,
      "step": 35600
    },
    {
      "entropy": 5.014977073669433,
      "epoch": 3.420268972142171,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003860678877953894,
      "loss": 4.5871,
      "mean_token_accuracy": 0.263104347884655,
      "num_tokens": 81652052.0,
      "step": 35605
    },
    {
      "entropy": 5.132188034057617,
      "epoch": 3.4207492795389047,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003860380711556077,
      "loss": 4.6766,
      "mean_token_accuracy": 0.24507495313882827,
      "num_tokens": 81663112.0,
      "step": 35610
    },
    {
      "entropy": 5.110806798934936,
      "epoch": 3.421229586935639,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038600825193785173,
      "loss": 4.6881,
      "mean_token_accuracy": 0.24801841378211975,
      "num_tokens": 81674022.0,
      "step": 35615
    },
    {
      "entropy": 5.123743057250977,
      "epoch": 3.4217098943323725,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003859784301428137,
      "loss": 4.651,
      "mean_token_accuracy": 0.25448338836431506,
      "num_tokens": 81685760.0,
      "step": 35620
    },
    {
      "entropy": 5.080193758010864,
      "epoch": 3.422190201729107,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003859486057711861,
      "loss": 4.5532,
      "mean_token_accuracy": 0.25283658504486084,
      "num_tokens": 81696479.0,
      "step": 35625
    },
    {
      "entropy": 5.058426761627198,
      "epoch": 3.4226705091258407,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003859187788236613,
      "loss": 4.6958,
      "mean_token_accuracy": 0.24173376858234405,
      "num_tokens": 81708426.0,
      "step": 35630
    },
    {
      "entropy": 5.093772029876709,
      "epoch": 3.4231508165225746,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038588894930093184,
      "loss": 4.6699,
      "mean_token_accuracy": 0.24928666800260543,
      "num_tokens": 81720651.0,
      "step": 35635
    },
    {
      "entropy": 5.079576015472412,
      "epoch": 3.4236311239193085,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038585911720369023,
      "loss": 4.6404,
      "mean_token_accuracy": 0.24633868038654327,
      "num_tokens": 81731691.0,
      "step": 35640
    },
    {
      "entropy": 5.0346263408660885,
      "epoch": 3.4241114313160423,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003858292825326291,
      "loss": 4.615,
      "mean_token_accuracy": 0.25262220203876495,
      "num_tokens": 81743567.0,
      "step": 35645
    },
    {
      "entropy": 5.082108116149902,
      "epoch": 3.4245917387127762,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003857994452884412,
      "loss": 4.62,
      "mean_token_accuracy": 0.25841368138790133,
      "num_tokens": 81754978.0,
      "step": 35650
    },
    {
      "entropy": 5.078841018676758,
      "epoch": 3.42507204610951,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003857696054718191,
      "loss": 4.6618,
      "mean_token_accuracy": 0.2412533715367317,
      "num_tokens": 81764911.0,
      "step": 35655
    },
    {
      "entropy": 5.026972532272339,
      "epoch": 3.425552353506244,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003857397630834557,
      "loss": 4.6579,
      "mean_token_accuracy": 0.25179278701543806,
      "num_tokens": 81776756.0,
      "step": 35660
    },
    {
      "entropy": 5.11190185546875,
      "epoch": 3.426032660902978,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038570991812404384,
      "loss": 4.7408,
      "mean_token_accuracy": 0.2428869053721428,
      "num_tokens": 81788592.0,
      "step": 35665
    },
    {
      "entropy": 5.117741394042969,
      "epoch": 3.4265129682997117,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003856800705942764,
      "loss": 4.625,
      "mean_token_accuracy": 0.24913697838783264,
      "num_tokens": 81799866.0,
      "step": 35670
    },
    {
      "entropy": 5.03779296875,
      "epoch": 3.4269932756964456,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038565022049484636,
      "loss": 4.6985,
      "mean_token_accuracy": 0.24533407241106034,
      "num_tokens": 81812210.0,
      "step": 35675
    },
    {
      "entropy": 5.0695325374603275,
      "epoch": 3.4274735830931795,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038562036782644675,
      "loss": 4.6429,
      "mean_token_accuracy": 0.25007969737052915,
      "num_tokens": 81825265.0,
      "step": 35680
    },
    {
      "entropy": 5.119020271301269,
      "epoch": 3.4279538904899134,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003855905125897708,
      "loss": 4.6707,
      "mean_token_accuracy": 0.24720313400030136,
      "num_tokens": 81837062.0,
      "step": 35685
    },
    {
      "entropy": 5.063393926620483,
      "epoch": 3.4284341978866473,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038556065478551147,
      "loss": 4.614,
      "mean_token_accuracy": 0.25321230888366697,
      "num_tokens": 81847773.0,
      "step": 35690
    },
    {
      "entropy": 5.05466947555542,
      "epoch": 3.428914505283381,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000385530794414362,
      "loss": 4.7143,
      "mean_token_accuracy": 0.2489491358399391,
      "num_tokens": 81860419.0,
      "step": 35695
    },
    {
      "entropy": 5.095835399627686,
      "epoch": 3.4293948126801155,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003855009314770157,
      "loss": 4.6294,
      "mean_token_accuracy": 0.25100259482860565,
      "num_tokens": 81871100.0,
      "step": 35700
    },
    {
      "entropy": 5.120609951019287,
      "epoch": 3.4298751200768494,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038547106597416593,
      "loss": 4.6809,
      "mean_token_accuracy": 0.24320197403430938,
      "num_tokens": 81883451.0,
      "step": 35705
    },
    {
      "entropy": 5.0456531047821045,
      "epoch": 3.4303554274735832,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000385441197906506,
      "loss": 4.585,
      "mean_token_accuracy": 0.24654542356729509,
      "num_tokens": 81893526.0,
      "step": 35710
    },
    {
      "entropy": 5.007334184646607,
      "epoch": 3.430835734870317,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038541132727472945,
      "loss": 4.5416,
      "mean_token_accuracy": 0.2509327694773674,
      "num_tokens": 81903810.0,
      "step": 35715
    },
    {
      "entropy": 5.059924983978272,
      "epoch": 3.431316042267051,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038538145407952964,
      "loss": 4.6527,
      "mean_token_accuracy": 0.2458035334944725,
      "num_tokens": 81914464.0,
      "step": 35720
    },
    {
      "entropy": 5.036932134628296,
      "epoch": 3.431796349663785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003853515783216003,
      "loss": 4.5491,
      "mean_token_accuracy": 0.2632763162255287,
      "num_tokens": 81925819.0,
      "step": 35725
    },
    {
      "entropy": 5.054652881622315,
      "epoch": 3.4322766570605188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000385321700001635,
      "loss": 4.6022,
      "mean_token_accuracy": 0.2511858016252518,
      "num_tokens": 81935853.0,
      "step": 35730
    },
    {
      "entropy": 5.120764112472534,
      "epoch": 3.4327569644572526,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003852918191203274,
      "loss": 4.7008,
      "mean_token_accuracy": 0.24836964756250382,
      "num_tokens": 81946475.0,
      "step": 35735
    },
    {
      "entropy": 5.1366204738616945,
      "epoch": 3.4332372718539865,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003852619356783712,
      "loss": 4.6968,
      "mean_token_accuracy": 0.24430963397026062,
      "num_tokens": 81957064.0,
      "step": 35740
    },
    {
      "entropy": 5.056974220275879,
      "epoch": 3.4337175792507204,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003852320496764603,
      "loss": 4.596,
      "mean_token_accuracy": 0.25681490153074266,
      "num_tokens": 81968029.0,
      "step": 35745
    },
    {
      "entropy": 5.069838809967041,
      "epoch": 3.4341978866474543,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038520216111528855,
      "loss": 4.7152,
      "mean_token_accuracy": 0.24832133948802948,
      "num_tokens": 81980004.0,
      "step": 35750
    },
    {
      "entropy": 5.105479145050049,
      "epoch": 3.434678194044188,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003851722699955499,
      "loss": 4.6146,
      "mean_token_accuracy": 0.24775831997394562,
      "num_tokens": 81991169.0,
      "step": 35755
    },
    {
      "entropy": 5.1883704662323,
      "epoch": 3.435158501440922,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003851423763179382,
      "loss": 4.8054,
      "mean_token_accuracy": 0.239046086370945,
      "num_tokens": 82002896.0,
      "step": 35760
    },
    {
      "entropy": 5.067167139053344,
      "epoch": 3.435638808837656,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038511248008314756,
      "loss": 4.6216,
      "mean_token_accuracy": 0.25888352692127226,
      "num_tokens": 82015248.0,
      "step": 35765
    },
    {
      "entropy": 5.057050800323486,
      "epoch": 3.43611911623439,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003850825812918722,
      "loss": 4.6763,
      "mean_token_accuracy": 0.24812908619642257,
      "num_tokens": 82025982.0,
      "step": 35770
    },
    {
      "entropy": 4.993900871276855,
      "epoch": 3.436599423631124,
      "grad_norm": 1.125,
      "learning_rate": 0.0003850526799448061,
      "loss": 4.5725,
      "mean_token_accuracy": 0.2543690577149391,
      "num_tokens": 82037053.0,
      "step": 35775
    },
    {
      "entropy": 5.083329963684082,
      "epoch": 3.437079731027858,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003850227760426436,
      "loss": 4.6466,
      "mean_token_accuracy": 0.2525723472237587,
      "num_tokens": 82048705.0,
      "step": 35780
    },
    {
      "entropy": 5.052794647216797,
      "epoch": 3.437560038424592,
      "grad_norm": 1.0,
      "learning_rate": 0.00038499286958607894,
      "loss": 4.6012,
      "mean_token_accuracy": 0.2520273387432098,
      "num_tokens": 82059373.0,
      "step": 35785
    },
    {
      "entropy": 5.079639625549317,
      "epoch": 3.438040345821326,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003849629605758065,
      "loss": 4.679,
      "mean_token_accuracy": 0.24850576519966125,
      "num_tokens": 82069851.0,
      "step": 35790
    },
    {
      "entropy": 5.102280807495117,
      "epoch": 3.4385206532180597,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003849330490125205,
      "loss": 4.6662,
      "mean_token_accuracy": 0.2416272208094597,
      "num_tokens": 82081167.0,
      "step": 35795
    },
    {
      "entropy": 5.0737368106842045,
      "epoch": 3.4390009606147935,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038490313489691566,
      "loss": 4.6467,
      "mean_token_accuracy": 0.2466164767742157,
      "num_tokens": 82092485.0,
      "step": 35800
    },
    {
      "entropy": 5.000869226455689,
      "epoch": 3.4394812680115274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003848732182296863,
      "loss": 4.5943,
      "mean_token_accuracy": 0.2601304709911346,
      "num_tokens": 82104022.0,
      "step": 35805
    },
    {
      "entropy": 5.09076623916626,
      "epoch": 3.4399615754082613,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038484329901152713,
      "loss": 4.6526,
      "mean_token_accuracy": 0.24942350685596465,
      "num_tokens": 82115086.0,
      "step": 35810
    },
    {
      "entropy": 5.195306873321533,
      "epoch": 3.440441882804995,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038481337724313264,
      "loss": 4.7771,
      "mean_token_accuracy": 0.24053261578083038,
      "num_tokens": 82126406.0,
      "step": 35815
    },
    {
      "entropy": 5.105406951904297,
      "epoch": 3.440922190201729,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003847834529251977,
      "loss": 4.6469,
      "mean_token_accuracy": 0.2459322139620781,
      "num_tokens": 82136874.0,
      "step": 35820
    },
    {
      "entropy": 4.977724218368531,
      "epoch": 3.441402497598463,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038475352605841693,
      "loss": 4.5624,
      "mean_token_accuracy": 0.2566766321659088,
      "num_tokens": 82148686.0,
      "step": 35825
    },
    {
      "entropy": 5.040128707885742,
      "epoch": 3.441882804995197,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003847235966434852,
      "loss": 4.6195,
      "mean_token_accuracy": 0.25262921005487443,
      "num_tokens": 82160608.0,
      "step": 35830
    },
    {
      "entropy": 5.159999084472656,
      "epoch": 3.4423631123919307,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003846936646810974,
      "loss": 4.7066,
      "mean_token_accuracy": 0.2464626356959343,
      "num_tokens": 82172045.0,
      "step": 35835
    },
    {
      "entropy": 5.10194993019104,
      "epoch": 3.4428434197886646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038466373017194834,
      "loss": 4.6192,
      "mean_token_accuracy": 0.2506742164492607,
      "num_tokens": 82184609.0,
      "step": 35840
    },
    {
      "entropy": 5.045957612991333,
      "epoch": 3.4433237271853985,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003846337931167332,
      "loss": 4.6519,
      "mean_token_accuracy": 0.2522635027766228,
      "num_tokens": 82195434.0,
      "step": 35845
    },
    {
      "entropy": 4.99964075088501,
      "epoch": 3.443804034582133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038460385351614683,
      "loss": 4.5739,
      "mean_token_accuracy": 0.2546292841434479,
      "num_tokens": 82206176.0,
      "step": 35850
    },
    {
      "entropy": 5.053601789474487,
      "epoch": 3.4442843419788662,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038457391137088455,
      "loss": 4.6671,
      "mean_token_accuracy": 0.2511560618877411,
      "num_tokens": 82217077.0,
      "step": 35855
    },
    {
      "entropy": 5.107504415512085,
      "epoch": 3.4447646493756006,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038454396668164136,
      "loss": 4.6816,
      "mean_token_accuracy": 0.2531796798110008,
      "num_tokens": 82227677.0,
      "step": 35860
    },
    {
      "entropy": 5.0559018611907955,
      "epoch": 3.4452449567723344,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003845140194491126,
      "loss": 4.5736,
      "mean_token_accuracy": 0.2522871419787407,
      "num_tokens": 82237841.0,
      "step": 35865
    },
    {
      "entropy": 5.108092403411865,
      "epoch": 3.4457252641690683,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038448406967399334,
      "loss": 4.6836,
      "mean_token_accuracy": 0.2533326894044876,
      "num_tokens": 82248977.0,
      "step": 35870
    },
    {
      "entropy": 5.154755163192749,
      "epoch": 3.446205571565802,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038445411735697917,
      "loss": 4.6445,
      "mean_token_accuracy": 0.24899294823408127,
      "num_tokens": 82260081.0,
      "step": 35875
    },
    {
      "entropy": 5.091596460342407,
      "epoch": 3.446685878962536,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003844241624987655,
      "loss": 4.7053,
      "mean_token_accuracy": 0.24618261456489562,
      "num_tokens": 82272498.0,
      "step": 35880
    },
    {
      "entropy": 5.145325803756714,
      "epoch": 3.44716618635927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003843942051000476,
      "loss": 4.6916,
      "mean_token_accuracy": 0.2526966631412506,
      "num_tokens": 82283341.0,
      "step": 35885
    },
    {
      "entropy": 5.095492553710938,
      "epoch": 3.447646493756004,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000384364245161521,
      "loss": 4.6533,
      "mean_token_accuracy": 0.2492258295416832,
      "num_tokens": 82293890.0,
      "step": 35890
    },
    {
      "entropy": 5.04125828742981,
      "epoch": 3.4481268011527377,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003843342826838815,
      "loss": 4.5957,
      "mean_token_accuracy": 0.25987059623003006,
      "num_tokens": 82303238.0,
      "step": 35895
    },
    {
      "entropy": 5.210150575637817,
      "epoch": 3.4486071085494716,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038430431766782463,
      "loss": 4.8069,
      "mean_token_accuracy": 0.23460250198841096,
      "num_tokens": 82314228.0,
      "step": 35900
    },
    {
      "entropy": 5.05591516494751,
      "epoch": 3.4490874159462055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000384274350114046,
      "loss": 4.6314,
      "mean_token_accuracy": 0.2493225783109665,
      "num_tokens": 82325034.0,
      "step": 35905
    },
    {
      "entropy": 5.077441453933716,
      "epoch": 3.4495677233429394,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038424438002324145,
      "loss": 4.7364,
      "mean_token_accuracy": 0.24234847724437714,
      "num_tokens": 82336063.0,
      "step": 35910
    },
    {
      "entropy": 5.194316291809082,
      "epoch": 3.4500480307396733,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038421440739610683,
      "loss": 4.6973,
      "mean_token_accuracy": 0.24313716143369674,
      "num_tokens": 82347587.0,
      "step": 35915
    },
    {
      "entropy": 5.19380669593811,
      "epoch": 3.450528338136407,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038418443223333797,
      "loss": 4.718,
      "mean_token_accuracy": 0.24863756448030472,
      "num_tokens": 82359939.0,
      "step": 35920
    },
    {
      "entropy": 5.00400938987732,
      "epoch": 3.451008645533141,
      "grad_norm": 0.875,
      "learning_rate": 0.0003841544545356308,
      "loss": 4.6182,
      "mean_token_accuracy": 0.2505972504615784,
      "num_tokens": 82372379.0,
      "step": 35925
    },
    {
      "entropy": 5.016212415695191,
      "epoch": 3.451488952929875,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038412447430368125,
      "loss": 4.6004,
      "mean_token_accuracy": 0.24992695748806,
      "num_tokens": 82384169.0,
      "step": 35930
    },
    {
      "entropy": 5.177416801452637,
      "epoch": 3.4519692603266092,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038409449153818556,
      "loss": 4.761,
      "mean_token_accuracy": 0.2394363284111023,
      "num_tokens": 82395002.0,
      "step": 35935
    },
    {
      "entropy": 5.14456582069397,
      "epoch": 3.452449567723343,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038406450623983964,
      "loss": 4.6405,
      "mean_token_accuracy": 0.2494342803955078,
      "num_tokens": 82405371.0,
      "step": 35940
    },
    {
      "entropy": 5.045395994186402,
      "epoch": 3.452929875120077,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038403451840933966,
      "loss": 4.5864,
      "mean_token_accuracy": 0.25370089411735536,
      "num_tokens": 82416948.0,
      "step": 35945
    },
    {
      "entropy": 5.007662057876587,
      "epoch": 3.453410182516811,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038400452804738204,
      "loss": 4.5428,
      "mean_token_accuracy": 0.2606812193989754,
      "num_tokens": 82428230.0,
      "step": 35950
    },
    {
      "entropy": 5.143924808502197,
      "epoch": 3.4538904899135447,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038397453515466297,
      "loss": 4.7058,
      "mean_token_accuracy": 0.2450960174202919,
      "num_tokens": 82439352.0,
      "step": 35955
    },
    {
      "entropy": 5.007000160217285,
      "epoch": 3.4543707973102786,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003839445397318787,
      "loss": 4.606,
      "mean_token_accuracy": 0.25452703684568406,
      "num_tokens": 82451621.0,
      "step": 35960
    },
    {
      "entropy": 5.04740834236145,
      "epoch": 3.4548511047070125,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003839145417797258,
      "loss": 4.6027,
      "mean_token_accuracy": 0.2588473901152611,
      "num_tokens": 82462250.0,
      "step": 35965
    },
    {
      "entropy": 5.032867050170898,
      "epoch": 3.4553314121037464,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003838845412989006,
      "loss": 4.5942,
      "mean_token_accuracy": 0.25449656397104264,
      "num_tokens": 82473575.0,
      "step": 35970
    },
    {
      "entropy": 5.0776654243469235,
      "epoch": 3.4558117195004803,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003838545382900997,
      "loss": 4.6213,
      "mean_token_accuracy": 0.2567702502012253,
      "num_tokens": 82485171.0,
      "step": 35975
    },
    {
      "entropy": 5.063100337982178,
      "epoch": 3.456292026897214,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003838245327540196,
      "loss": 4.5702,
      "mean_token_accuracy": 0.2639460489153862,
      "num_tokens": 82495993.0,
      "step": 35980
    },
    {
      "entropy": 5.110499906539917,
      "epoch": 3.456772334293948,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038379452469135706,
      "loss": 4.729,
      "mean_token_accuracy": 0.2415160894393921,
      "num_tokens": 82507509.0,
      "step": 35985
    },
    {
      "entropy": 5.10394058227539,
      "epoch": 3.457252641690682,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038376451410280864,
      "loss": 4.726,
      "mean_token_accuracy": 0.2437356010079384,
      "num_tokens": 82519716.0,
      "step": 35990
    },
    {
      "entropy": 5.116930723190308,
      "epoch": 3.457732949087416,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038373450098907124,
      "loss": 4.6645,
      "mean_token_accuracy": 0.24498708099126815,
      "num_tokens": 82530169.0,
      "step": 35995
    },
    {
      "entropy": 5.021606016159057,
      "epoch": 3.4582132564841497,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038370448535084156,
      "loss": 4.6287,
      "mean_token_accuracy": 0.25452626347541807,
      "num_tokens": 82541528.0,
      "step": 36000
    },
    {
      "epoch": 3.4582132564841497,
      "eval_entropy": 4.838066857771313,
      "eval_loss": 4.792507171630859,
      "eval_mean_token_accuracy": 0.250081086191628,
      "eval_num_tokens": 82541528.0,
      "eval_runtime": 26.5674,
      "eval_samples_per_second": 1235.159,
      "eval_steps_per_second": 154.4,
      "step": 36000
    },
    {
      "entropy": 5.11722617149353,
      "epoch": 3.4586935638808836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003836744671888165,
      "loss": 4.7415,
      "mean_token_accuracy": 0.23956319093704223,
      "num_tokens": 82553843.0,
      "step": 36005
    },
    {
      "entropy": 5.15065107345581,
      "epoch": 3.459173871277618,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038364444650369306,
      "loss": 4.6429,
      "mean_token_accuracy": 0.24674747735261918,
      "num_tokens": 82565453.0,
      "step": 36010
    },
    {
      "entropy": 5.093298006057739,
      "epoch": 3.4596541786743518,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038361442329616814,
      "loss": 4.6773,
      "mean_token_accuracy": 0.251237453520298,
      "num_tokens": 82576647.0,
      "step": 36015
    },
    {
      "entropy": 5.036974573135376,
      "epoch": 3.4601344860710856,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038358439756693886,
      "loss": 4.5987,
      "mean_token_accuracy": 0.25422193557024003,
      "num_tokens": 82587226.0,
      "step": 36020
    },
    {
      "entropy": 5.144786310195923,
      "epoch": 3.4606147934678195,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038355436931670225,
      "loss": 4.6059,
      "mean_token_accuracy": 0.2497960016131401,
      "num_tokens": 82599121.0,
      "step": 36025
    },
    {
      "entropy": 5.149803304672242,
      "epoch": 3.4610951008645534,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038352433854615557,
      "loss": 4.6134,
      "mean_token_accuracy": 0.26283191740512846,
      "num_tokens": 82609895.0,
      "step": 36030
    },
    {
      "entropy": 4.995981931686401,
      "epoch": 3.4615754082612873,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038349430525599606,
      "loss": 4.6265,
      "mean_token_accuracy": 0.2500505641102791,
      "num_tokens": 82622669.0,
      "step": 36035
    },
    {
      "entropy": 5.11412935256958,
      "epoch": 3.462055715658021,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038346426944692075,
      "loss": 4.7386,
      "mean_token_accuracy": 0.24186610132455827,
      "num_tokens": 82634895.0,
      "step": 36040
    },
    {
      "entropy": 5.092148733139038,
      "epoch": 3.462536023054755,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003834342311196274,
      "loss": 4.6032,
      "mean_token_accuracy": 0.2548403725028038,
      "num_tokens": 82645684.0,
      "step": 36045
    },
    {
      "entropy": 5.119237327575684,
      "epoch": 3.463016330451489,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038340419027481305,
      "loss": 4.6705,
      "mean_token_accuracy": 0.24663204252719878,
      "num_tokens": 82657455.0,
      "step": 36050
    },
    {
      "entropy": 5.175142574310303,
      "epoch": 3.463496637848223,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003833741469131753,
      "loss": 4.809,
      "mean_token_accuracy": 0.23849450200796127,
      "num_tokens": 82668611.0,
      "step": 36055
    },
    {
      "entropy": 5.022582340240478,
      "epoch": 3.4639769452449567,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003833441010354117,
      "loss": 4.552,
      "mean_token_accuracy": 0.2643838614225388,
      "num_tokens": 82680629.0,
      "step": 36060
    },
    {
      "entropy": 5.023467922210694,
      "epoch": 3.4644572526416906,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038331405264221974,
      "loss": 4.6483,
      "mean_token_accuracy": 0.25493009835481645,
      "num_tokens": 82690734.0,
      "step": 36065
    },
    {
      "entropy": 5.088310575485229,
      "epoch": 3.4649375600384245,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003832840017342972,
      "loss": 4.6588,
      "mean_token_accuracy": 0.2501319244503975,
      "num_tokens": 82703672.0,
      "step": 36070
    },
    {
      "entropy": 5.121178150177002,
      "epoch": 3.4654178674351583,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038325394831234164,
      "loss": 4.6789,
      "mean_token_accuracy": 0.24981866478919984,
      "num_tokens": 82715491.0,
      "step": 36075
    },
    {
      "entropy": 5.0753148078918455,
      "epoch": 3.465898174831892,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003832238923770508,
      "loss": 4.5405,
      "mean_token_accuracy": 0.25548545122146604,
      "num_tokens": 82726389.0,
      "step": 36080
    },
    {
      "entropy": 5.02631516456604,
      "epoch": 3.4663784822286265,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003831938339291225,
      "loss": 4.5697,
      "mean_token_accuracy": 0.2563122108578682,
      "num_tokens": 82738217.0,
      "step": 36085
    },
    {
      "entropy": 5.062917613983155,
      "epoch": 3.46685878962536,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038316377296925474,
      "loss": 4.6087,
      "mean_token_accuracy": 0.25574164986610415,
      "num_tokens": 82749473.0,
      "step": 36090
    },
    {
      "entropy": 5.177676725387573,
      "epoch": 3.4673390970220943,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038313370949814537,
      "loss": 4.7687,
      "mean_token_accuracy": 0.23873457610607146,
      "num_tokens": 82760485.0,
      "step": 36095
    },
    {
      "entropy": 5.037305927276611,
      "epoch": 3.467819404418828,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003831036435164923,
      "loss": 4.6221,
      "mean_token_accuracy": 0.25230002105236055,
      "num_tokens": 82772830.0,
      "step": 36100
    },
    {
      "entropy": 5.152843952178955,
      "epoch": 3.468299711815562,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003830735750249936,
      "loss": 4.7026,
      "mean_token_accuracy": 0.24051627963781358,
      "num_tokens": 82784600.0,
      "step": 36105
    },
    {
      "entropy": 5.170223569869995,
      "epoch": 3.468780019212296,
      "grad_norm": 1.0,
      "learning_rate": 0.0003830435040243475,
      "loss": 4.6923,
      "mean_token_accuracy": 0.24938689321279525,
      "num_tokens": 82795662.0,
      "step": 36110
    },
    {
      "entropy": 5.152642250061035,
      "epoch": 3.46926032660903,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000383013430515252,
      "loss": 4.7632,
      "mean_token_accuracy": 0.24946428090333939,
      "num_tokens": 82807271.0,
      "step": 36115
    },
    {
      "entropy": 5.063357925415039,
      "epoch": 3.4697406340057637,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003829833544984054,
      "loss": 4.6428,
      "mean_token_accuracy": 0.2531542256474495,
      "num_tokens": 82818593.0,
      "step": 36120
    },
    {
      "entropy": 5.115762376785279,
      "epoch": 3.4702209414024976,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003829532759745059,
      "loss": 4.6528,
      "mean_token_accuracy": 0.2534619107842445,
      "num_tokens": 82830338.0,
      "step": 36125
    },
    {
      "entropy": 5.0203986167907715,
      "epoch": 3.4707012487992315,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038292319494425195,
      "loss": 4.514,
      "mean_token_accuracy": 0.2642645850777626,
      "num_tokens": 82840863.0,
      "step": 36130
    },
    {
      "entropy": 4.985295391082763,
      "epoch": 3.4711815561959654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003828931114083419,
      "loss": 4.6016,
      "mean_token_accuracy": 0.2584876462817192,
      "num_tokens": 82852432.0,
      "step": 36135
    },
    {
      "entropy": 5.1155023097991945,
      "epoch": 3.4716618635926992,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000382863025367474,
      "loss": 4.6993,
      "mean_token_accuracy": 0.24567800462245942,
      "num_tokens": 82863638.0,
      "step": 36140
    },
    {
      "entropy": 5.078704166412353,
      "epoch": 3.472142170989433,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003828329368223471,
      "loss": 4.6248,
      "mean_token_accuracy": 0.25570025891065595,
      "num_tokens": 82874280.0,
      "step": 36145
    },
    {
      "entropy": 5.11669921875,
      "epoch": 3.472622478386167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003828028457736596,
      "loss": 4.7414,
      "mean_token_accuracy": 0.2539126634597778,
      "num_tokens": 82885105.0,
      "step": 36150
    },
    {
      "entropy": 5.117729187011719,
      "epoch": 3.473102785782901,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003827727522221101,
      "loss": 4.8128,
      "mean_token_accuracy": 0.24135739654302596,
      "num_tokens": 82898090.0,
      "step": 36155
    },
    {
      "entropy": 5.14274001121521,
      "epoch": 3.473583093179635,
      "grad_norm": 1.0,
      "learning_rate": 0.00038274265616839733,
      "loss": 4.6663,
      "mean_token_accuracy": 0.2469375103712082,
      "num_tokens": 82909794.0,
      "step": 36160
    },
    {
      "entropy": 5.082881164550781,
      "epoch": 3.4740634005763686,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038271255761322,
      "loss": 4.6567,
      "mean_token_accuracy": 0.24887287318706514,
      "num_tokens": 82921326.0,
      "step": 36165
    },
    {
      "entropy": 5.0274513244628904,
      "epoch": 3.474543707973103,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003826824565572769,
      "loss": 4.6547,
      "mean_token_accuracy": 0.24120912849903106,
      "num_tokens": 82932850.0,
      "step": 36170
    },
    {
      "entropy": 5.1782660484313965,
      "epoch": 3.475024015369837,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003826523530012669,
      "loss": 4.7577,
      "mean_token_accuracy": 0.24014445543289184,
      "num_tokens": 82944456.0,
      "step": 36175
    },
    {
      "entropy": 5.039668893814087,
      "epoch": 3.4755043227665707,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000382622246945889,
      "loss": 4.4969,
      "mean_token_accuracy": 0.2599898174405098,
      "num_tokens": 82956344.0,
      "step": 36180
    },
    {
      "entropy": 5.102537536621094,
      "epoch": 3.4759846301633046,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038259213839184205,
      "loss": 4.6537,
      "mean_token_accuracy": 0.24906317740678788,
      "num_tokens": 82967401.0,
      "step": 36185
    },
    {
      "entropy": 5.057475137710571,
      "epoch": 3.4764649375600385,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038256202733982515,
      "loss": 4.6582,
      "mean_token_accuracy": 0.2530850037932396,
      "num_tokens": 82977590.0,
      "step": 36190
    },
    {
      "entropy": 5.090711641311645,
      "epoch": 3.4769452449567724,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003825319137905373,
      "loss": 4.6452,
      "mean_token_accuracy": 0.25262016206979754,
      "num_tokens": 82990054.0,
      "step": 36195
    },
    {
      "entropy": 5.112039995193482,
      "epoch": 3.4774255523535063,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003825017977446778,
      "loss": 4.6268,
      "mean_token_accuracy": 0.2590195626020432,
      "num_tokens": 83001996.0,
      "step": 36200
    },
    {
      "entropy": 5.095945262908936,
      "epoch": 3.47790585975024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038247167920294574,
      "loss": 4.6847,
      "mean_token_accuracy": 0.23929998725652696,
      "num_tokens": 83013821.0,
      "step": 36205
    },
    {
      "entropy": 5.025370359420776,
      "epoch": 3.478386167146974,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038244155816604037,
      "loss": 4.6014,
      "mean_token_accuracy": 0.2541297674179077,
      "num_tokens": 83025079.0,
      "step": 36210
    },
    {
      "entropy": 5.055237197875977,
      "epoch": 3.478866474543708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003824114346346611,
      "loss": 4.6077,
      "mean_token_accuracy": 0.2503750368952751,
      "num_tokens": 83036728.0,
      "step": 36215
    },
    {
      "entropy": 5.089639616012573,
      "epoch": 3.479346781940442,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003823813086095073,
      "loss": 4.6653,
      "mean_token_accuracy": 0.2505496948957443,
      "num_tokens": 83048295.0,
      "step": 36220
    },
    {
      "entropy": 5.079287004470825,
      "epoch": 3.4798270893371757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038235118009127833,
      "loss": 4.6941,
      "mean_token_accuracy": 0.24984999746084213,
      "num_tokens": 83059872.0,
      "step": 36225
    },
    {
      "entropy": 5.03549280166626,
      "epoch": 3.4803073967339095,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003823210490806737,
      "loss": 4.5891,
      "mean_token_accuracy": 0.25525650233030317,
      "num_tokens": 83070842.0,
      "step": 36230
    },
    {
      "entropy": 5.155250024795532,
      "epoch": 3.4807877041306434,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000382290915578393,
      "loss": 4.7554,
      "mean_token_accuracy": 0.23989729434251786,
      "num_tokens": 83082919.0,
      "step": 36235
    },
    {
      "entropy": 5.072011661529541,
      "epoch": 3.4812680115273773,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003822607795851359,
      "loss": 4.6335,
      "mean_token_accuracy": 0.25233430713415145,
      "num_tokens": 83093303.0,
      "step": 36240
    },
    {
      "entropy": 4.967645931243896,
      "epoch": 3.4817483189241116,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003822306411016019,
      "loss": 4.5541,
      "mean_token_accuracy": 0.26196960657835006,
      "num_tokens": 83104218.0,
      "step": 36245
    },
    {
      "entropy": 5.092840909957886,
      "epoch": 3.4822286263208455,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003822005001284909,
      "loss": 4.7155,
      "mean_token_accuracy": 0.2447071298956871,
      "num_tokens": 83115573.0,
      "step": 36250
    },
    {
      "entropy": 5.15233039855957,
      "epoch": 3.4827089337175794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003821703566665026,
      "loss": 4.6298,
      "mean_token_accuracy": 0.2571637824177742,
      "num_tokens": 83126722.0,
      "step": 36255
    },
    {
      "entropy": 5.125375366210937,
      "epoch": 3.4831892411143133,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003821402107163368,
      "loss": 4.7172,
      "mean_token_accuracy": 0.23657969683408736,
      "num_tokens": 83138325.0,
      "step": 36260
    },
    {
      "entropy": 5.079007768630982,
      "epoch": 3.483669548511047,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003821100622786935,
      "loss": 4.5925,
      "mean_token_accuracy": 0.2575688973069191,
      "num_tokens": 83149434.0,
      "step": 36265
    },
    {
      "entropy": 5.180754280090332,
      "epoch": 3.484149855907781,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038207991135427255,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23160459995269775,
      "num_tokens": 83162068.0,
      "step": 36270
    },
    {
      "entropy": 5.192983150482178,
      "epoch": 3.484630163304515,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000382049757943774,
      "loss": 4.667,
      "mean_token_accuracy": 0.24925969839096068,
      "num_tokens": 83173993.0,
      "step": 36275
    },
    {
      "entropy": 5.123273944854736,
      "epoch": 3.485110470701249,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038201960204789796,
      "loss": 4.7649,
      "mean_token_accuracy": 0.2458132952451706,
      "num_tokens": 83186715.0,
      "step": 36280
    },
    {
      "entropy": 5.0580586910247805,
      "epoch": 3.4855907780979827,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003819894436673446,
      "loss": 4.6789,
      "mean_token_accuracy": 0.24805424660444259,
      "num_tokens": 83196299.0,
      "step": 36285
    },
    {
      "entropy": 5.104327917098999,
      "epoch": 3.4860710854947166,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038195928280281406,
      "loss": 4.7221,
      "mean_token_accuracy": 0.2479254722595215,
      "num_tokens": 83208361.0,
      "step": 36290
    },
    {
      "entropy": 5.145641088485718,
      "epoch": 3.4865513928914504,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003819291194550066,
      "loss": 4.6435,
      "mean_token_accuracy": 0.2531893253326416,
      "num_tokens": 83220273.0,
      "step": 36295
    },
    {
      "entropy": 5.161718702316284,
      "epoch": 3.4870317002881843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003818989536246224,
      "loss": 4.7428,
      "mean_token_accuracy": 0.24201322793960572,
      "num_tokens": 83231344.0,
      "step": 36300
    },
    {
      "entropy": 5.0649737358093265,
      "epoch": 3.487512007684918,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038186878531236196,
      "loss": 4.6339,
      "mean_token_accuracy": 0.2518267944455147,
      "num_tokens": 83244445.0,
      "step": 36305
    },
    {
      "entropy": 5.089413213729858,
      "epoch": 3.487992315081652,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038183861451892567,
      "loss": 4.6794,
      "mean_token_accuracy": 0.2505049079656601,
      "num_tokens": 83256837.0,
      "step": 36310
    },
    {
      "entropy": 5.103501653671264,
      "epoch": 3.488472622478386,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000381808441245014,
      "loss": 4.7106,
      "mean_token_accuracy": 0.2507676690816879,
      "num_tokens": 83267589.0,
      "step": 36315
    },
    {
      "entropy": 5.133641147613526,
      "epoch": 3.4889529298751203,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003817782654913274,
      "loss": 4.6993,
      "mean_token_accuracy": 0.2540870115160942,
      "num_tokens": 83278461.0,
      "step": 36320
    },
    {
      "entropy": 5.065153503417969,
      "epoch": 3.489433237271854,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003817480872585667,
      "loss": 4.5934,
      "mean_token_accuracy": 0.2532185599207878,
      "num_tokens": 83289970.0,
      "step": 36325
    },
    {
      "entropy": 5.002018117904663,
      "epoch": 3.489913544668588,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038171790654743234,
      "loss": 4.5366,
      "mean_token_accuracy": 0.2545260637998581,
      "num_tokens": 83301360.0,
      "step": 36330
    },
    {
      "entropy": 5.08944411277771,
      "epoch": 3.490393852065322,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000381687723358625,
      "loss": 4.7353,
      "mean_token_accuracy": 0.2448781430721283,
      "num_tokens": 83312164.0,
      "step": 36335
    },
    {
      "entropy": 5.084876680374146,
      "epoch": 3.490874159462056,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003816575376928456,
      "loss": 4.6959,
      "mean_token_accuracy": 0.24717308431863785,
      "num_tokens": 83324462.0,
      "step": 36340
    },
    {
      "entropy": 5.075300550460815,
      "epoch": 3.4913544668587897,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038162734955079483,
      "loss": 4.5711,
      "mean_token_accuracy": 0.24830948412418366,
      "num_tokens": 83335922.0,
      "step": 36345
    },
    {
      "entropy": 5.071036529541016,
      "epoch": 3.4918347742555236,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003815971589331736,
      "loss": 4.6477,
      "mean_token_accuracy": 0.2503596544265747,
      "num_tokens": 83348606.0,
      "step": 36350
    },
    {
      "entropy": 5.150634479522705,
      "epoch": 3.4923150816522575,
      "grad_norm": 1.0,
      "learning_rate": 0.0003815669658406829,
      "loss": 4.732,
      "mean_token_accuracy": 0.243729168176651,
      "num_tokens": 83360929.0,
      "step": 36355
    },
    {
      "entropy": 4.974408435821533,
      "epoch": 3.4927953890489913,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038153677027402374,
      "loss": 4.5285,
      "mean_token_accuracy": 0.26218522936105726,
      "num_tokens": 83372323.0,
      "step": 36360
    },
    {
      "entropy": 5.08025894165039,
      "epoch": 3.493275696445725,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038150657223389703,
      "loss": 4.6524,
      "mean_token_accuracy": 0.25619828402996064,
      "num_tokens": 83383629.0,
      "step": 36365
    },
    {
      "entropy": 5.154058456420898,
      "epoch": 3.493756003842459,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038147637172100397,
      "loss": 4.7315,
      "mean_token_accuracy": 0.24404130429029464,
      "num_tokens": 83395684.0,
      "step": 36370
    },
    {
      "entropy": 5.098542308807373,
      "epoch": 3.494236311239193,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003814461687360459,
      "loss": 4.6345,
      "mean_token_accuracy": 0.25017001330852506,
      "num_tokens": 83406835.0,
      "step": 36375
    },
    {
      "entropy": 5.0488463878631595,
      "epoch": 3.494716618635927,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038141596327972375,
      "loss": 4.6189,
      "mean_token_accuracy": 0.2558071821928024,
      "num_tokens": 83417591.0,
      "step": 36380
    },
    {
      "entropy": 4.995544910430908,
      "epoch": 3.4951969260326607,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003813857553527389,
      "loss": 4.5506,
      "mean_token_accuracy": 0.2571654736995697,
      "num_tokens": 83429998.0,
      "step": 36385
    },
    {
      "entropy": 5.128716659545899,
      "epoch": 3.4956772334293946,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038135554495579273,
      "loss": 4.7347,
      "mean_token_accuracy": 0.23999902456998826,
      "num_tokens": 83441886.0,
      "step": 36390
    },
    {
      "entropy": 5.154507637023926,
      "epoch": 3.496157540826129,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003813253320895866,
      "loss": 4.6912,
      "mean_token_accuracy": 0.24080813229084014,
      "num_tokens": 83453639.0,
      "step": 36395
    },
    {
      "entropy": 5.120505714416504,
      "epoch": 3.4966378482228624,
      "grad_norm": 0.9375,
      "learning_rate": 0.000381295116754822,
      "loss": 4.6835,
      "mean_token_accuracy": 0.2448027178645134,
      "num_tokens": 83465622.0,
      "step": 36400
    },
    {
      "entropy": 5.074857091903686,
      "epoch": 3.4971181556195967,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038126489895220047,
      "loss": 4.6772,
      "mean_token_accuracy": 0.24515732526779174,
      "num_tokens": 83476933.0,
      "step": 36405
    },
    {
      "entropy": 5.036981153488159,
      "epoch": 3.4975984630163306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038123467868242345,
      "loss": 4.5715,
      "mean_token_accuracy": 0.2554553300142288,
      "num_tokens": 83487919.0,
      "step": 36410
    },
    {
      "entropy": 5.041780185699463,
      "epoch": 3.4980787704130645,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038120445594619264,
      "loss": 4.6093,
      "mean_token_accuracy": 0.25497193038463595,
      "num_tokens": 83500037.0,
      "step": 36415
    },
    {
      "entropy": 4.968039226531983,
      "epoch": 3.4985590778097984,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003811742307442098,
      "loss": 4.575,
      "mean_token_accuracy": 0.2517978236079216,
      "num_tokens": 83512234.0,
      "step": 36420
    },
    {
      "entropy": 5.0802709579467775,
      "epoch": 3.4990393852065322,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038114400307717657,
      "loss": 4.5819,
      "mean_token_accuracy": 0.26136612743139265,
      "num_tokens": 83523504.0,
      "step": 36425
    },
    {
      "entropy": 5.144756746292114,
      "epoch": 3.499519692603266,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003811137729457947,
      "loss": 4.7001,
      "mean_token_accuracy": 0.24793779253959655,
      "num_tokens": 83534399.0,
      "step": 36430
    },
    {
      "entropy": 5.071603488922119,
      "epoch": 3.5,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003810835403507662,
      "loss": 4.6353,
      "mean_token_accuracy": 0.24901029467582703,
      "num_tokens": 83545307.0,
      "step": 36435
    },
    {
      "entropy": 5.084565258026123,
      "epoch": 3.500480307396734,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038105330529279285,
      "loss": 4.6862,
      "mean_token_accuracy": 0.2451920509338379,
      "num_tokens": 83555729.0,
      "step": 36440
    },
    {
      "entropy": 5.044751834869385,
      "epoch": 3.5009606147934678,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003810230677725767,
      "loss": 4.5835,
      "mean_token_accuracy": 0.25224649608135224,
      "num_tokens": 83566951.0,
      "step": 36445
    },
    {
      "entropy": 5.131327390670776,
      "epoch": 3.5014409221902016,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038099282779081975,
      "loss": 4.6616,
      "mean_token_accuracy": 0.25085354000329974,
      "num_tokens": 83579121.0,
      "step": 36450
    },
    {
      "entropy": 5.034185743331909,
      "epoch": 3.5019212295869355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038096258534822404,
      "loss": 4.56,
      "mean_token_accuracy": 0.2525182515382767,
      "num_tokens": 83589955.0,
      "step": 36455
    },
    {
      "entropy": 5.072691106796265,
      "epoch": 3.5024015369836694,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038093234044549174,
      "loss": 4.6405,
      "mean_token_accuracy": 0.2558046817779541,
      "num_tokens": 83603331.0,
      "step": 36460
    },
    {
      "entropy": 5.037310791015625,
      "epoch": 3.5028818443804033,
      "grad_norm": 1.0,
      "learning_rate": 0.00038090209308332503,
      "loss": 4.626,
      "mean_token_accuracy": 0.239207461476326,
      "num_tokens": 83615444.0,
      "step": 36465
    },
    {
      "entropy": 5.132150936126709,
      "epoch": 3.5033621517771376,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038087184326242627,
      "loss": 4.6787,
      "mean_token_accuracy": 0.2429385393857956,
      "num_tokens": 83626798.0,
      "step": 36470
    },
    {
      "entropy": 5.135977745056152,
      "epoch": 3.503842459173871,
      "grad_norm": 1.0,
      "learning_rate": 0.0003808415909834976,
      "loss": 4.687,
      "mean_token_accuracy": 0.2514459416270256,
      "num_tokens": 83639132.0,
      "step": 36475
    },
    {
      "entropy": 5.047981452941895,
      "epoch": 3.5043227665706054,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038081133624724147,
      "loss": 4.6887,
      "mean_token_accuracy": 0.24626527577638627,
      "num_tokens": 83650916.0,
      "step": 36480
    },
    {
      "entropy": 5.1005738258361815,
      "epoch": 3.5048030739673393,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00038078107905436035,
      "loss": 4.6621,
      "mean_token_accuracy": 0.2559020519256592,
      "num_tokens": 83662012.0,
      "step": 36485
    },
    {
      "entropy": 5.093000698089599,
      "epoch": 3.505283381364073,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038075081940555666,
      "loss": 4.6544,
      "mean_token_accuracy": 0.24324748814105987,
      "num_tokens": 83674191.0,
      "step": 36490
    },
    {
      "entropy": 5.000794410705566,
      "epoch": 3.505763688760807,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038072055730153297,
      "loss": 4.5432,
      "mean_token_accuracy": 0.2531591147184372,
      "num_tokens": 83685571.0,
      "step": 36495
    },
    {
      "entropy": 5.148982048034668,
      "epoch": 3.506243996157541,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038069029274299184,
      "loss": 4.7149,
      "mean_token_accuracy": 0.2397352933883667,
      "num_tokens": 83696779.0,
      "step": 36500
    },
    {
      "entropy": 5.079680061340332,
      "epoch": 3.506724303554275,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003806600257306359,
      "loss": 4.6001,
      "mean_token_accuracy": 0.25482336431741714,
      "num_tokens": 83708704.0,
      "step": 36505
    },
    {
      "entropy": 4.973321056365966,
      "epoch": 3.5072046109510087,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038062975626516806,
      "loss": 4.5447,
      "mean_token_accuracy": 0.25617378056049345,
      "num_tokens": 83719551.0,
      "step": 36510
    },
    {
      "entropy": 5.136297655105591,
      "epoch": 3.5076849183477425,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038059948434729086,
      "loss": 4.772,
      "mean_token_accuracy": 0.24116030633449553,
      "num_tokens": 83730758.0,
      "step": 36515
    },
    {
      "entropy": 5.0705506801605225,
      "epoch": 3.5081652257444764,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038056920997770713,
      "loss": 4.6544,
      "mean_token_accuracy": 0.24877442717552184,
      "num_tokens": 83742076.0,
      "step": 36520
    },
    {
      "entropy": 5.036797666549683,
      "epoch": 3.5086455331412103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038053893315711997,
      "loss": 4.5915,
      "mean_token_accuracy": 0.24971298277378082,
      "num_tokens": 83752095.0,
      "step": 36525
    },
    {
      "entropy": 5.113968801498413,
      "epoch": 3.509125840537944,
      "grad_norm": 1.078125,
      "learning_rate": 0.000380508653886232,
      "loss": 4.6927,
      "mean_token_accuracy": 0.2538919895887375,
      "num_tokens": 83762652.0,
      "step": 36530
    },
    {
      "entropy": 5.125010442733765,
      "epoch": 3.509606147934678,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003804783721657465,
      "loss": 4.707,
      "mean_token_accuracy": 0.2500274136662483,
      "num_tokens": 83772889.0,
      "step": 36535
    },
    {
      "entropy": 5.05591893196106,
      "epoch": 3.510086455331412,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038044808799636635,
      "loss": 4.6134,
      "mean_token_accuracy": 0.2496625229716301,
      "num_tokens": 83783840.0,
      "step": 36540
    },
    {
      "entropy": 5.065688419342041,
      "epoch": 3.5105667627281463,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003804178013787947,
      "loss": 4.5496,
      "mean_token_accuracy": 0.26345977783203123,
      "num_tokens": 83793763.0,
      "step": 36545
    },
    {
      "entropy": 5.114035606384277,
      "epoch": 3.5110470701248797,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038038751231373477,
      "loss": 4.6879,
      "mean_token_accuracy": 0.2438541293144226,
      "num_tokens": 83806194.0,
      "step": 36550
    },
    {
      "entropy": 5.069313383102417,
      "epoch": 3.511527377521614,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038035722080188974,
      "loss": 4.6413,
      "mean_token_accuracy": 0.2599907472729683,
      "num_tokens": 83817023.0,
      "step": 36555
    },
    {
      "entropy": 5.005367946624756,
      "epoch": 3.5120076849183475,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038032692684396286,
      "loss": 4.6369,
      "mean_token_accuracy": 0.2589208632707596,
      "num_tokens": 83827323.0,
      "step": 36560
    },
    {
      "entropy": 5.000881576538086,
      "epoch": 3.512487992315082,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003802966304406575,
      "loss": 4.6434,
      "mean_token_accuracy": 0.254885396361351,
      "num_tokens": 83838421.0,
      "step": 36565
    },
    {
      "entropy": 5.0703741073608395,
      "epoch": 3.5129682997118157,
      "grad_norm": 1.0,
      "learning_rate": 0.00038026633159267707,
      "loss": 4.5488,
      "mean_token_accuracy": 0.26358395516872407,
      "num_tokens": 83850650.0,
      "step": 36570
    },
    {
      "entropy": 5.077552413940429,
      "epoch": 3.5134486071085496,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038023603030072503,
      "loss": 4.6108,
      "mean_token_accuracy": 0.25405322313308715,
      "num_tokens": 83861331.0,
      "step": 36575
    },
    {
      "entropy": 5.060527801513672,
      "epoch": 3.5139289145052834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003802057265655048,
      "loss": 4.6198,
      "mean_token_accuracy": 0.25126429945230483,
      "num_tokens": 83873740.0,
      "step": 36580
    },
    {
      "entropy": 5.100219297409057,
      "epoch": 3.5144092219020173,
      "grad_norm": 1.0,
      "learning_rate": 0.00038017542038772003,
      "loss": 4.655,
      "mean_token_accuracy": 0.24868163764476775,
      "num_tokens": 83885020.0,
      "step": 36585
    },
    {
      "entropy": 5.155364894866944,
      "epoch": 3.514889529298751,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038014511176807426,
      "loss": 4.7789,
      "mean_token_accuracy": 0.23745510429143907,
      "num_tokens": 83897102.0,
      "step": 36590
    },
    {
      "entropy": 5.088635730743408,
      "epoch": 3.515369836695485,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038011480070727124,
      "loss": 4.6443,
      "mean_token_accuracy": 0.24901620000600816,
      "num_tokens": 83908431.0,
      "step": 36595
    },
    {
      "entropy": 5.028873491287231,
      "epoch": 3.515850144092219,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003800844872060147,
      "loss": 4.5985,
      "mean_token_accuracy": 0.2581069618463516,
      "num_tokens": 83920039.0,
      "step": 36600
    },
    {
      "entropy": 5.099091339111328,
      "epoch": 3.516330451488953,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038005417126500836,
      "loss": 4.6821,
      "mean_token_accuracy": 0.24124883413314818,
      "num_tokens": 83931131.0,
      "step": 36605
    },
    {
      "entropy": 5.024585437774658,
      "epoch": 3.5168107588856867,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038002385288495613,
      "loss": 4.5642,
      "mean_token_accuracy": 0.25346441119909285,
      "num_tokens": 83943601.0,
      "step": 36610
    },
    {
      "entropy": 5.078319454193116,
      "epoch": 3.5172910662824206,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003799935320665619,
      "loss": 4.727,
      "mean_token_accuracy": 0.25004153698682785,
      "num_tokens": 83954209.0,
      "step": 36615
    },
    {
      "entropy": 5.112721920013428,
      "epoch": 3.517771373679155,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037996320881052956,
      "loss": 4.6706,
      "mean_token_accuracy": 0.2541840896010399,
      "num_tokens": 83965313.0,
      "step": 36620
    },
    {
      "entropy": 5.0514708995819095,
      "epoch": 3.5182516810758884,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037993288311756327,
      "loss": 4.6272,
      "mean_token_accuracy": 0.2581818088889122,
      "num_tokens": 83976623.0,
      "step": 36625
    },
    {
      "entropy": 5.111215591430664,
      "epoch": 3.5187319884726227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000379902554988367,
      "loss": 4.6447,
      "mean_token_accuracy": 0.24739768952131272,
      "num_tokens": 83986929.0,
      "step": 36630
    },
    {
      "entropy": 5.08362774848938,
      "epoch": 3.519212295869356,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003798722244236449,
      "loss": 4.5524,
      "mean_token_accuracy": 0.25451887398958206,
      "num_tokens": 83997833.0,
      "step": 36635
    },
    {
      "entropy": 5.083107566833496,
      "epoch": 3.5196926032660905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003798418914241011,
      "loss": 4.6811,
      "mean_token_accuracy": 0.25154520869255065,
      "num_tokens": 84008560.0,
      "step": 36640
    },
    {
      "entropy": 5.1032506942749025,
      "epoch": 3.5201729106628243,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037981155599043997,
      "loss": 4.655,
      "mean_token_accuracy": 0.2519940912723541,
      "num_tokens": 84019776.0,
      "step": 36645
    },
    {
      "entropy": 5.186181592941284,
      "epoch": 3.520653218059558,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003797812181233657,
      "loss": 4.7856,
      "mean_token_accuracy": 0.23799768090248108,
      "num_tokens": 84031194.0,
      "step": 36650
    },
    {
      "entropy": 5.100554084777832,
      "epoch": 3.521133525456292,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037975087782358264,
      "loss": 4.6146,
      "mean_token_accuracy": 0.2516957759857178,
      "num_tokens": 84044751.0,
      "step": 36655
    },
    {
      "entropy": 5.047068119049072,
      "epoch": 3.521613832853026,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003797205350917953,
      "loss": 4.6453,
      "mean_token_accuracy": 0.2521489590406418,
      "num_tokens": 84055980.0,
      "step": 36660
    },
    {
      "entropy": 5.085846471786499,
      "epoch": 3.52209414024976,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037969018992870814,
      "loss": 4.6408,
      "mean_token_accuracy": 0.2503954544663429,
      "num_tokens": 84068369.0,
      "step": 36665
    },
    {
      "entropy": 5.12131781578064,
      "epoch": 3.5225744476464937,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037965984233502553,
      "loss": 4.7005,
      "mean_token_accuracy": 0.24634699076414107,
      "num_tokens": 84081037.0,
      "step": 36670
    },
    {
      "entropy": 5.1365625858306885,
      "epoch": 3.5230547550432276,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003796294923114522,
      "loss": 4.7433,
      "mean_token_accuracy": 0.2449925184249878,
      "num_tokens": 84091786.0,
      "step": 36675
    },
    {
      "entropy": 5.161068058013916,
      "epoch": 3.5235350624399615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037959913985869276,
      "loss": 4.7579,
      "mean_token_accuracy": 0.24172043353319167,
      "num_tokens": 84103605.0,
      "step": 36680
    },
    {
      "entropy": 5.061544799804688,
      "epoch": 3.5240153698366954,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037956878497745185,
      "loss": 4.5638,
      "mean_token_accuracy": 0.2572044789791107,
      "num_tokens": 84113940.0,
      "step": 36685
    },
    {
      "entropy": 5.051302242279053,
      "epoch": 3.5244956772334293,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003795384276684343,
      "loss": 4.587,
      "mean_token_accuracy": 0.2560321196913719,
      "num_tokens": 84125045.0,
      "step": 36690
    },
    {
      "entropy": 5.1738848209381105,
      "epoch": 3.524975984630163,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003795080679323449,
      "loss": 4.7452,
      "mean_token_accuracy": 0.25118776559829714,
      "num_tokens": 84137190.0,
      "step": 36695
    },
    {
      "entropy": 5.011337852478027,
      "epoch": 3.525456292026897,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003794777057698884,
      "loss": 4.5879,
      "mean_token_accuracy": 0.25955586731433866,
      "num_tokens": 84148465.0,
      "step": 36700
    },
    {
      "entropy": 5.113687038421631,
      "epoch": 3.5259365994236314,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003794473411817699,
      "loss": 4.7206,
      "mean_token_accuracy": 0.2512251675128937,
      "num_tokens": 84159752.0,
      "step": 36705
    },
    {
      "entropy": 5.108015394210815,
      "epoch": 3.526416906820365,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003794169741686942,
      "loss": 4.5929,
      "mean_token_accuracy": 0.25025183856487276,
      "num_tokens": 84171835.0,
      "step": 36710
    },
    {
      "entropy": 5.078171968460083,
      "epoch": 3.526897214217099,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003793866047313665,
      "loss": 4.6463,
      "mean_token_accuracy": 0.25045702904462813,
      "num_tokens": 84182487.0,
      "step": 36715
    },
    {
      "entropy": 5.014062690734863,
      "epoch": 3.527377521613833,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037935623287049173,
      "loss": 4.6075,
      "mean_token_accuracy": 0.2554659277200699,
      "num_tokens": 84194607.0,
      "step": 36720
    },
    {
      "entropy": 5.1786425590515135,
      "epoch": 3.527857829010567,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037932585858677515,
      "loss": 4.731,
      "mean_token_accuracy": 0.2430495500564575,
      "num_tokens": 84205599.0,
      "step": 36725
    },
    {
      "entropy": 5.114451742172241,
      "epoch": 3.5283381364073008,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037929548188092186,
      "loss": 4.7191,
      "mean_token_accuracy": 0.24889148026704788,
      "num_tokens": 84217576.0,
      "step": 36730
    },
    {
      "entropy": 5.026782512664795,
      "epoch": 3.5288184438040346,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037926510275363725,
      "loss": 4.5879,
      "mean_token_accuracy": 0.2567594930529594,
      "num_tokens": 84228774.0,
      "step": 36735
    },
    {
      "entropy": 5.1033789157867435,
      "epoch": 3.5292987512007685,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003792347212056265,
      "loss": 4.6983,
      "mean_token_accuracy": 0.25046224147081375,
      "num_tokens": 84240187.0,
      "step": 36740
    },
    {
      "entropy": 5.043985319137573,
      "epoch": 3.5297790585975024,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037920433723759506,
      "loss": 4.5902,
      "mean_token_accuracy": 0.25571478307247164,
      "num_tokens": 84250965.0,
      "step": 36745
    },
    {
      "entropy": 5.182170295715332,
      "epoch": 3.5302593659942363,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003791739508502483,
      "loss": 4.6986,
      "mean_token_accuracy": 0.23960365056991578,
      "num_tokens": 84262048.0,
      "step": 36750
    },
    {
      "entropy": 5.076257371902466,
      "epoch": 3.53073967339097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037914356204429183,
      "loss": 4.6129,
      "mean_token_accuracy": 0.25314805656671524,
      "num_tokens": 84273455.0,
      "step": 36755
    },
    {
      "entropy": 5.0670084953308105,
      "epoch": 3.531219980787704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037911317082043103,
      "loss": 4.6679,
      "mean_token_accuracy": 0.25066041499376296,
      "num_tokens": 84284034.0,
      "step": 36760
    },
    {
      "entropy": 5.002759647369385,
      "epoch": 3.531700288184438,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003790827771793715,
      "loss": 4.6043,
      "mean_token_accuracy": 0.25627346336841583,
      "num_tokens": 84296044.0,
      "step": 36765
    },
    {
      "entropy": 5.060204410552979,
      "epoch": 3.532180595581172,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000379052381121819,
      "loss": 4.6172,
      "mean_token_accuracy": 0.2508239015936852,
      "num_tokens": 84308590.0,
      "step": 36770
    },
    {
      "entropy": 5.060040521621704,
      "epoch": 3.5326609029779057,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003790219826484792,
      "loss": 4.548,
      "mean_token_accuracy": 0.2559355929493904,
      "num_tokens": 84319089.0,
      "step": 36775
    },
    {
      "entropy": 5.131367874145508,
      "epoch": 3.53314121037464,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037899158176005784,
      "loss": 4.6184,
      "mean_token_accuracy": 0.25430234521627426,
      "num_tokens": 84331120.0,
      "step": 36780
    },
    {
      "entropy": 5.004497766494751,
      "epoch": 3.5336215177713735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037896117845726067,
      "loss": 4.6668,
      "mean_token_accuracy": 0.25458366274833677,
      "num_tokens": 84343103.0,
      "step": 36785
    },
    {
      "entropy": 5.027828311920166,
      "epoch": 3.534101825168108,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037893077274079375,
      "loss": 4.5979,
      "mean_token_accuracy": 0.25008914768695834,
      "num_tokens": 84353818.0,
      "step": 36790
    },
    {
      "entropy": 5.084591102600098,
      "epoch": 3.5345821325648417,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037890036461136285,
      "loss": 4.6317,
      "mean_token_accuracy": 0.2583279117941856,
      "num_tokens": 84365981.0,
      "step": 36795
    },
    {
      "entropy": 5.131192827224732,
      "epoch": 3.5350624399615755,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000378869954069674,
      "loss": 4.6919,
      "mean_token_accuracy": 0.2484836310148239,
      "num_tokens": 84376095.0,
      "step": 36800
    },
    {
      "entropy": 5.068400239944458,
      "epoch": 3.5355427473583094,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003788395411164332,
      "loss": 4.6269,
      "mean_token_accuracy": 0.25151005685329436,
      "num_tokens": 84387273.0,
      "step": 36805
    },
    {
      "entropy": 5.054103994369507,
      "epoch": 3.5360230547550433,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037880912575234666,
      "loss": 4.619,
      "mean_token_accuracy": 0.2619991585612297,
      "num_tokens": 84398075.0,
      "step": 36810
    },
    {
      "entropy": 5.141785955429077,
      "epoch": 3.536503362151777,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003787787079781205,
      "loss": 4.6634,
      "mean_token_accuracy": 0.2462481066584587,
      "num_tokens": 84408883.0,
      "step": 36815
    },
    {
      "entropy": 5.093524551391601,
      "epoch": 3.536983669548511,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003787482877944608,
      "loss": 4.6758,
      "mean_token_accuracy": 0.2506498873233795,
      "num_tokens": 84419848.0,
      "step": 36820
    },
    {
      "entropy": 5.09059772491455,
      "epoch": 3.537463976945245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000378717865202074,
      "loss": 4.6973,
      "mean_token_accuracy": 0.24778785258531572,
      "num_tokens": 84431385.0,
      "step": 36825
    },
    {
      "entropy": 5.053072929382324,
      "epoch": 3.537944284341979,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037868744020166635,
      "loss": 4.6153,
      "mean_token_accuracy": 0.2549866482615471,
      "num_tokens": 84442569.0,
      "step": 36830
    },
    {
      "entropy": 5.015812587738037,
      "epoch": 3.5384245917387127,
      "grad_norm": 1.0,
      "learning_rate": 0.00037865701279394414,
      "loss": 4.619,
      "mean_token_accuracy": 0.251385198533535,
      "num_tokens": 84455587.0,
      "step": 36835
    },
    {
      "entropy": 5.018901109695435,
      "epoch": 3.5389048991354466,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037862658297961393,
      "loss": 4.6396,
      "mean_token_accuracy": 0.24822321832180022,
      "num_tokens": 84467378.0,
      "step": 36840
    },
    {
      "entropy": 5.117156553268432,
      "epoch": 3.5393852065321805,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003785961507593823,
      "loss": 4.6915,
      "mean_token_accuracy": 0.25221733152866366,
      "num_tokens": 84479430.0,
      "step": 36845
    },
    {
      "entropy": 5.19115252494812,
      "epoch": 3.5398655139289144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003785657161339555,
      "loss": 4.853,
      "mean_token_accuracy": 0.23352140635252,
      "num_tokens": 84492052.0,
      "step": 36850
    },
    {
      "entropy": 5.109949445724487,
      "epoch": 3.5403458213256487,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037853527910404033,
      "loss": 4.6943,
      "mean_token_accuracy": 0.24694526940584183,
      "num_tokens": 84503714.0,
      "step": 36855
    },
    {
      "entropy": 5.078173351287842,
      "epoch": 3.540826128722382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003785048396703435,
      "loss": 4.6505,
      "mean_token_accuracy": 0.25500172525644305,
      "num_tokens": 84514095.0,
      "step": 36860
    },
    {
      "entropy": 5.11562066078186,
      "epoch": 3.5413064361191164,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037847439783357165,
      "loss": 4.7245,
      "mean_token_accuracy": 0.24837725907564162,
      "num_tokens": 84525574.0,
      "step": 36865
    },
    {
      "entropy": 5.116574096679687,
      "epoch": 3.54178674351585,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003784439535944314,
      "loss": 4.65,
      "mean_token_accuracy": 0.2445897117257118,
      "num_tokens": 84538032.0,
      "step": 36870
    },
    {
      "entropy": 5.043939542770386,
      "epoch": 3.542267050912584,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003784135069536298,
      "loss": 4.5612,
      "mean_token_accuracy": 0.25666681081056597,
      "num_tokens": 84548380.0,
      "step": 36875
    },
    {
      "entropy": 5.132465505599976,
      "epoch": 3.542747358309318,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037838305791187363,
      "loss": 4.7742,
      "mean_token_accuracy": 0.24348633587360383,
      "num_tokens": 84560252.0,
      "step": 36880
    },
    {
      "entropy": 5.0812300682067875,
      "epoch": 3.543227665706052,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037835260646986986,
      "loss": 4.6144,
      "mean_token_accuracy": 0.25557042062282564,
      "num_tokens": 84572400.0,
      "step": 36885
    },
    {
      "entropy": 5.1480710983276365,
      "epoch": 3.543707973102786,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037832215262832545,
      "loss": 4.6894,
      "mean_token_accuracy": 0.24463952332735062,
      "num_tokens": 84582858.0,
      "step": 36890
    },
    {
      "entropy": 5.08629207611084,
      "epoch": 3.5441882804995197,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003782916963879474,
      "loss": 4.6,
      "mean_token_accuracy": 0.2635972946882248,
      "num_tokens": 84594860.0,
      "step": 36895
    },
    {
      "entropy": 5.095694494247437,
      "epoch": 3.5446685878962536,
      "grad_norm": 1.09375,
      "learning_rate": 0.000378261237749443,
      "loss": 4.6688,
      "mean_token_accuracy": 0.24842114746570587,
      "num_tokens": 84607320.0,
      "step": 36900
    },
    {
      "entropy": 5.130174875259399,
      "epoch": 3.5451488952929875,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003782307767135192,
      "loss": 4.7421,
      "mean_token_accuracy": 0.24644280672073365,
      "num_tokens": 84619006.0,
      "step": 36905
    },
    {
      "entropy": 5.182825231552124,
      "epoch": 3.5456292026897214,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037820031328088325,
      "loss": 4.726,
      "mean_token_accuracy": 0.24514499306678772,
      "num_tokens": 84630904.0,
      "step": 36910
    },
    {
      "entropy": 5.103760623931885,
      "epoch": 3.5461095100864553,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003781698474522425,
      "loss": 4.66,
      "mean_token_accuracy": 0.25196399837732314,
      "num_tokens": 84641374.0,
      "step": 36915
    },
    {
      "entropy": 5.030815458297729,
      "epoch": 3.546589817483189,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037813937922830435,
      "loss": 4.5611,
      "mean_token_accuracy": 0.25622759759426117,
      "num_tokens": 84652640.0,
      "step": 36920
    },
    {
      "entropy": 5.030396938323975,
      "epoch": 3.547070124879923,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037810890860977595,
      "loss": 4.6128,
      "mean_token_accuracy": 0.25398995280265807,
      "num_tokens": 84663929.0,
      "step": 36925
    },
    {
      "entropy": 5.08392333984375,
      "epoch": 3.547550432276657,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037807843559736494,
      "loss": 4.6715,
      "mean_token_accuracy": 0.2440377801656723,
      "num_tokens": 84675017.0,
      "step": 36930
    },
    {
      "entropy": 5.078500175476075,
      "epoch": 3.5480307396733908,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037804796019177874,
      "loss": 4.6678,
      "mean_token_accuracy": 0.2465991348028183,
      "num_tokens": 84686633.0,
      "step": 36935
    },
    {
      "entropy": 5.14223051071167,
      "epoch": 3.548511047070125,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003780174823937248,
      "loss": 4.7019,
      "mean_token_accuracy": 0.24328419119119643,
      "num_tokens": 84696987.0,
      "step": 36940
    },
    {
      "entropy": 5.110996913909912,
      "epoch": 3.5489913544668585,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037798700220391087,
      "loss": 4.6368,
      "mean_token_accuracy": 0.2574117824435234,
      "num_tokens": 84707837.0,
      "step": 36945
    },
    {
      "entropy": 5.039188528060913,
      "epoch": 3.549471661863593,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037795651962304456,
      "loss": 4.6178,
      "mean_token_accuracy": 0.2560281068086624,
      "num_tokens": 84719115.0,
      "step": 36950
    },
    {
      "entropy": 5.031380319595337,
      "epoch": 3.5499519692603267,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037792603465183354,
      "loss": 4.6508,
      "mean_token_accuracy": 0.2466077208518982,
      "num_tokens": 84730588.0,
      "step": 36955
    },
    {
      "entropy": 5.102978372573853,
      "epoch": 3.5504322766570606,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003778955472909857,
      "loss": 4.6251,
      "mean_token_accuracy": 0.25075575709342957,
      "num_tokens": 84742334.0,
      "step": 36960
    },
    {
      "entropy": 5.1673510551452635,
      "epoch": 3.5509125840537945,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037786505754120874,
      "loss": 4.7598,
      "mean_token_accuracy": 0.23687791526317598,
      "num_tokens": 84753157.0,
      "step": 36965
    },
    {
      "entropy": 5.069588613510132,
      "epoch": 3.5513928914505284,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037783456540321065,
      "loss": 4.6117,
      "mean_token_accuracy": 0.2484051451086998,
      "num_tokens": 84764718.0,
      "step": 36970
    },
    {
      "entropy": 5.024899291992187,
      "epoch": 3.5518731988472623,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037780407087769925,
      "loss": 4.5998,
      "mean_token_accuracy": 0.2556925430893898,
      "num_tokens": 84775836.0,
      "step": 36975
    },
    {
      "entropy": 5.065531492233276,
      "epoch": 3.552353506243996,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003777735739653826,
      "loss": 4.6905,
      "mean_token_accuracy": 0.2479138806462288,
      "num_tokens": 84787585.0,
      "step": 36980
    },
    {
      "entropy": 5.071279001235962,
      "epoch": 3.55283381364073,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003777430746669687,
      "loss": 4.6742,
      "mean_token_accuracy": 0.2456093043088913,
      "num_tokens": 84798356.0,
      "step": 36985
    },
    {
      "entropy": 5.153820514678955,
      "epoch": 3.553314121037464,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037771257298316573,
      "loss": 4.6947,
      "mean_token_accuracy": 0.25190759152173997,
      "num_tokens": 84809476.0,
      "step": 36990
    },
    {
      "entropy": 5.066720867156983,
      "epoch": 3.553794428434198,
      "grad_norm": 1.0,
      "learning_rate": 0.0003776820689146818,
      "loss": 4.6235,
      "mean_token_accuracy": 0.2515781044960022,
      "num_tokens": 84820467.0,
      "step": 36995
    },
    {
      "entropy": 5.170472097396851,
      "epoch": 3.5542747358309317,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037765156246222513,
      "loss": 4.7366,
      "mean_token_accuracy": 0.24037210643291473,
      "num_tokens": 84832725.0,
      "step": 37000
    },
    {
      "entropy": 5.072093057632446,
      "epoch": 3.5547550432276656,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037762105362650403,
      "loss": 4.6318,
      "mean_token_accuracy": 0.25407868921756743,
      "num_tokens": 84845068.0,
      "step": 37005
    },
    {
      "entropy": 5.073831033706665,
      "epoch": 3.5552353506243994,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037759054240822675,
      "loss": 4.6534,
      "mean_token_accuracy": 0.2534131556749344,
      "num_tokens": 84856463.0,
      "step": 37010
    },
    {
      "entropy": 5.114542388916016,
      "epoch": 3.5557156580211338,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003775600288081018,
      "loss": 4.6624,
      "mean_token_accuracy": 0.25050514489412307,
      "num_tokens": 84866922.0,
      "step": 37015
    },
    {
      "entropy": 5.150750064849854,
      "epoch": 3.556195965417867,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003775295128268374,
      "loss": 4.7642,
      "mean_token_accuracy": 0.24165007323026658,
      "num_tokens": 84877642.0,
      "step": 37020
    },
    {
      "entropy": 5.1161542415618895,
      "epoch": 3.5566762728146015,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00037749899446514216,
      "loss": 4.7644,
      "mean_token_accuracy": 0.23605671226978303,
      "num_tokens": 84890022.0,
      "step": 37025
    },
    {
      "entropy": 5.064982795715332,
      "epoch": 3.5571565802113354,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003774684737237247,
      "loss": 4.6014,
      "mean_token_accuracy": 0.2549568757414818,
      "num_tokens": 84902831.0,
      "step": 37030
    },
    {
      "entropy": 5.009907627105713,
      "epoch": 3.5576368876080693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003774379506032935,
      "loss": 4.4761,
      "mean_token_accuracy": 0.2617789193987846,
      "num_tokens": 84913734.0,
      "step": 37035
    },
    {
      "entropy": 5.056560850143432,
      "epoch": 3.558117195004803,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037740742510455726,
      "loss": 4.6306,
      "mean_token_accuracy": 0.24979381412267684,
      "num_tokens": 84926384.0,
      "step": 37040
    },
    {
      "entropy": 5.121348094940186,
      "epoch": 3.558597502401537,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003773768972282248,
      "loss": 4.6983,
      "mean_token_accuracy": 0.24870953857898712,
      "num_tokens": 84937129.0,
      "step": 37045
    },
    {
      "entropy": 5.088742685317993,
      "epoch": 3.559077809798271,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003773463669750046,
      "loss": 4.7059,
      "mean_token_accuracy": 0.24439526200294495,
      "num_tokens": 84948009.0,
      "step": 37050
    },
    {
      "entropy": 5.088961935043335,
      "epoch": 3.559558117195005,
      "grad_norm": 1.0,
      "learning_rate": 0.0003773158343456058,
      "loss": 4.6307,
      "mean_token_accuracy": 0.25386869311332705,
      "num_tokens": 84958870.0,
      "step": 37055
    },
    {
      "entropy": 5.233843374252319,
      "epoch": 3.5600384245917387,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037728529934073714,
      "loss": 4.8096,
      "mean_token_accuracy": 0.23361001461744307,
      "num_tokens": 84969873.0,
      "step": 37060
    },
    {
      "entropy": 5.09761414527893,
      "epoch": 3.5605187319884726,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003772547619611075,
      "loss": 4.6863,
      "mean_token_accuracy": 0.24947050511837005,
      "num_tokens": 84981785.0,
      "step": 37065
    },
    {
      "entropy": 5.130055570602417,
      "epoch": 3.5609990393852065,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003772242222074259,
      "loss": 4.6937,
      "mean_token_accuracy": 0.242839315533638,
      "num_tokens": 84992449.0,
      "step": 37070
    },
    {
      "entropy": 5.096578645706177,
      "epoch": 3.5614793467819403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037719368008040145,
      "loss": 4.6724,
      "mean_token_accuracy": 0.24725667834281922,
      "num_tokens": 85004121.0,
      "step": 37075
    },
    {
      "entropy": 5.110057735443116,
      "epoch": 3.561959654178674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037716313558074327,
      "loss": 4.6648,
      "mean_token_accuracy": 0.2527973115444183,
      "num_tokens": 85015590.0,
      "step": 37080
    },
    {
      "entropy": 5.146784734725952,
      "epoch": 3.562439961575408,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003771325887091603,
      "loss": 4.7459,
      "mean_token_accuracy": 0.24695681035518646,
      "num_tokens": 85027178.0,
      "step": 37085
    },
    {
      "entropy": 5.054149103164673,
      "epoch": 3.5629202689721424,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000377102039466362,
      "loss": 4.6337,
      "mean_token_accuracy": 0.26136402040719986,
      "num_tokens": 85039606.0,
      "step": 37090
    },
    {
      "entropy": 5.171673917770386,
      "epoch": 3.563400576368876,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003770714878530575,
      "loss": 4.7584,
      "mean_token_accuracy": 0.23984458297491074,
      "num_tokens": 85051792.0,
      "step": 37095
    },
    {
      "entropy": 5.073290252685547,
      "epoch": 3.56388088376561,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037704093386995607,
      "loss": 4.6131,
      "mean_token_accuracy": 0.2566264942288399,
      "num_tokens": 85063798.0,
      "step": 37100
    },
    {
      "entropy": 5.117003774642944,
      "epoch": 3.5643611911623436,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003770103775177673,
      "loss": 4.6028,
      "mean_token_accuracy": 0.26112805902957914,
      "num_tokens": 85074557.0,
      "step": 37105
    },
    {
      "entropy": 5.106589317321777,
      "epoch": 3.564841498559078,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037697981879720034,
      "loss": 4.7096,
      "mean_token_accuracy": 0.24336211532354354,
      "num_tokens": 85086640.0,
      "step": 37110
    },
    {
      "entropy": 5.042007732391357,
      "epoch": 3.565321805955812,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037694925770896486,
      "loss": 4.5147,
      "mean_token_accuracy": 0.2611681506037712,
      "num_tokens": 85097171.0,
      "step": 37115
    },
    {
      "entropy": 5.123412322998047,
      "epoch": 3.5658021133525457,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037691869425377037,
      "loss": 4.7115,
      "mean_token_accuracy": 0.2471804216504097,
      "num_tokens": 85108378.0,
      "step": 37120
    },
    {
      "entropy": 5.1445152282714846,
      "epoch": 3.5662824207492796,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003768881284323263,
      "loss": 4.716,
      "mean_token_accuracy": 0.23971171230077742,
      "num_tokens": 85120195.0,
      "step": 37125
    },
    {
      "entropy": 5.0118225574493405,
      "epoch": 3.5667627281460135,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003768575602453426,
      "loss": 4.5659,
      "mean_token_accuracy": 0.26260753720998764,
      "num_tokens": 85130475.0,
      "step": 37130
    },
    {
      "entropy": 4.959185695648193,
      "epoch": 3.5672430355427474,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037682698969352876,
      "loss": 4.5073,
      "mean_token_accuracy": 0.2651381492614746,
      "num_tokens": 85142355.0,
      "step": 37135
    },
    {
      "entropy": 5.026103973388672,
      "epoch": 3.5677233429394812,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003767964167775946,
      "loss": 4.5646,
      "mean_token_accuracy": 0.2579309016466141,
      "num_tokens": 85152975.0,
      "step": 37140
    },
    {
      "entropy": 5.126038455963135,
      "epoch": 3.568203650336215,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003767658414982498,
      "loss": 4.6682,
      "mean_token_accuracy": 0.24799676090478898,
      "num_tokens": 85163501.0,
      "step": 37145
    },
    {
      "entropy": 5.131547641754151,
      "epoch": 3.568683957732949,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003767352638562044,
      "loss": 4.7453,
      "mean_token_accuracy": 0.2453390419483185,
      "num_tokens": 85175696.0,
      "step": 37150
    },
    {
      "entropy": 5.0793821811676025,
      "epoch": 3.569164265129683,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003767046838521683,
      "loss": 4.6597,
      "mean_token_accuracy": 0.2534979835152626,
      "num_tokens": 85187565.0,
      "step": 37155
    },
    {
      "entropy": 4.998664712905883,
      "epoch": 3.5696445725264168,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003766741014868515,
      "loss": 4.5135,
      "mean_token_accuracy": 0.2590597227215767,
      "num_tokens": 85198549.0,
      "step": 37160
    },
    {
      "entropy": 5.029041862487793,
      "epoch": 3.570124879923151,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037664351676096385,
      "loss": 4.5945,
      "mean_token_accuracy": 0.25564354062080386,
      "num_tokens": 85210452.0,
      "step": 37165
    },
    {
      "entropy": 5.126746129989624,
      "epoch": 3.5706051873198845,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037661292967521563,
      "loss": 4.7242,
      "mean_token_accuracy": 0.24574075341224672,
      "num_tokens": 85223384.0,
      "step": 37170
    },
    {
      "entropy": 5.172893810272217,
      "epoch": 3.571085494716619,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037658234023031694,
      "loss": 4.716,
      "mean_token_accuracy": 0.24297018647193908,
      "num_tokens": 85234117.0,
      "step": 37175
    },
    {
      "entropy": 5.1059094905853275,
      "epoch": 3.5715658021133523,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003765517484269779,
      "loss": 4.6704,
      "mean_token_accuracy": 0.2505803808569908,
      "num_tokens": 85245850.0,
      "step": 37180
    },
    {
      "entropy": 5.053035163879395,
      "epoch": 3.5720461095100866,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037652115426590874,
      "loss": 4.6684,
      "mean_token_accuracy": 0.25507454127073287,
      "num_tokens": 85258476.0,
      "step": 37185
    },
    {
      "entropy": 5.0444536209106445,
      "epoch": 3.5725264169068205,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037649055774782,
      "loss": 4.5993,
      "mean_token_accuracy": 0.2518114119768143,
      "num_tokens": 85269291.0,
      "step": 37190
    },
    {
      "entropy": 5.159016799926758,
      "epoch": 3.5730067243035544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003764599588734218,
      "loss": 4.6788,
      "mean_token_accuracy": 0.25475382804870605,
      "num_tokens": 85280515.0,
      "step": 37195
    },
    {
      "entropy": 5.1930326461792,
      "epoch": 3.5734870317002883,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037642935764342454,
      "loss": 4.6619,
      "mean_token_accuracy": 0.24712493270635605,
      "num_tokens": 85291980.0,
      "step": 37200
    },
    {
      "entropy": 5.08789963722229,
      "epoch": 3.573967339097022,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003763987540585389,
      "loss": 4.7267,
      "mean_token_accuracy": 0.248262919485569,
      "num_tokens": 85303383.0,
      "step": 37205
    },
    {
      "entropy": 5.1748607635498045,
      "epoch": 3.574447646493756,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003763681481194752,
      "loss": 4.802,
      "mean_token_accuracy": 0.22915845066308976,
      "num_tokens": 85316475.0,
      "step": 37210
    },
    {
      "entropy": 5.099907922744751,
      "epoch": 3.57492795389049,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00037633753982694413,
      "loss": 4.6682,
      "mean_token_accuracy": 0.24434349536895753,
      "num_tokens": 85328844.0,
      "step": 37215
    },
    {
      "entropy": 5.044738912582398,
      "epoch": 3.5754082612872238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037630692918165634,
      "loss": 4.5709,
      "mean_token_accuracy": 0.25019311904907227,
      "num_tokens": 85340243.0,
      "step": 37220
    },
    {
      "entropy": 5.018146944046021,
      "epoch": 3.5758885686839577,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003762763161843224,
      "loss": 4.5079,
      "mean_token_accuracy": 0.2559737339615822,
      "num_tokens": 85349874.0,
      "step": 37225
    },
    {
      "entropy": 5.166643762588501,
      "epoch": 3.5763688760806915,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003762457008356531,
      "loss": 4.8189,
      "mean_token_accuracy": 0.2364949107170105,
      "num_tokens": 85362552.0,
      "step": 37230
    },
    {
      "entropy": 5.1092382907867435,
      "epoch": 3.5768491834774254,
      "grad_norm": 1.0,
      "learning_rate": 0.00037621508313635945,
      "loss": 4.6801,
      "mean_token_accuracy": 0.2484647139906883,
      "num_tokens": 85373018.0,
      "step": 37235
    },
    {
      "entropy": 5.04711537361145,
      "epoch": 3.5773294908741593,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003761844630871519,
      "loss": 4.6537,
      "mean_token_accuracy": 0.24833656400442122,
      "num_tokens": 85384386.0,
      "step": 37240
    },
    {
      "entropy": 5.104425668716431,
      "epoch": 3.577809798270893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003761538406887417,
      "loss": 4.595,
      "mean_token_accuracy": 0.2501459434628487,
      "num_tokens": 85394797.0,
      "step": 37245
    },
    {
      "entropy": 5.011469602584839,
      "epoch": 3.5782901056676275,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003761232159418397,
      "loss": 4.5893,
      "mean_token_accuracy": 0.2565186157822609,
      "num_tokens": 85406790.0,
      "step": 37250
    },
    {
      "entropy": 5.108412837982177,
      "epoch": 3.578770413064361,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003760925888471568,
      "loss": 4.5808,
      "mean_token_accuracy": 0.25363024324178696,
      "num_tokens": 85417187.0,
      "step": 37255
    },
    {
      "entropy": 5.114837121963501,
      "epoch": 3.5792507204610953,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003760619594054042,
      "loss": 4.6501,
      "mean_token_accuracy": 0.24643144458532334,
      "num_tokens": 85429562.0,
      "step": 37260
    },
    {
      "entropy": 5.126291751861572,
      "epoch": 3.579731027857829,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037603132761729303,
      "loss": 4.6973,
      "mean_token_accuracy": 0.24665552228689194,
      "num_tokens": 85441151.0,
      "step": 37265
    },
    {
      "entropy": 5.071020412445068,
      "epoch": 3.580211335254563,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003760006934835344,
      "loss": 4.624,
      "mean_token_accuracy": 0.26001393646001814,
      "num_tokens": 85452547.0,
      "step": 37270
    },
    {
      "entropy": 5.194815444946289,
      "epoch": 3.580691642651297,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003759700570048396,
      "loss": 4.8053,
      "mean_token_accuracy": 0.24358469098806382,
      "num_tokens": 85463467.0,
      "step": 37275
    },
    {
      "entropy": 5.017001104354859,
      "epoch": 3.581171950048031,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003759394181819199,
      "loss": 4.5857,
      "mean_token_accuracy": 0.25571144074201585,
      "num_tokens": 85475417.0,
      "step": 37280
    },
    {
      "entropy": 5.080109596252441,
      "epoch": 3.5816522574447647,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037590877701548667,
      "loss": 4.7062,
      "mean_token_accuracy": 0.2471271976828575,
      "num_tokens": 85486644.0,
      "step": 37285
    },
    {
      "entropy": 5.076833057403564,
      "epoch": 3.5821325648414986,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003758781335062512,
      "loss": 4.6188,
      "mean_token_accuracy": 0.2515591368079185,
      "num_tokens": 85497811.0,
      "step": 37290
    },
    {
      "entropy": 5.140811824798584,
      "epoch": 3.5826128722382324,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003758474876549251,
      "loss": 4.6629,
      "mean_token_accuracy": 0.2579734280705452,
      "num_tokens": 85508924.0,
      "step": 37295
    },
    {
      "entropy": 5.003505659103394,
      "epoch": 3.5830931796349663,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037581683946221974,
      "loss": 4.5937,
      "mean_token_accuracy": 0.2578735679388046,
      "num_tokens": 85520541.0,
      "step": 37300
    },
    {
      "entropy": 5.057085990905762,
      "epoch": 3.5835734870317,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003757861889288467,
      "loss": 4.6468,
      "mean_token_accuracy": 0.25001031905412674,
      "num_tokens": 85531717.0,
      "step": 37305
    },
    {
      "entropy": 5.08507022857666,
      "epoch": 3.584053794428434,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037575553605551765,
      "loss": 4.6401,
      "mean_token_accuracy": 0.25046221613883973,
      "num_tokens": 85542145.0,
      "step": 37310
    },
    {
      "entropy": 5.091464900970459,
      "epoch": 3.584534101825168,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037572488084294423,
      "loss": 4.6164,
      "mean_token_accuracy": 0.26090596318244935,
      "num_tokens": 85554584.0,
      "step": 37315
    },
    {
      "entropy": 5.058615112304688,
      "epoch": 3.585014409221902,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037569422329183824,
      "loss": 4.5632,
      "mean_token_accuracy": 0.2559929579496384,
      "num_tokens": 85565218.0,
      "step": 37320
    },
    {
      "entropy": 5.0683098316192625,
      "epoch": 3.585494716618636,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003756635634029112,
      "loss": 4.7572,
      "mean_token_accuracy": 0.24580509662628175,
      "num_tokens": 85577365.0,
      "step": 37325
    },
    {
      "entropy": 5.181476306915283,
      "epoch": 3.5859750240153696,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037563290117687535,
      "loss": 4.826,
      "mean_token_accuracy": 0.23955522030591964,
      "num_tokens": 85589639.0,
      "step": 37330
    },
    {
      "entropy": 5.145947122573853,
      "epoch": 3.586455331412104,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003756022366144423,
      "loss": 4.6315,
      "mean_token_accuracy": 0.24678362309932708,
      "num_tokens": 85601058.0,
      "step": 37335
    },
    {
      "entropy": 5.157214021682739,
      "epoch": 3.586935638808838,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003755715697163239,
      "loss": 4.6205,
      "mean_token_accuracy": 0.24688201397657394,
      "num_tokens": 85611966.0,
      "step": 37340
    },
    {
      "entropy": 5.052561950683594,
      "epoch": 3.5874159462055717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037554090048323247,
      "loss": 4.6382,
      "mean_token_accuracy": 0.25397345423698425,
      "num_tokens": 85623292.0,
      "step": 37345
    },
    {
      "entropy": 5.136478424072266,
      "epoch": 3.5878962536023056,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037551022891587976,
      "loss": 4.7515,
      "mean_token_accuracy": 0.2419552981853485,
      "num_tokens": 85634997.0,
      "step": 37350
    },
    {
      "entropy": 5.073695278167724,
      "epoch": 3.5883765609990395,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037547955501497806,
      "loss": 4.6261,
      "mean_token_accuracy": 0.25159311592578887,
      "num_tokens": 85646540.0,
      "step": 37355
    },
    {
      "entropy": 5.066670274734497,
      "epoch": 3.5888568683957733,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003754488787812394,
      "loss": 4.6737,
      "mean_token_accuracy": 0.24403929710388184,
      "num_tokens": 85658431.0,
      "step": 37360
    },
    {
      "entropy": 5.07287220954895,
      "epoch": 3.589337175792507,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003754182002153761,
      "loss": 4.5302,
      "mean_token_accuracy": 0.26011413484811785,
      "num_tokens": 85668639.0,
      "step": 37365
    },
    {
      "entropy": 5.060079145431518,
      "epoch": 3.589817483189241,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003753875193181003,
      "loss": 4.6538,
      "mean_token_accuracy": 0.25065494775772096,
      "num_tokens": 85680429.0,
      "step": 37370
    },
    {
      "entropy": 5.044514417648315,
      "epoch": 3.590297790585975,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037535683609012446,
      "loss": 4.5973,
      "mean_token_accuracy": 0.2538568675518036,
      "num_tokens": 85691358.0,
      "step": 37375
    },
    {
      "entropy": 5.097279357910156,
      "epoch": 3.590778097982709,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037532615053216084,
      "loss": 4.6492,
      "mean_token_accuracy": 0.2554454177618027,
      "num_tokens": 85702393.0,
      "step": 37380
    },
    {
      "entropy": 5.091383361816407,
      "epoch": 3.5912584053794427,
      "grad_norm": 0.96875,
      "learning_rate": 0.000375295462644922,
      "loss": 4.7048,
      "mean_token_accuracy": 0.24712061733007432,
      "num_tokens": 85715363.0,
      "step": 37385
    },
    {
      "entropy": 5.186736297607422,
      "epoch": 3.5917387127761766,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003752647724291202,
      "loss": 4.716,
      "mean_token_accuracy": 0.24656790345907212,
      "num_tokens": 85726270.0,
      "step": 37390
    },
    {
      "entropy": 5.1081531047821045,
      "epoch": 3.5922190201729105,
      "grad_norm": 1.0,
      "learning_rate": 0.00037523407988546815,
      "loss": 4.6449,
      "mean_token_accuracy": 0.2519206777215004,
      "num_tokens": 85736891.0,
      "step": 37395
    },
    {
      "entropy": 5.087747621536255,
      "epoch": 3.592699327569645,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003752033850146785,
      "loss": 4.6404,
      "mean_token_accuracy": 0.2538626179099083,
      "num_tokens": 85747901.0,
      "step": 37400
    },
    {
      "entropy": 5.030999422073364,
      "epoch": 3.5931796349663783,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003751726878174637,
      "loss": 4.5534,
      "mean_token_accuracy": 0.24899375140666963,
      "num_tokens": 85758320.0,
      "step": 37405
    },
    {
      "entropy": 5.162257099151612,
      "epoch": 3.5936599423631126,
      "grad_norm": 1.0,
      "learning_rate": 0.0003751419882945365,
      "loss": 4.6962,
      "mean_token_accuracy": 0.24782577455043792,
      "num_tokens": 85770004.0,
      "step": 37410
    },
    {
      "entropy": 5.10714201927185,
      "epoch": 3.594140249759846,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003751112864466097,
      "loss": 4.694,
      "mean_token_accuracy": 0.24449634701013565,
      "num_tokens": 85784105.0,
      "step": 37415
    },
    {
      "entropy": 5.061793231964112,
      "epoch": 3.5946205571565804,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003750805822743962,
      "loss": 4.5555,
      "mean_token_accuracy": 0.2593277469277382,
      "num_tokens": 85795166.0,
      "step": 37420
    },
    {
      "entropy": 5.141040325164795,
      "epoch": 3.5951008645533142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003750498757786087,
      "loss": 4.7133,
      "mean_token_accuracy": 0.24777896106243133,
      "num_tokens": 85805334.0,
      "step": 37425
    },
    {
      "entropy": 5.115486192703247,
      "epoch": 3.595581171950048,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003750191669599602,
      "loss": 4.7233,
      "mean_token_accuracy": 0.2427632838487625,
      "num_tokens": 85817548.0,
      "step": 37430
    },
    {
      "entropy": 5.086406278610229,
      "epoch": 3.596061479346782,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003749884558191636,
      "loss": 4.6967,
      "mean_token_accuracy": 0.2428709536790848,
      "num_tokens": 85830770.0,
      "step": 37435
    },
    {
      "entropy": 5.081411218643188,
      "epoch": 3.596541786743516,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037495774235693193,
      "loss": 4.6649,
      "mean_token_accuracy": 0.2557657897472382,
      "num_tokens": 85841682.0,
      "step": 37440
    },
    {
      "entropy": 5.114785385131836,
      "epoch": 3.5970220941402498,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037492702657397837,
      "loss": 4.6389,
      "mean_token_accuracy": 0.25576379895210266,
      "num_tokens": 85853657.0,
      "step": 37445
    },
    {
      "entropy": 5.14649338722229,
      "epoch": 3.5975024015369836,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000374896308471016,
      "loss": 4.7442,
      "mean_token_accuracy": 0.24357293397188187,
      "num_tokens": 85864460.0,
      "step": 37450
    },
    {
      "entropy": 5.092826175689697,
      "epoch": 3.5979827089337175,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037486558804875785,
      "loss": 4.6137,
      "mean_token_accuracy": 0.24906432926654815,
      "num_tokens": 85875865.0,
      "step": 37455
    },
    {
      "entropy": 5.110413551330566,
      "epoch": 3.5984630163304514,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003748348653079174,
      "loss": 4.735,
      "mean_token_accuracy": 0.24760830402374268,
      "num_tokens": 85887101.0,
      "step": 37460
    },
    {
      "entropy": 5.102922105789185,
      "epoch": 3.5989433237271853,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037480414024920777,
      "loss": 4.6884,
      "mean_token_accuracy": 0.24789487272500993,
      "num_tokens": 85897491.0,
      "step": 37465
    },
    {
      "entropy": 5.096307897567749,
      "epoch": 3.599423631123919,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003747734128733424,
      "loss": 4.6881,
      "mean_token_accuracy": 0.24766882956027986,
      "num_tokens": 85908044.0,
      "step": 37470
    },
    {
      "entropy": 5.1168773651123045,
      "epoch": 3.5999039385206535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003747426831810346,
      "loss": 4.6869,
      "mean_token_accuracy": 0.24947619587182998,
      "num_tokens": 85919223.0,
      "step": 37475
    },
    {
      "entropy": 5.09006519317627,
      "epoch": 3.600384245917387,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037471195117299796,
      "loss": 4.6945,
      "mean_token_accuracy": 0.24930300116539,
      "num_tokens": 85930590.0,
      "step": 37480
    },
    {
      "entropy": 5.0655198097229,
      "epoch": 3.6008645533141213,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003746812168499458,
      "loss": 4.6101,
      "mean_token_accuracy": 0.26505458503961565,
      "num_tokens": 85942623.0,
      "step": 37485
    },
    {
      "entropy": 5.113658380508423,
      "epoch": 3.6013448607108547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003746504802125918,
      "loss": 4.6581,
      "mean_token_accuracy": 0.24950801432132722,
      "num_tokens": 85953596.0,
      "step": 37490
    },
    {
      "entropy": 5.107417201995849,
      "epoch": 3.601825168107589,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037461974126164953,
      "loss": 4.6422,
      "mean_token_accuracy": 0.25002007335424425,
      "num_tokens": 85964729.0,
      "step": 37495
    },
    {
      "entropy": 5.064407444000244,
      "epoch": 3.602305475504323,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037458899999783267,
      "loss": 4.6696,
      "mean_token_accuracy": 0.2472187325358391,
      "num_tokens": 85976350.0,
      "step": 37500
    },
    {
      "entropy": 4.994870901107788,
      "epoch": 3.6027857829010568,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037455825642185504,
      "loss": 4.478,
      "mean_token_accuracy": 0.2645552784204483,
      "num_tokens": 85988576.0,
      "step": 37505
    },
    {
      "entropy": 5.05402717590332,
      "epoch": 3.6032660902977907,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003745275105344303,
      "loss": 4.6474,
      "mean_token_accuracy": 0.25074800550937654,
      "num_tokens": 85998312.0,
      "step": 37510
    },
    {
      "entropy": 5.0709826946258545,
      "epoch": 3.6037463976945245,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003744967623362723,
      "loss": 4.7144,
      "mean_token_accuracy": 0.25072051435709,
      "num_tokens": 86009848.0,
      "step": 37515
    },
    {
      "entropy": 5.1340203285217285,
      "epoch": 3.6042267050912584,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003744660118280949,
      "loss": 4.6607,
      "mean_token_accuracy": 0.249435156583786,
      "num_tokens": 86022243.0,
      "step": 37520
    },
    {
      "entropy": 5.0720299243927,
      "epoch": 3.6047070124879923,
      "grad_norm": 1.015625,
      "learning_rate": 0.000374435259010612,
      "loss": 4.6918,
      "mean_token_accuracy": 0.2517037570476532,
      "num_tokens": 86033659.0,
      "step": 37525
    },
    {
      "entropy": 5.070416498184204,
      "epoch": 3.605187319884726,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037440450388453767,
      "loss": 4.5693,
      "mean_token_accuracy": 0.25942796319723127,
      "num_tokens": 86045167.0,
      "step": 37530
    },
    {
      "entropy": 5.0563232421875,
      "epoch": 3.60566762728146,
      "grad_norm": 1.03125,
      "learning_rate": 0.000374373746450586,
      "loss": 4.614,
      "mean_token_accuracy": 0.253194235265255,
      "num_tokens": 86057016.0,
      "step": 37535
    },
    {
      "entropy": 5.060325336456299,
      "epoch": 3.606147934678194,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037434298670947097,
      "loss": 4.6488,
      "mean_token_accuracy": 0.2497337356209755,
      "num_tokens": 86068326.0,
      "step": 37540
    },
    {
      "entropy": 5.058252143859863,
      "epoch": 3.606628242074928,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037431222466190677,
      "loss": 4.5559,
      "mean_token_accuracy": 0.25614576041698456,
      "num_tokens": 86080042.0,
      "step": 37545
    },
    {
      "entropy": 5.088401460647583,
      "epoch": 3.6071085494716617,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037428146030860765,
      "loss": 4.6637,
      "mean_token_accuracy": 0.24101950377225875,
      "num_tokens": 86090641.0,
      "step": 37550
    },
    {
      "entropy": 5.091852521896362,
      "epoch": 3.6075888568683956,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003742506936502878,
      "loss": 4.6509,
      "mean_token_accuracy": 0.2584911659359932,
      "num_tokens": 86101472.0,
      "step": 37555
    },
    {
      "entropy": 5.082197380065918,
      "epoch": 3.60806916426513,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003742199246876616,
      "loss": 4.598,
      "mean_token_accuracy": 0.25803781896829603,
      "num_tokens": 86113761.0,
      "step": 37560
    },
    {
      "entropy": 5.180209159851074,
      "epoch": 3.6085494716618634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037418915342144333,
      "loss": 4.7506,
      "mean_token_accuracy": 0.24518842995166779,
      "num_tokens": 86125406.0,
      "step": 37565
    },
    {
      "entropy": 5.00511417388916,
      "epoch": 3.6090297790585977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003741583798523474,
      "loss": 4.5402,
      "mean_token_accuracy": 0.2563123762607574,
      "num_tokens": 86137235.0,
      "step": 37570
    },
    {
      "entropy": 5.070316314697266,
      "epoch": 3.6095100864553316,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003741276039810884,
      "loss": 4.6439,
      "mean_token_accuracy": 0.25325279533863065,
      "num_tokens": 86148666.0,
      "step": 37575
    },
    {
      "entropy": 5.167660093307495,
      "epoch": 3.6099903938520654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003740968258083808,
      "loss": 4.7029,
      "mean_token_accuracy": 0.24197864532470703,
      "num_tokens": 86159559.0,
      "step": 37580
    },
    {
      "entropy": 5.066628313064575,
      "epoch": 3.6104707012487993,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003740660453349391,
      "loss": 4.6288,
      "mean_token_accuracy": 0.2527072161436081,
      "num_tokens": 86172384.0,
      "step": 37585
    },
    {
      "entropy": 5.1307916164398195,
      "epoch": 3.610951008645533,
      "grad_norm": 1.015625,
      "learning_rate": 0.000374035262561478,
      "loss": 4.6304,
      "mean_token_accuracy": 0.24917416423559188,
      "num_tokens": 86184075.0,
      "step": 37590
    },
    {
      "entropy": 5.060346841812134,
      "epoch": 3.611431316042267,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037400447748871224,
      "loss": 4.6215,
      "mean_token_accuracy": 0.25630906522274016,
      "num_tokens": 86194800.0,
      "step": 37595
    },
    {
      "entropy": 4.998279762268067,
      "epoch": 3.611911623439001,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037397369011735643,
      "loss": 4.5384,
      "mean_token_accuracy": 0.2529678553342819,
      "num_tokens": 86206752.0,
      "step": 37600
    },
    {
      "entropy": 5.076165962219238,
      "epoch": 3.612391930835735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003739429004481254,
      "loss": 4.6058,
      "mean_token_accuracy": 0.2605464980006218,
      "num_tokens": 86218645.0,
      "step": 37605
    },
    {
      "entropy": 5.044993829727173,
      "epoch": 3.6128722382324687,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037391210848173413,
      "loss": 4.6331,
      "mean_token_accuracy": 0.2548112586140633,
      "num_tokens": 86231485.0,
      "step": 37610
    },
    {
      "entropy": 5.088577508926392,
      "epoch": 3.6133525456292026,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037388131421889733,
      "loss": 4.706,
      "mean_token_accuracy": 0.24424276500940323,
      "num_tokens": 86244171.0,
      "step": 37615
    },
    {
      "entropy": 5.181043338775635,
      "epoch": 3.6138328530259365,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003738505176603301,
      "loss": 4.7301,
      "mean_token_accuracy": 0.24273558259010314,
      "num_tokens": 86255014.0,
      "step": 37620
    },
    {
      "entropy": 5.19847526550293,
      "epoch": 3.6143131604226704,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037381971880674726,
      "loss": 4.8065,
      "mean_token_accuracy": 0.23629067689180375,
      "num_tokens": 86265067.0,
      "step": 37625
    },
    {
      "entropy": 5.170110368728638,
      "epoch": 3.6147934678194042,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037378891765886405,
      "loss": 4.7648,
      "mean_token_accuracy": 0.2392004981637001,
      "num_tokens": 86276089.0,
      "step": 37630
    },
    {
      "entropy": 5.068682384490967,
      "epoch": 3.6152737752161386,
      "grad_norm": 1.0,
      "learning_rate": 0.00037375811421739555,
      "loss": 4.597,
      "mean_token_accuracy": 0.25616201013326645,
      "num_tokens": 86287601.0,
      "step": 37635
    },
    {
      "entropy": 5.111126804351807,
      "epoch": 3.615754082612872,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037372730848305683,
      "loss": 4.7578,
      "mean_token_accuracy": 0.2404927134513855,
      "num_tokens": 86299265.0,
      "step": 37640
    },
    {
      "entropy": 5.142370891571045,
      "epoch": 3.6162343900096063,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037369650045656315,
      "loss": 4.6739,
      "mean_token_accuracy": 0.2470065698027611,
      "num_tokens": 86310250.0,
      "step": 37645
    },
    {
      "entropy": 5.110120677947998,
      "epoch": 3.61671469740634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037366569013862986,
      "loss": 4.6413,
      "mean_token_accuracy": 0.25905528515577314,
      "num_tokens": 86322523.0,
      "step": 37650
    },
    {
      "entropy": 5.044010400772095,
      "epoch": 3.617195004803074,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003736348775299722,
      "loss": 4.5561,
      "mean_token_accuracy": 0.2550064459443092,
      "num_tokens": 86334143.0,
      "step": 37655
    },
    {
      "entropy": 5.024511337280273,
      "epoch": 3.617675312199808,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037360406263130556,
      "loss": 4.5527,
      "mean_token_accuracy": 0.25469929426908494,
      "num_tokens": 86345310.0,
      "step": 37660
    },
    {
      "entropy": 5.095580291748047,
      "epoch": 3.618155619596542,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037357324544334537,
      "loss": 4.6653,
      "mean_token_accuracy": 0.24907104074954986,
      "num_tokens": 86356838.0,
      "step": 37665
    },
    {
      "entropy": 5.075345897674561,
      "epoch": 3.6186359269932757,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003735424259668072,
      "loss": 4.6156,
      "mean_token_accuracy": 0.26202905029058454,
      "num_tokens": 86369346.0,
      "step": 37670
    },
    {
      "entropy": 5.086515951156616,
      "epoch": 3.6191162343900096,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003735116042024064,
      "loss": 4.6121,
      "mean_token_accuracy": 0.2564449727535248,
      "num_tokens": 86380441.0,
      "step": 37675
    },
    {
      "entropy": 5.080757188796997,
      "epoch": 3.6195965417867435,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003734807801508587,
      "loss": 4.6771,
      "mean_token_accuracy": 0.2445537492632866,
      "num_tokens": 86391858.0,
      "step": 37680
    },
    {
      "entropy": 5.1455412864685055,
      "epoch": 3.6200768491834774,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037344995381287984,
      "loss": 4.7626,
      "mean_token_accuracy": 0.2384191036224365,
      "num_tokens": 86402763.0,
      "step": 37685
    },
    {
      "entropy": 5.080848503112793,
      "epoch": 3.6205571565802113,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037341912518918524,
      "loss": 4.6269,
      "mean_token_accuracy": 0.25846137702465055,
      "num_tokens": 86413116.0,
      "step": 37690
    },
    {
      "entropy": 5.052998399734497,
      "epoch": 3.621037463976945,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037338829428049087,
      "loss": 4.6357,
      "mean_token_accuracy": 0.2544378384947777,
      "num_tokens": 86424566.0,
      "step": 37695
    },
    {
      "entropy": 5.0788122653961185,
      "epoch": 3.621517771373679,
      "grad_norm": 1.0,
      "learning_rate": 0.00037335746108751246,
      "loss": 4.736,
      "mean_token_accuracy": 0.2458861768245697,
      "num_tokens": 86435049.0,
      "step": 37700
    },
    {
      "entropy": 5.075373935699463,
      "epoch": 3.621998078770413,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003733266256109658,
      "loss": 4.6774,
      "mean_token_accuracy": 0.24338336735963823,
      "num_tokens": 86446186.0,
      "step": 37705
    },
    {
      "entropy": 5.252439355850219,
      "epoch": 3.6224783861671472,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037329578785156686,
      "loss": 4.7205,
      "mean_token_accuracy": 0.2431373968720436,
      "num_tokens": 86456291.0,
      "step": 37710
    },
    {
      "entropy": 5.170227527618408,
      "epoch": 3.6229586935638807,
      "grad_norm": 1.0,
      "learning_rate": 0.0003732649478100317,
      "loss": 4.7218,
      "mean_token_accuracy": 0.2478360116481781,
      "num_tokens": 86467345.0,
      "step": 37715
    },
    {
      "entropy": 5.059037542343139,
      "epoch": 3.623439000960615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037323410548707617,
      "loss": 4.6122,
      "mean_token_accuracy": 0.2590093672275543,
      "num_tokens": 86478254.0,
      "step": 37720
    },
    {
      "entropy": 5.122468376159668,
      "epoch": 3.6239193083573484,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003732032608834164,
      "loss": 4.7721,
      "mean_token_accuracy": 0.23883745670318604,
      "num_tokens": 86491622.0,
      "step": 37725
    },
    {
      "entropy": 5.103543043136597,
      "epoch": 3.6243996157540828,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037317241399976856,
      "loss": 4.5677,
      "mean_token_accuracy": 0.26052537709474566,
      "num_tokens": 86502118.0,
      "step": 37730
    },
    {
      "entropy": 5.059629583358765,
      "epoch": 3.6248799231508166,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037314156483684875,
      "loss": 4.5967,
      "mean_token_accuracy": 0.25427010357379914,
      "num_tokens": 86513277.0,
      "step": 37735
    },
    {
      "entropy": 5.17595739364624,
      "epoch": 3.6253602305475505,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003731107133953732,
      "loss": 4.8435,
      "mean_token_accuracy": 0.24113130420446396,
      "num_tokens": 86524849.0,
      "step": 37740
    },
    {
      "entropy": 5.124017381668091,
      "epoch": 3.6258405379442844,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037307985967605827,
      "loss": 4.6276,
      "mean_token_accuracy": 0.25483871102333067,
      "num_tokens": 86535802.0,
      "step": 37745
    },
    {
      "entropy": 5.048529243469238,
      "epoch": 3.6263208453410183,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037304900367962017,
      "loss": 4.5834,
      "mean_token_accuracy": 0.25676089227199556,
      "num_tokens": 86547268.0,
      "step": 37750
    },
    {
      "entropy": 5.067370843887329,
      "epoch": 3.626801152737752,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003730181454067754,
      "loss": 4.592,
      "mean_token_accuracy": 0.25217443257570266,
      "num_tokens": 86558899.0,
      "step": 37755
    },
    {
      "entropy": 5.15465784072876,
      "epoch": 3.627281460134486,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037298728485824043,
      "loss": 4.7475,
      "mean_token_accuracy": 0.24565772861242294,
      "num_tokens": 86569894.0,
      "step": 37760
    },
    {
      "entropy": 5.128888463973999,
      "epoch": 3.62776176753122,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003729564220347315,
      "loss": 4.7018,
      "mean_token_accuracy": 0.2507124558091164,
      "num_tokens": 86580080.0,
      "step": 37765
    },
    {
      "entropy": 5.0718008518219,
      "epoch": 3.628242074927954,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003729255569369654,
      "loss": 4.634,
      "mean_token_accuracy": 0.2506013810634613,
      "num_tokens": 86591242.0,
      "step": 37770
    },
    {
      "entropy": 5.0222594261169435,
      "epoch": 3.6287223823246877,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003728946895656586,
      "loss": 4.5468,
      "mean_token_accuracy": 0.2568160057067871,
      "num_tokens": 86602239.0,
      "step": 37775
    },
    {
      "entropy": 5.039238786697387,
      "epoch": 3.6292026897214216,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003728638199215279,
      "loss": 4.6949,
      "mean_token_accuracy": 0.24305164366960524,
      "num_tokens": 86614105.0,
      "step": 37780
    },
    {
      "entropy": 5.138247871398926,
      "epoch": 3.629682997118156,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037283294800528985,
      "loss": 4.6811,
      "mean_token_accuracy": 0.2535844728350639,
      "num_tokens": 86624873.0,
      "step": 37785
    },
    {
      "entropy": 5.088062858581543,
      "epoch": 3.6301633045148893,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003728020738176612,
      "loss": 4.6119,
      "mean_token_accuracy": 0.2538898676633835,
      "num_tokens": 86637208.0,
      "step": 37790
    },
    {
      "entropy": 5.172359848022461,
      "epoch": 3.6306436119116237,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003727711973593589,
      "loss": 4.7288,
      "mean_token_accuracy": 0.24510256201028824,
      "num_tokens": 86649349.0,
      "step": 37795
    },
    {
      "entropy": 5.1340892791748045,
      "epoch": 3.631123919308357,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003727403186310996,
      "loss": 4.7015,
      "mean_token_accuracy": 0.24962761402130126,
      "num_tokens": 86661626.0,
      "step": 37800
    },
    {
      "entropy": 5.099214887619018,
      "epoch": 3.6316042267050914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037270943763360046,
      "loss": 4.7154,
      "mean_token_accuracy": 0.2505449503660202,
      "num_tokens": 86671642.0,
      "step": 37805
    },
    {
      "entropy": 5.160044431686401,
      "epoch": 3.6320845341018253,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003726785543675782,
      "loss": 4.6864,
      "mean_token_accuracy": 0.24632345736026764,
      "num_tokens": 86683665.0,
      "step": 37810
    },
    {
      "entropy": 5.086599206924438,
      "epoch": 3.632564841498559,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037264766883375003,
      "loss": 4.5943,
      "mean_token_accuracy": 0.2520718052983284,
      "num_tokens": 86693909.0,
      "step": 37815
    },
    {
      "entropy": 5.141820096969605,
      "epoch": 3.633045148895293,
      "grad_norm": 1.0,
      "learning_rate": 0.0003726167810328329,
      "loss": 4.7116,
      "mean_token_accuracy": 0.24550954401493072,
      "num_tokens": 86706467.0,
      "step": 37820
    },
    {
      "entropy": 5.15879192352295,
      "epoch": 3.633525456292027,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000372585890965544,
      "loss": 4.6831,
      "mean_token_accuracy": 0.2438836306333542,
      "num_tokens": 86718226.0,
      "step": 37825
    },
    {
      "entropy": 5.150217771530151,
      "epoch": 3.634005763688761,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003725549986326004,
      "loss": 4.7415,
      "mean_token_accuracy": 0.24738034158945083,
      "num_tokens": 86730438.0,
      "step": 37830
    },
    {
      "entropy": 5.087861204147339,
      "epoch": 3.6344860710854947,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003725241040347196,
      "loss": 4.6511,
      "mean_token_accuracy": 0.2454354003071785,
      "num_tokens": 86742245.0,
      "step": 37835
    },
    {
      "entropy": 5.055248022079468,
      "epoch": 3.6349663784822286,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003724932071726186,
      "loss": 4.6279,
      "mean_token_accuracy": 0.25477882027626036,
      "num_tokens": 86753626.0,
      "step": 37840
    },
    {
      "entropy": 5.1023846626281735,
      "epoch": 3.6354466858789625,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037246230804701476,
      "loss": 4.6346,
      "mean_token_accuracy": 0.25732103586196897,
      "num_tokens": 86765507.0,
      "step": 37845
    },
    {
      "entropy": 5.034156465530396,
      "epoch": 3.6359269932756964,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037243140665862556,
      "loss": 4.5693,
      "mean_token_accuracy": 0.25694352239370344,
      "num_tokens": 86776339.0,
      "step": 37850
    },
    {
      "entropy": 5.09523115158081,
      "epoch": 3.6364073006724302,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037240050300816843,
      "loss": 4.7131,
      "mean_token_accuracy": 0.24878025949001312,
      "num_tokens": 86787836.0,
      "step": 37855
    },
    {
      "entropy": 5.100965118408203,
      "epoch": 3.636887608069164,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037236959709636085,
      "loss": 4.6225,
      "mean_token_accuracy": 0.2582941517233849,
      "num_tokens": 86799995.0,
      "step": 37860
    },
    {
      "entropy": 5.183699369430542,
      "epoch": 3.637367915465898,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037233868892392035,
      "loss": 4.7536,
      "mean_token_accuracy": 0.2423551693558693,
      "num_tokens": 86812426.0,
      "step": 37865
    },
    {
      "entropy": 5.104881477355957,
      "epoch": 3.6378482228626323,
      "grad_norm": 1.0,
      "learning_rate": 0.0003723077784915645,
      "loss": 4.6877,
      "mean_token_accuracy": 0.24555708169937135,
      "num_tokens": 86822305.0,
      "step": 37870
    },
    {
      "entropy": 5.088838529586792,
      "epoch": 3.6383285302593658,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037227686580001095,
      "loss": 4.6255,
      "mean_token_accuracy": 0.24925509244203567,
      "num_tokens": 86834487.0,
      "step": 37875
    },
    {
      "entropy": 5.1229222297668455,
      "epoch": 3.6388088376561,
      "grad_norm": 1.0,
      "learning_rate": 0.0003722459508499775,
      "loss": 4.6248,
      "mean_token_accuracy": 0.25296334475278853,
      "num_tokens": 86845439.0,
      "step": 37880
    },
    {
      "entropy": 5.040551424026489,
      "epoch": 3.639289145052834,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003722150336421818,
      "loss": 4.6132,
      "mean_token_accuracy": 0.25646093040704726,
      "num_tokens": 86857348.0,
      "step": 37885
    },
    {
      "entropy": 5.093715763092041,
      "epoch": 3.639769452449568,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003721841141773417,
      "loss": 4.6532,
      "mean_token_accuracy": 0.24836108535528184,
      "num_tokens": 86868707.0,
      "step": 37890
    },
    {
      "entropy": 5.045411396026611,
      "epoch": 3.6402497598463017,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037215319245617503,
      "loss": 4.6126,
      "mean_token_accuracy": 0.25628345608711245,
      "num_tokens": 86879865.0,
      "step": 37895
    },
    {
      "entropy": 5.0547863960266115,
      "epoch": 3.6407300672430356,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037212226847939975,
      "loss": 4.6091,
      "mean_token_accuracy": 0.25798009932041166,
      "num_tokens": 86891369.0,
      "step": 37900
    },
    {
      "entropy": 5.154679441452027,
      "epoch": 3.6412103746397695,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003720913422477338,
      "loss": 4.7081,
      "mean_token_accuracy": 0.24967972040176392,
      "num_tokens": 86904702.0,
      "step": 37905
    },
    {
      "entropy": 5.09899263381958,
      "epoch": 3.6416906820365034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037206041376189515,
      "loss": 4.6515,
      "mean_token_accuracy": 0.24440739899873734,
      "num_tokens": 86914546.0,
      "step": 37910
    },
    {
      "entropy": 4.964228010177612,
      "epoch": 3.6421709894332372,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003720294830226018,
      "loss": 4.5197,
      "mean_token_accuracy": 0.26240006238222124,
      "num_tokens": 86925383.0,
      "step": 37915
    },
    {
      "entropy": 5.056173944473267,
      "epoch": 3.642651296829971,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003719985500305721,
      "loss": 4.6932,
      "mean_token_accuracy": 0.24181782454252243,
      "num_tokens": 86937763.0,
      "step": 37920
    },
    {
      "entropy": 5.115506410598755,
      "epoch": 3.643131604226705,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003719676147865241,
      "loss": 4.6623,
      "mean_token_accuracy": 0.2508893057703972,
      "num_tokens": 86948488.0,
      "step": 37925
    },
    {
      "entropy": 5.107485294342041,
      "epoch": 3.643611911623439,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000371936677291176,
      "loss": 4.6085,
      "mean_token_accuracy": 0.25395099371671676,
      "num_tokens": 86960457.0,
      "step": 37930
    },
    {
      "entropy": 5.053769254684449,
      "epoch": 3.6440922190201728,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003719057375452461,
      "loss": 4.6141,
      "mean_token_accuracy": 0.2524847835302353,
      "num_tokens": 86971636.0,
      "step": 37935
    },
    {
      "entropy": 5.135983657836914,
      "epoch": 3.6445725264169067,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003718747955494527,
      "loss": 4.7563,
      "mean_token_accuracy": 0.23872699588537216,
      "num_tokens": 86981717.0,
      "step": 37940
    },
    {
      "entropy": 4.975488185882568,
      "epoch": 3.645052833813641,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037184385130451425,
      "loss": 4.5009,
      "mean_token_accuracy": 0.25380056351423264,
      "num_tokens": 86994461.0,
      "step": 37945
    },
    {
      "entropy": 5.113227987289429,
      "epoch": 3.6455331412103744,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037181290481114903,
      "loss": 4.6681,
      "mean_token_accuracy": 0.24778452068567275,
      "num_tokens": 87005716.0,
      "step": 37950
    },
    {
      "entropy": 5.1196362495422365,
      "epoch": 3.6460134486071087,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003717819560700758,
      "loss": 4.6491,
      "mean_token_accuracy": 0.24705199152231216,
      "num_tokens": 87017809.0,
      "step": 37955
    },
    {
      "entropy": 5.040398025512696,
      "epoch": 3.6464937560038426,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003717510050820128,
      "loss": 4.5905,
      "mean_token_accuracy": 0.25308541506528853,
      "num_tokens": 87028376.0,
      "step": 37960
    },
    {
      "entropy": 5.086601638793946,
      "epoch": 3.6469740634005765,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037172005184767885,
      "loss": 4.6465,
      "mean_token_accuracy": 0.2514526903629303,
      "num_tokens": 87039338.0,
      "step": 37965
    },
    {
      "entropy": 5.078248023986816,
      "epoch": 3.6474543707973104,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003716890963677924,
      "loss": 4.5713,
      "mean_token_accuracy": 0.26180371791124346,
      "num_tokens": 87050404.0,
      "step": 37970
    },
    {
      "entropy": 5.113364839553833,
      "epoch": 3.6479346781940443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003716581386430723,
      "loss": 4.6523,
      "mean_token_accuracy": 0.24585482478141785,
      "num_tokens": 87061020.0,
      "step": 37975
    },
    {
      "entropy": 5.137400960922241,
      "epoch": 3.648414985590778,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003716271786742372,
      "loss": 4.7335,
      "mean_token_accuracy": 0.24480593651533128,
      "num_tokens": 87073579.0,
      "step": 37980
    },
    {
      "entropy": 5.102648401260376,
      "epoch": 3.648895292987512,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037159621646200595,
      "loss": 4.6536,
      "mean_token_accuracy": 0.25213366448879243,
      "num_tokens": 87084782.0,
      "step": 37985
    },
    {
      "entropy": 5.160861539840698,
      "epoch": 3.649375600384246,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003715652520070974,
      "loss": 4.6699,
      "mean_token_accuracy": 0.2411545917391777,
      "num_tokens": 87096398.0,
      "step": 37990
    },
    {
      "entropy": 5.09494194984436,
      "epoch": 3.64985590778098,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003715342853102305,
      "loss": 4.6652,
      "mean_token_accuracy": 0.24746763259172438,
      "num_tokens": 87108685.0,
      "step": 37995
    },
    {
      "entropy": 5.217353248596192,
      "epoch": 3.6503362151777137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003715033163721241,
      "loss": 4.839,
      "mean_token_accuracy": 0.23443522900342942,
      "num_tokens": 87119105.0,
      "step": 38000
    },
    {
      "entropy": 5.111135387420655,
      "epoch": 3.6508165225744476,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003714723451934972,
      "loss": 4.6377,
      "mean_token_accuracy": 0.25259246975183486,
      "num_tokens": 87130382.0,
      "step": 38005
    },
    {
      "entropy": 5.024069547653198,
      "epoch": 3.6512968299711814,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037144137177506887,
      "loss": 4.5615,
      "mean_token_accuracy": 0.25948636829853056,
      "num_tokens": 87141455.0,
      "step": 38010
    },
    {
      "entropy": 5.090369844436646,
      "epoch": 3.6517771373679153,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003714103961175584,
      "loss": 4.6321,
      "mean_token_accuracy": 0.2545865148305893,
      "num_tokens": 87151901.0,
      "step": 38015
    },
    {
      "entropy": 5.116042995452881,
      "epoch": 3.6522574447646496,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003713794182216847,
      "loss": 4.7259,
      "mean_token_accuracy": 0.24248694628477097,
      "num_tokens": 87163190.0,
      "step": 38020
    },
    {
      "entropy": 5.071746397018432,
      "epoch": 3.652737752161383,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037134843808816717,
      "loss": 4.6373,
      "mean_token_accuracy": 0.25728211402893064,
      "num_tokens": 87176118.0,
      "step": 38025
    },
    {
      "entropy": 5.010204696655274,
      "epoch": 3.6532180595581174,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037131745571772497,
      "loss": 4.5205,
      "mean_token_accuracy": 0.2567053154110909,
      "num_tokens": 87187685.0,
      "step": 38030
    },
    {
      "entropy": 5.080197334289551,
      "epoch": 3.653698366954851,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037128647111107736,
      "loss": 4.6189,
      "mean_token_accuracy": 0.25379740744829177,
      "num_tokens": 87199253.0,
      "step": 38035
    },
    {
      "entropy": 5.11206202507019,
      "epoch": 3.654178674351585,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000371255484268944,
      "loss": 4.6548,
      "mean_token_accuracy": 0.2520978763699532,
      "num_tokens": 87210086.0,
      "step": 38040
    },
    {
      "entropy": 5.110010480880737,
      "epoch": 3.654658981748319,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037122449519204397,
      "loss": 4.6849,
      "mean_token_accuracy": 0.24849056601524352,
      "num_tokens": 87220834.0,
      "step": 38045
    },
    {
      "entropy": 5.19793701171875,
      "epoch": 3.655139289145053,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037119350388109697,
      "loss": 4.7471,
      "mean_token_accuracy": 0.23724036812782287,
      "num_tokens": 87231621.0,
      "step": 38050
    },
    {
      "entropy": 5.0546119689941404,
      "epoch": 3.655619596541787,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037116251033682236,
      "loss": 4.6097,
      "mean_token_accuracy": 0.25672447234392165,
      "num_tokens": 87243242.0,
      "step": 38055
    },
    {
      "entropy": 5.131210517883301,
      "epoch": 3.6560999039385207,
      "grad_norm": 1.0,
      "learning_rate": 0.0003711315145599398,
      "loss": 4.7478,
      "mean_token_accuracy": 0.24562825709581376,
      "num_tokens": 87254097.0,
      "step": 38060
    },
    {
      "entropy": 5.176368474960327,
      "epoch": 3.6565802113352546,
      "grad_norm": 1.03125,
      "learning_rate": 0.000371100516551169,
      "loss": 4.7371,
      "mean_token_accuracy": 0.24558158814907075,
      "num_tokens": 87265478.0,
      "step": 38065
    },
    {
      "entropy": 5.145399570465088,
      "epoch": 3.6570605187319885,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003710695163112295,
      "loss": 4.6383,
      "mean_token_accuracy": 0.24609354585409166,
      "num_tokens": 87276740.0,
      "step": 38070
    },
    {
      "entropy": 5.054460859298706,
      "epoch": 3.6575408261287223,
      "grad_norm": 1.0,
      "learning_rate": 0.0003710385138408412,
      "loss": 4.6537,
      "mean_token_accuracy": 0.24853895008563995,
      "num_tokens": 87288387.0,
      "step": 38075
    },
    {
      "entropy": 5.0839540481567385,
      "epoch": 3.658021133525456,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003710075091407237,
      "loss": 4.696,
      "mean_token_accuracy": 0.25374274104833605,
      "num_tokens": 87298859.0,
      "step": 38080
    },
    {
      "entropy": 5.1137611865997314,
      "epoch": 3.65850144092219,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003709765022115969,
      "loss": 4.6005,
      "mean_token_accuracy": 0.2584525913000107,
      "num_tokens": 87311380.0,
      "step": 38085
    },
    {
      "entropy": 5.045075082778931,
      "epoch": 3.658981748318924,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003709454930541808,
      "loss": 4.5495,
      "mean_token_accuracy": 0.2567291930317879,
      "num_tokens": 87321689.0,
      "step": 38090
    },
    {
      "entropy": 5.090453052520752,
      "epoch": 3.659462055715658,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037091448166919516,
      "loss": 4.668,
      "mean_token_accuracy": 0.24770600646734237,
      "num_tokens": 87334659.0,
      "step": 38095
    },
    {
      "entropy": 5.068580770492554,
      "epoch": 3.6599423631123917,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037088346805736014,
      "loss": 4.675,
      "mean_token_accuracy": 0.25145077854394915,
      "num_tokens": 87347452.0,
      "step": 38100
    },
    {
      "entropy": 5.118318939208985,
      "epoch": 3.660422670509126,
      "grad_norm": 1.0,
      "learning_rate": 0.0003708524522193957,
      "loss": 4.6895,
      "mean_token_accuracy": 0.25291628688573836,
      "num_tokens": 87358296.0,
      "step": 38105
    },
    {
      "entropy": 5.102107572555542,
      "epoch": 3.6609029779058595,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037082143415602186,
      "loss": 4.591,
      "mean_token_accuracy": 0.25062034130096433,
      "num_tokens": 87369624.0,
      "step": 38110
    },
    {
      "entropy": 5.0493217468261715,
      "epoch": 3.661383285302594,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003707904138679589,
      "loss": 4.615,
      "mean_token_accuracy": 0.2520409613847733,
      "num_tokens": 87381870.0,
      "step": 38115
    },
    {
      "entropy": 5.091123723983765,
      "epoch": 3.6618635926993277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000370759391355927,
      "loss": 4.6519,
      "mean_token_accuracy": 0.25782447010278703,
      "num_tokens": 87393486.0,
      "step": 38120
    },
    {
      "entropy": 5.0814080238342285,
      "epoch": 3.6623439000960616,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003707283666206464,
      "loss": 4.62,
      "mean_token_accuracy": 0.2555075153708458,
      "num_tokens": 87404081.0,
      "step": 38125
    },
    {
      "entropy": 5.105334520339966,
      "epoch": 3.6628242074927955,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003706973396628373,
      "loss": 4.613,
      "mean_token_accuracy": 0.25475423634052274,
      "num_tokens": 87415086.0,
      "step": 38130
    },
    {
      "entropy": 5.019858169555664,
      "epoch": 3.6633045148895294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003706663104832202,
      "loss": 4.5796,
      "mean_token_accuracy": 0.2581296429038048,
      "num_tokens": 87425601.0,
      "step": 38135
    },
    {
      "entropy": 5.038957118988037,
      "epoch": 3.6637848222862632,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037063527908251536,
      "loss": 4.6126,
      "mean_token_accuracy": 0.2526656612753868,
      "num_tokens": 87437912.0,
      "step": 38140
    },
    {
      "entropy": 5.157754945755005,
      "epoch": 3.664265129682997,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003706042454614434,
      "loss": 4.6394,
      "mean_token_accuracy": 0.2520978018641472,
      "num_tokens": 87449484.0,
      "step": 38145
    },
    {
      "entropy": 5.060836744308472,
      "epoch": 3.664745437079731,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003705732096207247,
      "loss": 4.5814,
      "mean_token_accuracy": 0.26135337054729463,
      "num_tokens": 87461101.0,
      "step": 38150
    },
    {
      "entropy": 5.068535280227661,
      "epoch": 3.665225744476465,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037054217156107987,
      "loss": 4.5808,
      "mean_token_accuracy": 0.25019862949848176,
      "num_tokens": 87472819.0,
      "step": 38155
    },
    {
      "entropy": 5.048040294647217,
      "epoch": 3.6657060518731988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037051113128322953,
      "loss": 4.5885,
      "mean_token_accuracy": 0.25586462318897246,
      "num_tokens": 87484387.0,
      "step": 38160
    },
    {
      "entropy": 5.10573091506958,
      "epoch": 3.6661863592699326,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037048008878789437,
      "loss": 4.6975,
      "mean_token_accuracy": 0.24632806032896043,
      "num_tokens": 87495874.0,
      "step": 38165
    },
    {
      "entropy": 5.1110485076904295,
      "epoch": 3.6666666666666665,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037044904407579506,
      "loss": 4.7649,
      "mean_token_accuracy": 0.24759913831949235,
      "num_tokens": 87506473.0,
      "step": 38170
    },
    {
      "entropy": 5.078562831878662,
      "epoch": 3.6671469740634004,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037041799714765226,
      "loss": 4.6026,
      "mean_token_accuracy": 0.2558997943997383,
      "num_tokens": 87517995.0,
      "step": 38175
    },
    {
      "entropy": 5.132196140289307,
      "epoch": 3.6676272814601347,
      "grad_norm": 0.921875,
      "learning_rate": 0.000370386948004187,
      "loss": 4.624,
      "mean_token_accuracy": 0.2482762187719345,
      "num_tokens": 87529811.0,
      "step": 38180
    },
    {
      "entropy": 5.028821134567261,
      "epoch": 3.668107588856868,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037035589664612004,
      "loss": 4.5909,
      "mean_token_accuracy": 0.24969737827777863,
      "num_tokens": 87541064.0,
      "step": 38185
    },
    {
      "entropy": 5.0892415046691895,
      "epoch": 3.6685878962536025,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003703248430741723,
      "loss": 4.5832,
      "mean_token_accuracy": 0.2495742365717888,
      "num_tokens": 87552998.0,
      "step": 38190
    },
    {
      "entropy": 5.069949388504028,
      "epoch": 3.6690682036503364,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003702937872890647,
      "loss": 4.6344,
      "mean_token_accuracy": 0.2587790325284004,
      "num_tokens": 87564850.0,
      "step": 38195
    },
    {
      "entropy": 5.067332649230957,
      "epoch": 3.6695485110470702,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003702627292915184,
      "loss": 4.662,
      "mean_token_accuracy": 0.25104972571134565,
      "num_tokens": 87576216.0,
      "step": 38200
    },
    {
      "entropy": 5.153991603851319,
      "epoch": 3.670028818443804,
      "grad_norm": 0.84375,
      "learning_rate": 0.0003702316690822544,
      "loss": 4.7258,
      "mean_token_accuracy": 0.2543079242110252,
      "num_tokens": 87588621.0,
      "step": 38205
    },
    {
      "entropy": 5.110312414169312,
      "epoch": 3.670509125840538,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037020060666199386,
      "loss": 4.6455,
      "mean_token_accuracy": 0.24760898649692537,
      "num_tokens": 87600554.0,
      "step": 38210
    },
    {
      "entropy": 5.069766998291016,
      "epoch": 3.670989433237272,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003701695420314578,
      "loss": 4.6538,
      "mean_token_accuracy": 0.25722844451665877,
      "num_tokens": 87611501.0,
      "step": 38215
    },
    {
      "entropy": 5.027830505371094,
      "epoch": 3.6714697406340058,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037013847519136765,
      "loss": 4.5333,
      "mean_token_accuracy": 0.2629034250974655,
      "num_tokens": 87623529.0,
      "step": 38220
    },
    {
      "entropy": 5.066942024230957,
      "epoch": 3.6719500480307397,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003701074061424446,
      "loss": 4.652,
      "mean_token_accuracy": 0.25573812127113343,
      "num_tokens": 87635818.0,
      "step": 38225
    },
    {
      "entropy": 5.065303134918213,
      "epoch": 3.6724303554274735,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037007633488541,
      "loss": 4.6527,
      "mean_token_accuracy": 0.24580302089452744,
      "num_tokens": 87647866.0,
      "step": 38230
    },
    {
      "entropy": 5.092589235305786,
      "epoch": 3.6729106628242074,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037004526142098517,
      "loss": 4.652,
      "mean_token_accuracy": 0.2516085088253021,
      "num_tokens": 87659252.0,
      "step": 38235
    },
    {
      "entropy": 5.093813085556031,
      "epoch": 3.6733909702209413,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037001418574989165,
      "loss": 4.5976,
      "mean_token_accuracy": 0.2567723974585533,
      "num_tokens": 87671279.0,
      "step": 38240
    },
    {
      "entropy": 5.20090880393982,
      "epoch": 3.673871277617675,
      "grad_norm": 1.0,
      "learning_rate": 0.00036998310787285084,
      "loss": 4.7842,
      "mean_token_accuracy": 0.24645301699638367,
      "num_tokens": 87682887.0,
      "step": 38245
    },
    {
      "entropy": 5.115007925033569,
      "epoch": 3.674351585014409,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003699520277905843,
      "loss": 4.6859,
      "mean_token_accuracy": 0.2534624397754669,
      "num_tokens": 87694228.0,
      "step": 38250
    },
    {
      "entropy": 5.055486154556275,
      "epoch": 3.6748318924111434,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003699209455038137,
      "loss": 4.6306,
      "mean_token_accuracy": 0.25532738119363785,
      "num_tokens": 87706020.0,
      "step": 38255
    },
    {
      "entropy": 5.031919240951538,
      "epoch": 3.675312199807877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036988986101326056,
      "loss": 4.587,
      "mean_token_accuracy": 0.25429045408964157,
      "num_tokens": 87716438.0,
      "step": 38260
    },
    {
      "entropy": 5.0591593265533445,
      "epoch": 3.675792507204611,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003698587743196465,
      "loss": 4.585,
      "mean_token_accuracy": 0.26038565784692763,
      "num_tokens": 87727162.0,
      "step": 38265
    },
    {
      "entropy": 5.00170202255249,
      "epoch": 3.6762728146013446,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036982768542369353,
      "loss": 4.5469,
      "mean_token_accuracy": 0.2604043498635292,
      "num_tokens": 87739114.0,
      "step": 38270
    },
    {
      "entropy": 5.074012899398804,
      "epoch": 3.676753121998079,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036979659432612324,
      "loss": 4.681,
      "mean_token_accuracy": 0.2551993265748024,
      "num_tokens": 87751608.0,
      "step": 38275
    },
    {
      "entropy": 5.149392795562744,
      "epoch": 3.677233429394813,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003697655010276575,
      "loss": 4.6885,
      "mean_token_accuracy": 0.24589761942625046,
      "num_tokens": 87763096.0,
      "step": 38280
    },
    {
      "entropy": 5.050564479827881,
      "epoch": 3.6777137367915467,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003697344055290182,
      "loss": 4.5474,
      "mean_token_accuracy": 0.2611980676651001,
      "num_tokens": 87774043.0,
      "step": 38285
    },
    {
      "entropy": 5.127991819381714,
      "epoch": 3.6781940441882806,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003697033078309274,
      "loss": 4.7745,
      "mean_token_accuracy": 0.24007734507322312,
      "num_tokens": 87786645.0,
      "step": 38290
    },
    {
      "entropy": 5.165225601196289,
      "epoch": 3.6786743515850144,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036967220793410687,
      "loss": 4.749,
      "mean_token_accuracy": 0.24531423896551133,
      "num_tokens": 87797294.0,
      "step": 38295
    },
    {
      "entropy": 5.131644535064697,
      "epoch": 3.6791546589817483,
      "grad_norm": 1.0,
      "learning_rate": 0.0003696411058392789,
      "loss": 4.6296,
      "mean_token_accuracy": 0.2493642210960388,
      "num_tokens": 87809174.0,
      "step": 38300
    },
    {
      "entropy": 5.106382417678833,
      "epoch": 3.679634966378482,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003696100015471654,
      "loss": 4.6966,
      "mean_token_accuracy": 0.2448267012834549,
      "num_tokens": 87820289.0,
      "step": 38305
    },
    {
      "entropy": 5.086899518966675,
      "epoch": 3.680115273775216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003695788950584886,
      "loss": 4.6182,
      "mean_token_accuracy": 0.2558057069778442,
      "num_tokens": 87830692.0,
      "step": 38310
    },
    {
      "entropy": 4.9945155620574955,
      "epoch": 3.68059558117195,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036954778637397084,
      "loss": 4.5194,
      "mean_token_accuracy": 0.2634935021400452,
      "num_tokens": 87841588.0,
      "step": 38315
    },
    {
      "entropy": 5.094189596176148,
      "epoch": 3.681075888568684,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036951667549433404,
      "loss": 4.6534,
      "mean_token_accuracy": 0.25254047513008115,
      "num_tokens": 87852129.0,
      "step": 38320
    },
    {
      "entropy": 5.088709783554077,
      "epoch": 3.6815561959654177,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036948556242030083,
      "loss": 4.7148,
      "mean_token_accuracy": 0.250695475935936,
      "num_tokens": 87863669.0,
      "step": 38325
    },
    {
      "entropy": 5.099450397491455,
      "epoch": 3.682036503362152,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036945444715259334,
      "loss": 4.613,
      "mean_token_accuracy": 0.25125362575054166,
      "num_tokens": 87875019.0,
      "step": 38330
    },
    {
      "entropy": 5.155366563796997,
      "epoch": 3.6825168107588855,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003694233296919341,
      "loss": 4.664,
      "mean_token_accuracy": 0.25433524399995805,
      "num_tokens": 87886041.0,
      "step": 38335
    },
    {
      "entropy": 5.077705049514771,
      "epoch": 3.68299711815562,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036939221003904547,
      "loss": 4.635,
      "mean_token_accuracy": 0.24922849088907242,
      "num_tokens": 87897182.0,
      "step": 38340
    },
    {
      "entropy": 5.060061740875244,
      "epoch": 3.6834774255523532,
      "grad_norm": 0.859375,
      "learning_rate": 0.00036936108819465006,
      "loss": 4.661,
      "mean_token_accuracy": 0.2514450952410698,
      "num_tokens": 87909625.0,
      "step": 38345
    },
    {
      "entropy": 5.144206428527832,
      "epoch": 3.6839577329490876,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036932996415947047,
      "loss": 4.7495,
      "mean_token_accuracy": 0.24253323972225188,
      "num_tokens": 87922531.0,
      "step": 38350
    },
    {
      "entropy": 5.093444204330444,
      "epoch": 3.6844380403458215,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036929883793422906,
      "loss": 4.6199,
      "mean_token_accuracy": 0.2518089234828949,
      "num_tokens": 87932926.0,
      "step": 38355
    },
    {
      "entropy": 5.113624715805054,
      "epoch": 3.6849183477425553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003692677095196487,
      "loss": 4.6927,
      "mean_token_accuracy": 0.25047756880521777,
      "num_tokens": 87944149.0,
      "step": 38360
    },
    {
      "entropy": 5.070630884170532,
      "epoch": 3.685398655139289,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000369236578916452,
      "loss": 4.586,
      "mean_token_accuracy": 0.25583580881357193,
      "num_tokens": 87956300.0,
      "step": 38365
    },
    {
      "entropy": 5.030020427703858,
      "epoch": 3.685878962536023,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036920544612536186,
      "loss": 4.572,
      "mean_token_accuracy": 0.25993893295526505,
      "num_tokens": 87968961.0,
      "step": 38370
    },
    {
      "entropy": 5.047727251052857,
      "epoch": 3.686359269932757,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003691743111471009,
      "loss": 4.6178,
      "mean_token_accuracy": 0.25484521836042406,
      "num_tokens": 87981767.0,
      "step": 38375
    },
    {
      "entropy": 5.154641103744507,
      "epoch": 3.686839577329491,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003691431739823922,
      "loss": 4.7572,
      "mean_token_accuracy": 0.24311138987541198,
      "num_tokens": 87993079.0,
      "step": 38380
    },
    {
      "entropy": 5.135982847213745,
      "epoch": 3.6873198847262247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036911203463195843,
      "loss": 4.6482,
      "mean_token_accuracy": 0.2579922527074814,
      "num_tokens": 88003752.0,
      "step": 38385
    },
    {
      "entropy": 5.017614555358887,
      "epoch": 3.6878001921229586,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036908089309652266,
      "loss": 4.5589,
      "mean_token_accuracy": 0.2617955669760704,
      "num_tokens": 88014082.0,
      "step": 38390
    },
    {
      "entropy": 5.122862148284912,
      "epoch": 3.6882804995196925,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000369049749376808,
      "loss": 4.7106,
      "mean_token_accuracy": 0.24453695267438888,
      "num_tokens": 88025577.0,
      "step": 38395
    },
    {
      "entropy": 5.0725806713104244,
      "epoch": 3.6887608069164264,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036901860347353735,
      "loss": 4.6547,
      "mean_token_accuracy": 0.2540478348731995,
      "num_tokens": 88037124.0,
      "step": 38400
    },
    {
      "entropy": 5.201787662506104,
      "epoch": 3.6892411143131603,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000368987455387434,
      "loss": 4.7755,
      "mean_token_accuracy": 0.2373130962252617,
      "num_tokens": 88050008.0,
      "step": 38405
    },
    {
      "entropy": 5.169263076782227,
      "epoch": 3.689721421709894,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003689563051192209,
      "loss": 4.7406,
      "mean_token_accuracy": 0.24134527742862702,
      "num_tokens": 88063592.0,
      "step": 38410
    },
    {
      "entropy": 5.151019430160522,
      "epoch": 3.6902017291066285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036892515266962145,
      "loss": 4.6823,
      "mean_token_accuracy": 0.24287305921316146,
      "num_tokens": 88075658.0,
      "step": 38415
    },
    {
      "entropy": 5.117627716064453,
      "epoch": 3.690682036503362,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003688939980393588,
      "loss": 4.6067,
      "mean_token_accuracy": 0.2616917505860329,
      "num_tokens": 88087149.0,
      "step": 38420
    },
    {
      "entropy": 5.082990741729736,
      "epoch": 3.6911623439000962,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003688628412291564,
      "loss": 4.635,
      "mean_token_accuracy": 0.25454208105802534,
      "num_tokens": 88097870.0,
      "step": 38425
    },
    {
      "entropy": 5.082919454574585,
      "epoch": 3.69164265129683,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036883168223973754,
      "loss": 4.6874,
      "mean_token_accuracy": 0.24908167868852615,
      "num_tokens": 88108315.0,
      "step": 38430
    },
    {
      "entropy": 5.052491569519043,
      "epoch": 3.692122958693564,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003688005210718257,
      "loss": 4.665,
      "mean_token_accuracy": 0.2536837592720985,
      "num_tokens": 88119209.0,
      "step": 38435
    },
    {
      "entropy": 5.115154170989991,
      "epoch": 3.692603266090298,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003687693577261442,
      "loss": 4.6001,
      "mean_token_accuracy": 0.24841394871473313,
      "num_tokens": 88129630.0,
      "step": 38440
    },
    {
      "entropy": 5.0822145462036135,
      "epoch": 3.6930835734870318,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036873819220341667,
      "loss": 4.6614,
      "mean_token_accuracy": 0.2492525726556778,
      "num_tokens": 88142664.0,
      "step": 38445
    },
    {
      "entropy": 5.039520359039306,
      "epoch": 3.6935638808837656,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036870702450436655,
      "loss": 4.6094,
      "mean_token_accuracy": 0.25730950236320493,
      "num_tokens": 88154045.0,
      "step": 38450
    },
    {
      "entropy": 5.086952543258667,
      "epoch": 3.6940441882804995,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036867585462971775,
      "loss": 4.656,
      "mean_token_accuracy": 0.25460516661405563,
      "num_tokens": 88165980.0,
      "step": 38455
    },
    {
      "entropy": 5.153704643249512,
      "epoch": 3.6945244956772334,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003686446825801937,
      "loss": 4.7067,
      "mean_token_accuracy": 0.2454301580786705,
      "num_tokens": 88177797.0,
      "step": 38460
    },
    {
      "entropy": 5.0727245807647705,
      "epoch": 3.6950048030739673,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003686135083565181,
      "loss": 4.6022,
      "mean_token_accuracy": 0.2541072636842728,
      "num_tokens": 88189703.0,
      "step": 38465
    },
    {
      "entropy": 5.1217052936553955,
      "epoch": 3.695485110470701,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036858233195941493,
      "loss": 4.6328,
      "mean_token_accuracy": 0.25338649600744245,
      "num_tokens": 88200117.0,
      "step": 38470
    },
    {
      "entropy": 4.995954990386963,
      "epoch": 3.695965417867435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003685511533896079,
      "loss": 4.5578,
      "mean_token_accuracy": 0.2570286735892296,
      "num_tokens": 88211277.0,
      "step": 38475
    },
    {
      "entropy": 5.081033515930176,
      "epoch": 3.696445725264169,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036851997264782083,
      "loss": 4.67,
      "mean_token_accuracy": 0.2559676319360733,
      "num_tokens": 88222369.0,
      "step": 38480
    },
    {
      "entropy": 5.110042762756348,
      "epoch": 3.696926032660903,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036848878973477767,
      "loss": 4.6777,
      "mean_token_accuracy": 0.2429370880126953,
      "num_tokens": 88233353.0,
      "step": 38485
    },
    {
      "entropy": 5.1450334072113035,
      "epoch": 3.697406340057637,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036845760465120247,
      "loss": 4.6608,
      "mean_token_accuracy": 0.25573973953723905,
      "num_tokens": 88244443.0,
      "step": 38490
    },
    {
      "entropy": 5.169345045089722,
      "epoch": 3.6978866474543706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003684264173978192,
      "loss": 4.7092,
      "mean_token_accuracy": 0.24799252450466155,
      "num_tokens": 88254834.0,
      "step": 38495
    },
    {
      "entropy": 5.162010526657104,
      "epoch": 3.698366954851105,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003683952279753519,
      "loss": 4.6559,
      "mean_token_accuracy": 0.2525052845478058,
      "num_tokens": 88267535.0,
      "step": 38500
    },
    {
      "entropy": 5.109923934936523,
      "epoch": 3.6988472622478388,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036836403638452477,
      "loss": 4.6727,
      "mean_token_accuracy": 0.2479187250137329,
      "num_tokens": 88277945.0,
      "step": 38505
    },
    {
      "entropy": 5.119804334640503,
      "epoch": 3.6993275696445727,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003683328426260619,
      "loss": 4.6938,
      "mean_token_accuracy": 0.2553853169083595,
      "num_tokens": 88289117.0,
      "step": 38510
    },
    {
      "entropy": 5.009228181838989,
      "epoch": 3.6998078770413065,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036830164670068754,
      "loss": 4.6177,
      "mean_token_accuracy": 0.2593109041452408,
      "num_tokens": 88301010.0,
      "step": 38515
    },
    {
      "entropy": 5.101396465301514,
      "epoch": 3.7002881844380404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036827044860912605,
      "loss": 4.6431,
      "mean_token_accuracy": 0.25681435912847517,
      "num_tokens": 88311997.0,
      "step": 38520
    },
    {
      "entropy": 5.180732536315918,
      "epoch": 3.7007684918347743,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003682392483521017,
      "loss": 4.7159,
      "mean_token_accuracy": 0.2544684767723083,
      "num_tokens": 88324212.0,
      "step": 38525
    },
    {
      "entropy": 5.121018362045288,
      "epoch": 3.701248799231508,
      "grad_norm": 1.0,
      "learning_rate": 0.00036820804593033893,
      "loss": 4.633,
      "mean_token_accuracy": 0.2462209314107895,
      "num_tokens": 88336212.0,
      "step": 38530
    },
    {
      "entropy": 5.069417428970337,
      "epoch": 3.701729106628242,
      "grad_norm": 0.96875,
      "learning_rate": 0.000368176841344562,
      "loss": 4.5819,
      "mean_token_accuracy": 0.25885077714920046,
      "num_tokens": 88347651.0,
      "step": 38535
    },
    {
      "entropy": 5.120373964309692,
      "epoch": 3.702209414024976,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036814563459549557,
      "loss": 4.6748,
      "mean_token_accuracy": 0.24834855198860167,
      "num_tokens": 88359386.0,
      "step": 38540
    },
    {
      "entropy": 5.0997549533844,
      "epoch": 3.70268972142171,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036811442568386403,
      "loss": 4.7084,
      "mean_token_accuracy": 0.2453974023461342,
      "num_tokens": 88372527.0,
      "step": 38545
    },
    {
      "entropy": 5.077084589004516,
      "epoch": 3.7031700288184437,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036808321461039207,
      "loss": 4.6386,
      "mean_token_accuracy": 0.24501306265592576,
      "num_tokens": 88384974.0,
      "step": 38550
    },
    {
      "entropy": 5.088533926010132,
      "epoch": 3.7036503362151776,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003680520013758042,
      "loss": 4.6544,
      "mean_token_accuracy": 0.2566433772444725,
      "num_tokens": 88396754.0,
      "step": 38555
    },
    {
      "entropy": 5.102467775344849,
      "epoch": 3.7041306436119115,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036802078598082523,
      "loss": 4.6139,
      "mean_token_accuracy": 0.2524507835507393,
      "num_tokens": 88407706.0,
      "step": 38560
    },
    {
      "entropy": 5.042420434951782,
      "epoch": 3.704610951008646,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003679895684261798,
      "loss": 4.5402,
      "mean_token_accuracy": 0.26206666231155396,
      "num_tokens": 88418735.0,
      "step": 38565
    },
    {
      "entropy": 5.043337059020996,
      "epoch": 3.7050912584053792,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003679583487125927,
      "loss": 4.6286,
      "mean_token_accuracy": 0.25510458946228026,
      "num_tokens": 88432135.0,
      "step": 38570
    },
    {
      "entropy": 5.170778036117554,
      "epoch": 3.7055715658021136,
      "grad_norm": 1.0,
      "learning_rate": 0.0003679271268407888,
      "loss": 4.8193,
      "mean_token_accuracy": 0.2381758451461792,
      "num_tokens": 88444864.0,
      "step": 38575
    },
    {
      "entropy": 5.1153356552124025,
      "epoch": 3.706051873198847,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003678959028114929,
      "loss": 4.6377,
      "mean_token_accuracy": 0.247845122218132,
      "num_tokens": 88456044.0,
      "step": 38580
    },
    {
      "entropy": 5.1868256568908695,
      "epoch": 3.7065321805955813,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036786467662543,
      "loss": 4.7171,
      "mean_token_accuracy": 0.24605813324451448,
      "num_tokens": 88466711.0,
      "step": 38585
    },
    {
      "entropy": 5.112851810455322,
      "epoch": 3.707012487992315,
      "grad_norm": 1.0,
      "learning_rate": 0.0003678334482833251,
      "loss": 4.6824,
      "mean_token_accuracy": 0.2425534188747406,
      "num_tokens": 88477568.0,
      "step": 38590
    },
    {
      "entropy": 5.068500709533692,
      "epoch": 3.707492795389049,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036780221778590316,
      "loss": 4.5763,
      "mean_token_accuracy": 0.2650339603424072,
      "num_tokens": 88488456.0,
      "step": 38595
    },
    {
      "entropy": 5.023856687545776,
      "epoch": 3.707973102785783,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003677709851338893,
      "loss": 4.5925,
      "mean_token_accuracy": 0.25468298196792605,
      "num_tokens": 88498990.0,
      "step": 38600
    },
    {
      "entropy": 5.08435206413269,
      "epoch": 3.708453410182517,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036773975032800867,
      "loss": 4.6031,
      "mean_token_accuracy": 0.25503617376089094,
      "num_tokens": 88509344.0,
      "step": 38605
    },
    {
      "entropy": 5.081196451187134,
      "epoch": 3.7089337175792507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003677085133689864,
      "loss": 4.6372,
      "mean_token_accuracy": 0.24500472694635392,
      "num_tokens": 88521686.0,
      "step": 38610
    },
    {
      "entropy": 5.048555707931518,
      "epoch": 3.7094140249759846,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036767727425754767,
      "loss": 4.6253,
      "mean_token_accuracy": 0.25094345062971113,
      "num_tokens": 88532590.0,
      "step": 38615
    },
    {
      "entropy": 5.115468454360962,
      "epoch": 3.7098943323727185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003676460329944179,
      "loss": 4.6341,
      "mean_token_accuracy": 0.253052793443203,
      "num_tokens": 88543522.0,
      "step": 38620
    },
    {
      "entropy": 5.148989248275757,
      "epoch": 3.7103746397694524,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003676147895803223,
      "loss": 4.7145,
      "mean_token_accuracy": 0.24457293897867202,
      "num_tokens": 88553927.0,
      "step": 38625
    },
    {
      "entropy": 5.107468986511231,
      "epoch": 3.7108549471661862,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003675835440159863,
      "loss": 4.6599,
      "mean_token_accuracy": 0.25474999099969864,
      "num_tokens": 88565417.0,
      "step": 38630
    },
    {
      "entropy": 5.12046856880188,
      "epoch": 3.71133525456292,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003675522963021354,
      "loss": 4.6598,
      "mean_token_accuracy": 0.25391160100698473,
      "num_tokens": 88577535.0,
      "step": 38635
    },
    {
      "entropy": 5.150574636459351,
      "epoch": 3.7118155619596545,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036752104643949493,
      "loss": 4.6452,
      "mean_token_accuracy": 0.25338971614837646,
      "num_tokens": 88589248.0,
      "step": 38640
    },
    {
      "entropy": 5.042962646484375,
      "epoch": 3.712295869356388,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003674897944287904,
      "loss": 4.5919,
      "mean_token_accuracy": 0.25254558622837064,
      "num_tokens": 88600132.0,
      "step": 38645
    },
    {
      "entropy": 5.0118980884552,
      "epoch": 3.712776176753122,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003674585402707477,
      "loss": 4.6047,
      "mean_token_accuracy": 0.2553155556321144,
      "num_tokens": 88612734.0,
      "step": 38650
    },
    {
      "entropy": 5.11864824295044,
      "epoch": 3.7132564841498557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003674272839660921,
      "loss": 4.6038,
      "mean_token_accuracy": 0.2574846923351288,
      "num_tokens": 88623443.0,
      "step": 38655
    },
    {
      "entropy": 5.165136671066284,
      "epoch": 3.71373679154659,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036739602551554945,
      "loss": 4.702,
      "mean_token_accuracy": 0.24244782328605652,
      "num_tokens": 88635258.0,
      "step": 38660
    },
    {
      "entropy": 5.073174047470093,
      "epoch": 3.714217098943324,
      "grad_norm": 1.0,
      "learning_rate": 0.0003673647649198454,
      "loss": 4.6142,
      "mean_token_accuracy": 0.2523558303713799,
      "num_tokens": 88646782.0,
      "step": 38665
    },
    {
      "entropy": 5.062476110458374,
      "epoch": 3.7146974063400577,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036733350217970593,
      "loss": 4.6363,
      "mean_token_accuracy": 0.24883365333080293,
      "num_tokens": 88658683.0,
      "step": 38670
    },
    {
      "entropy": 5.111922407150269,
      "epoch": 3.7151777137367916,
      "grad_norm": 1.0,
      "learning_rate": 0.00036730223729585654,
      "loss": 4.7065,
      "mean_token_accuracy": 0.25253400653600694,
      "num_tokens": 88669672.0,
      "step": 38675
    },
    {
      "entropy": 5.05165662765503,
      "epoch": 3.7156580211335255,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003672709702690234,
      "loss": 4.5858,
      "mean_token_accuracy": 0.2552440583705902,
      "num_tokens": 88681581.0,
      "step": 38680
    },
    {
      "entropy": 5.133931398391724,
      "epoch": 3.7161383285302594,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036723970109993223,
      "loss": 4.6505,
      "mean_token_accuracy": 0.25421790331602095,
      "num_tokens": 88693992.0,
      "step": 38685
    },
    {
      "entropy": 5.101785945892334,
      "epoch": 3.7166186359269933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003672084297893091,
      "loss": 4.6841,
      "mean_token_accuracy": 0.24772959649562837,
      "num_tokens": 88706147.0,
      "step": 38690
    },
    {
      "entropy": 5.149497652053833,
      "epoch": 3.717098943323727,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036717715633788014,
      "loss": 4.6858,
      "mean_token_accuracy": 0.24786393642425536,
      "num_tokens": 88718781.0,
      "step": 38695
    },
    {
      "entropy": 5.216381359100342,
      "epoch": 3.717579250720461,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036714588074637116,
      "loss": 4.759,
      "mean_token_accuracy": 0.25113029032945633,
      "num_tokens": 88730577.0,
      "step": 38700
    },
    {
      "entropy": 5.1039539813995365,
      "epoch": 3.718059558117195,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003671146030155086,
      "loss": 4.5993,
      "mean_token_accuracy": 0.2569827139377594,
      "num_tokens": 88741859.0,
      "step": 38705
    },
    {
      "entropy": 5.029372882843018,
      "epoch": 3.718539865513929,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003670833231460184,
      "loss": 4.5655,
      "mean_token_accuracy": 0.25873389542102815,
      "num_tokens": 88753367.0,
      "step": 38710
    },
    {
      "entropy": 5.198507642745971,
      "epoch": 3.7190201729106627,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036705204113862683,
      "loss": 4.7137,
      "mean_token_accuracy": 0.24958897083997728,
      "num_tokens": 88765277.0,
      "step": 38715
    },
    {
      "entropy": 5.015054082870483,
      "epoch": 3.7195004803073966,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003670207569940603,
      "loss": 4.5472,
      "mean_token_accuracy": 0.2553865358233452,
      "num_tokens": 88776695.0,
      "step": 38720
    },
    {
      "entropy": 5.1216127395629885,
      "epoch": 3.719980787704131,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036698947071304495,
      "loss": 4.75,
      "mean_token_accuracy": 0.2421002835035324,
      "num_tokens": 88788766.0,
      "step": 38725
    },
    {
      "entropy": 5.11321702003479,
      "epoch": 3.7204610951008643,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036695818229630735,
      "loss": 4.698,
      "mean_token_accuracy": 0.24866437017917634,
      "num_tokens": 88800812.0,
      "step": 38730
    },
    {
      "entropy": 5.08926420211792,
      "epoch": 3.7209414024975986,
      "grad_norm": 1.0,
      "learning_rate": 0.0003669268917445737,
      "loss": 4.5968,
      "mean_token_accuracy": 0.2588414415717125,
      "num_tokens": 88812908.0,
      "step": 38735
    },
    {
      "entropy": 4.998296213150025,
      "epoch": 3.7214217098943325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003668955990585706,
      "loss": 4.4575,
      "mean_token_accuracy": 0.26801868677139284,
      "num_tokens": 88823248.0,
      "step": 38740
    },
    {
      "entropy": 5.051337671279907,
      "epoch": 3.7219020172910664,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036686430423902466,
      "loss": 4.6396,
      "mean_token_accuracy": 0.25335901528596877,
      "num_tokens": 88834423.0,
      "step": 38745
    },
    {
      "entropy": 5.145777797698974,
      "epoch": 3.7223823246878003,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003668330072866623,
      "loss": 4.6913,
      "mean_token_accuracy": 0.2466604456305504,
      "num_tokens": 88845802.0,
      "step": 38750
    },
    {
      "entropy": 5.073604536056519,
      "epoch": 3.722862632084534,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036680170820221013,
      "loss": 4.6174,
      "mean_token_accuracy": 0.2498373955488205,
      "num_tokens": 88857023.0,
      "step": 38755
    },
    {
      "entropy": 5.130909156799317,
      "epoch": 3.723342939481268,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000366770406986395,
      "loss": 4.7123,
      "mean_token_accuracy": 0.24981154799461364,
      "num_tokens": 88869233.0,
      "step": 38760
    },
    {
      "entropy": 5.09631314277649,
      "epoch": 3.723823246878002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003667391036399435,
      "loss": 4.6226,
      "mean_token_accuracy": 0.2500057712197304,
      "num_tokens": 88880907.0,
      "step": 38765
    },
    {
      "entropy": 5.0434082508087155,
      "epoch": 3.724303554274736,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003667077981635824,
      "loss": 4.5282,
      "mean_token_accuracy": 0.26535087674856184,
      "num_tokens": 88891431.0,
      "step": 38770
    },
    {
      "entropy": 5.0661791324615475,
      "epoch": 3.7247838616714697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036667649055803855,
      "loss": 4.5888,
      "mean_token_accuracy": 0.2573632672429085,
      "num_tokens": 88902206.0,
      "step": 38775
    },
    {
      "entropy": 5.0722630500793455,
      "epoch": 3.7252641690682036,
      "grad_norm": 1.0,
      "learning_rate": 0.00036664518082403876,
      "loss": 4.6669,
      "mean_token_accuracy": 0.25162563025951384,
      "num_tokens": 88912969.0,
      "step": 38780
    },
    {
      "entropy": 5.1457624435424805,
      "epoch": 3.7257444764649374,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003666138689623101,
      "loss": 4.6871,
      "mean_token_accuracy": 0.25196520537137984,
      "num_tokens": 88924699.0,
      "step": 38785
    },
    {
      "entropy": 5.10528826713562,
      "epoch": 3.7262247838616713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036658255497357936,
      "loss": 4.561,
      "mean_token_accuracy": 0.257609498500824,
      "num_tokens": 88936540.0,
      "step": 38790
    },
    {
      "entropy": 4.992747783660889,
      "epoch": 3.726705091258405,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00036655123885857366,
      "loss": 4.5158,
      "mean_token_accuracy": 0.26353667080402376,
      "num_tokens": 88948056.0,
      "step": 38795
    },
    {
      "entropy": 5.064490032196045,
      "epoch": 3.7271853986551395,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036651992061802007,
      "loss": 4.6814,
      "mean_token_accuracy": 0.2458980441093445,
      "num_tokens": 88959225.0,
      "step": 38800
    },
    {
      "entropy": 5.038723373413086,
      "epoch": 3.727665706051873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003664886002526457,
      "loss": 4.5807,
      "mean_token_accuracy": 0.2544922441244125,
      "num_tokens": 88970749.0,
      "step": 38805
    },
    {
      "entropy": 5.144995832443238,
      "epoch": 3.7281460134486073,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036645727776317763,
      "loss": 4.6989,
      "mean_token_accuracy": 0.24638806730508805,
      "num_tokens": 88983149.0,
      "step": 38810
    },
    {
      "entropy": 5.015345335006714,
      "epoch": 3.728626320845341,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003664259531503432,
      "loss": 4.5559,
      "mean_token_accuracy": 0.2532812938094139,
      "num_tokens": 88994650.0,
      "step": 38815
    },
    {
      "entropy": 5.072617769241333,
      "epoch": 3.729106628242075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003663946264148696,
      "loss": 4.6467,
      "mean_token_accuracy": 0.2514279708266258,
      "num_tokens": 89005932.0,
      "step": 38820
    },
    {
      "entropy": 5.040671968460083,
      "epoch": 3.729586935638809,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003663632975574841,
      "loss": 4.5734,
      "mean_token_accuracy": 0.25849454700946806,
      "num_tokens": 89018758.0,
      "step": 38825
    },
    {
      "entropy": 5.100713682174683,
      "epoch": 3.730067243035543,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003663319665789142,
      "loss": 4.6639,
      "mean_token_accuracy": 0.25500238835811617,
      "num_tokens": 89029194.0,
      "step": 38830
    },
    {
      "entropy": 5.135214233398438,
      "epoch": 3.7305475504322767,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003663006334798872,
      "loss": 4.68,
      "mean_token_accuracy": 0.24878722429275513,
      "num_tokens": 89040998.0,
      "step": 38835
    },
    {
      "entropy": 5.131090545654297,
      "epoch": 3.7310278578290106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003662692982611306,
      "loss": 4.5937,
      "mean_token_accuracy": 0.2539515569806099,
      "num_tokens": 89051790.0,
      "step": 38840
    },
    {
      "entropy": 5.043944787979126,
      "epoch": 3.7315081652257445,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036623796092337196,
      "loss": 4.6424,
      "mean_token_accuracy": 0.25520167499780655,
      "num_tokens": 89062695.0,
      "step": 38845
    },
    {
      "entropy": 5.083025884628296,
      "epoch": 3.7319884726224783,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003662066214673388,
      "loss": 4.6394,
      "mean_token_accuracy": 0.25567148476839063,
      "num_tokens": 89074208.0,
      "step": 38850
    },
    {
      "entropy": 5.227126693725586,
      "epoch": 3.7324687800192122,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003661752798937586,
      "loss": 4.7399,
      "mean_token_accuracy": 0.23968783020973206,
      "num_tokens": 89085127.0,
      "step": 38855
    },
    {
      "entropy": 5.155915021896362,
      "epoch": 3.732949087415946,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036614393620335923,
      "loss": 4.6601,
      "mean_token_accuracy": 0.245440474152565,
      "num_tokens": 89096511.0,
      "step": 38860
    },
    {
      "entropy": 5.054928207397461,
      "epoch": 3.73342939481268,
      "grad_norm": 1.0,
      "learning_rate": 0.0003661125903968683,
      "loss": 4.5561,
      "mean_token_accuracy": 0.2537227481603622,
      "num_tokens": 89107603.0,
      "step": 38865
    },
    {
      "entropy": 5.0408666133880615,
      "epoch": 3.733909702209414,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003660812424750135,
      "loss": 4.598,
      "mean_token_accuracy": 0.2565795391798019,
      "num_tokens": 89118004.0,
      "step": 38870
    },
    {
      "entropy": 5.065018033981323,
      "epoch": 3.734390009606148,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003660498924385228,
      "loss": 4.6355,
      "mean_token_accuracy": 0.25155327171087266,
      "num_tokens": 89129893.0,
      "step": 38875
    },
    {
      "entropy": 5.061264181137085,
      "epoch": 3.7348703170028816,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036601854028812394,
      "loss": 4.571,
      "mean_token_accuracy": 0.2593329295516014,
      "num_tokens": 89140500.0,
      "step": 38880
    },
    {
      "entropy": 5.070968914031982,
      "epoch": 3.735350624399616,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036598718602454474,
      "loss": 4.5843,
      "mean_token_accuracy": 0.2591773644089699,
      "num_tokens": 89150567.0,
      "step": 38885
    },
    {
      "entropy": 5.089797067642212,
      "epoch": 3.7358309317963494,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003659558296485133,
      "loss": 4.6354,
      "mean_token_accuracy": 0.2582481399178505,
      "num_tokens": 89162400.0,
      "step": 38890
    },
    {
      "entropy": 5.116664552688599,
      "epoch": 3.7363112391930837,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036592447116075765,
      "loss": 4.6971,
      "mean_token_accuracy": 0.24563159942626953,
      "num_tokens": 89173889.0,
      "step": 38895
    },
    {
      "entropy": 5.2170634269714355,
      "epoch": 3.7367915465898176,
      "grad_norm": 1.0,
      "learning_rate": 0.00036589311056200573,
      "loss": 4.809,
      "mean_token_accuracy": 0.24192871451377868,
      "num_tokens": 89185287.0,
      "step": 38900
    },
    {
      "entropy": 5.146986675262451,
      "epoch": 3.7372718539865515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003658617478529857,
      "loss": 4.7621,
      "mean_token_accuracy": 0.24672640413045882,
      "num_tokens": 89197976.0,
      "step": 38905
    },
    {
      "entropy": 5.092033004760742,
      "epoch": 3.7377521613832854,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036583038303442567,
      "loss": 4.6682,
      "mean_token_accuracy": 0.24752541184425353,
      "num_tokens": 89209078.0,
      "step": 38910
    },
    {
      "entropy": 5.210891437530518,
      "epoch": 3.7382324687800192,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036579901610705377,
      "loss": 4.7172,
      "mean_token_accuracy": 0.24299340546131135,
      "num_tokens": 89221838.0,
      "step": 38915
    },
    {
      "entropy": 5.084416723251342,
      "epoch": 3.738712776176753,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036576764707159853,
      "loss": 4.6479,
      "mean_token_accuracy": 0.2530025288462639,
      "num_tokens": 89232272.0,
      "step": 38920
    },
    {
      "entropy": 5.110701942443848,
      "epoch": 3.739193083573487,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003657362759287879,
      "loss": 4.6449,
      "mean_token_accuracy": 0.2488437920808792,
      "num_tokens": 89244601.0,
      "step": 38925
    },
    {
      "entropy": 5.100154876708984,
      "epoch": 3.739673390970221,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036570490267935044,
      "loss": 4.6455,
      "mean_token_accuracy": 0.25198967605829237,
      "num_tokens": 89255962.0,
      "step": 38930
    },
    {
      "entropy": 5.0598423957824705,
      "epoch": 3.7401536983669548,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036567352732401443,
      "loss": 4.6213,
      "mean_token_accuracy": 0.2509034886956215,
      "num_tokens": 89267786.0,
      "step": 38935
    },
    {
      "entropy": 5.082919645309448,
      "epoch": 3.7406340057636887,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036564214986350844,
      "loss": 4.66,
      "mean_token_accuracy": 0.24941446632146835,
      "num_tokens": 89280220.0,
      "step": 38940
    },
    {
      "entropy": 5.137206077575684,
      "epoch": 3.7411143131604225,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036561077029856077,
      "loss": 4.6293,
      "mean_token_accuracy": 0.24889074712991716,
      "num_tokens": 89291183.0,
      "step": 38945
    },
    {
      "entropy": 5.070282697677612,
      "epoch": 3.7415946205571564,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036557938862990017,
      "loss": 4.6821,
      "mean_token_accuracy": 0.24759533256292343,
      "num_tokens": 89302565.0,
      "step": 38950
    },
    {
      "entropy": 5.058570718765258,
      "epoch": 3.7420749279538903,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00036554800485825506,
      "loss": 4.6021,
      "mean_token_accuracy": 0.25205858051776886,
      "num_tokens": 89314295.0,
      "step": 38955
    },
    {
      "entropy": 5.039202451705933,
      "epoch": 3.7425552353506246,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036551661898435414,
      "loss": 4.5378,
      "mean_token_accuracy": 0.2644787490367889,
      "num_tokens": 89325047.0,
      "step": 38960
    },
    {
      "entropy": 5.104923963546753,
      "epoch": 3.743035542747358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003654852310089262,
      "loss": 4.6801,
      "mean_token_accuracy": 0.2549544423818588,
      "num_tokens": 89336309.0,
      "step": 38965
    },
    {
      "entropy": 5.10462646484375,
      "epoch": 3.7435158501440924,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036545384093269984,
      "loss": 4.549,
      "mean_token_accuracy": 0.25772915333509444,
      "num_tokens": 89348422.0,
      "step": 38970
    },
    {
      "entropy": 5.072885370254516,
      "epoch": 3.7439961575408263,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036542244875640385,
      "loss": 4.615,
      "mean_token_accuracy": 0.2565992161631584,
      "num_tokens": 89360297.0,
      "step": 38975
    },
    {
      "entropy": 5.069131135940552,
      "epoch": 3.74447646493756,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036539105448076713,
      "loss": 4.601,
      "mean_token_accuracy": 0.26005024611949923,
      "num_tokens": 89372498.0,
      "step": 38980
    },
    {
      "entropy": 5.041598463058472,
      "epoch": 3.744956772334294,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003653596581065185,
      "loss": 4.6037,
      "mean_token_accuracy": 0.2564409777522087,
      "num_tokens": 89386176.0,
      "step": 38985
    },
    {
      "entropy": 5.027814102172852,
      "epoch": 3.745437079731028,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003653282596343869,
      "loss": 4.5805,
      "mean_token_accuracy": 0.2566273152828217,
      "num_tokens": 89397779.0,
      "step": 38990
    },
    {
      "entropy": 5.111563301086425,
      "epoch": 3.745917387127762,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003652968590651014,
      "loss": 4.6141,
      "mean_token_accuracy": 0.25729541629552843,
      "num_tokens": 89409313.0,
      "step": 38995
    },
    {
      "entropy": 5.138322734832764,
      "epoch": 3.7463976945244957,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036526545639939097,
      "loss": 4.6527,
      "mean_token_accuracy": 0.2581737250089645,
      "num_tokens": 89421715.0,
      "step": 39000
    },
    {
      "epoch": 3.7463976945244957,
      "eval_entropy": 4.879512735426804,
      "eval_loss": 4.754510879516602,
      "eval_mean_token_accuracy": 0.25433530046199077,
      "eval_num_tokens": 89421715.0,
      "eval_runtime": 26.6864,
      "eval_samples_per_second": 1229.653,
      "eval_steps_per_second": 153.711,
      "step": 39000
    },
    {
      "entropy": 5.0712611198425295,
      "epoch": 3.7468780019212296,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003652340516379846,
      "loss": 4.5986,
      "mean_token_accuracy": 0.25506399720907214,
      "num_tokens": 89432223.0,
      "step": 39005
    },
    {
      "entropy": 5.015200853347778,
      "epoch": 3.7473583093179634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036520264478161164,
      "loss": 4.5838,
      "mean_token_accuracy": 0.2656658411026001,
      "num_tokens": 89442960.0,
      "step": 39010
    },
    {
      "entropy": 5.111458730697632,
      "epoch": 3.7478386167146973,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003651712358310009,
      "loss": 4.6558,
      "mean_token_accuracy": 0.24895536601543428,
      "num_tokens": 89454038.0,
      "step": 39015
    },
    {
      "entropy": 5.263994646072388,
      "epoch": 3.748318924111431,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036513982478688197,
      "loss": 4.8029,
      "mean_token_accuracy": 0.23638329952955245,
      "num_tokens": 89466716.0,
      "step": 39020
    },
    {
      "entropy": 5.129837560653686,
      "epoch": 3.748799231508165,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036510841164998385,
      "loss": 4.6438,
      "mean_token_accuracy": 0.24912787824869156,
      "num_tokens": 89478444.0,
      "step": 39025
    },
    {
      "entropy": 5.051269102096557,
      "epoch": 3.749279538904899,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003650769964210361,
      "loss": 4.6192,
      "mean_token_accuracy": 0.25106038749217985,
      "num_tokens": 89489515.0,
      "step": 39030
    },
    {
      "entropy": 5.083862924575806,
      "epoch": 3.7497598463016333,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003650455791007679,
      "loss": 4.6668,
      "mean_token_accuracy": 0.2524017602205276,
      "num_tokens": 89500220.0,
      "step": 39035
    },
    {
      "entropy": 5.111151885986328,
      "epoch": 3.7502401536983667,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036501415968990874,
      "loss": 4.5898,
      "mean_token_accuracy": 0.2522943913936615,
      "num_tokens": 89511141.0,
      "step": 39040
    },
    {
      "entropy": 5.033183765411377,
      "epoch": 3.750720461095101,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003649827381891881,
      "loss": 4.5724,
      "mean_token_accuracy": 0.25301480293273926,
      "num_tokens": 89523085.0,
      "step": 39045
    },
    {
      "entropy": 5.017771196365357,
      "epoch": 3.751200768491835,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003649513145993355,
      "loss": 4.6072,
      "mean_token_accuracy": 0.2563282251358032,
      "num_tokens": 89534048.0,
      "step": 39050
    },
    {
      "entropy": 5.126554489135742,
      "epoch": 3.751681075888569,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003649198889210804,
      "loss": 4.6726,
      "mean_token_accuracy": 0.24880437105894088,
      "num_tokens": 89545149.0,
      "step": 39055
    },
    {
      "entropy": 5.070091485977173,
      "epoch": 3.7521613832853027,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003648884611551525,
      "loss": 4.6247,
      "mean_token_accuracy": 0.2528557628393173,
      "num_tokens": 89558233.0,
      "step": 39060
    },
    {
      "entropy": 5.122643899917603,
      "epoch": 3.7526416906820366,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036485703130228147,
      "loss": 4.6509,
      "mean_token_accuracy": 0.24945441782474517,
      "num_tokens": 89569775.0,
      "step": 39065
    },
    {
      "entropy": 5.0733390808105465,
      "epoch": 3.7531219980787704,
      "grad_norm": 1.0,
      "learning_rate": 0.00036482559936319703,
      "loss": 4.6608,
      "mean_token_accuracy": 0.2534037515521049,
      "num_tokens": 89581572.0,
      "step": 39070
    },
    {
      "entropy": 5.03880615234375,
      "epoch": 3.7536023054755043,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036479416533862885,
      "loss": 4.6147,
      "mean_token_accuracy": 0.253967148065567,
      "num_tokens": 89592856.0,
      "step": 39075
    },
    {
      "entropy": 5.066880750656128,
      "epoch": 3.754082612872238,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036476272922930674,
      "loss": 4.608,
      "mean_token_accuracy": 0.25648281574249265,
      "num_tokens": 89604245.0,
      "step": 39080
    },
    {
      "entropy": 5.109600353240967,
      "epoch": 3.754562920268972,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036473129103596077,
      "loss": 4.6704,
      "mean_token_accuracy": 0.24748291671276093,
      "num_tokens": 89615741.0,
      "step": 39085
    },
    {
      "entropy": 5.0357239723205565,
      "epoch": 3.755043227665706,
      "grad_norm": 1.0,
      "learning_rate": 0.0003646998507593205,
      "loss": 4.5889,
      "mean_token_accuracy": 0.2574347421526909,
      "num_tokens": 89625969.0,
      "step": 39090
    },
    {
      "entropy": 5.026542234420776,
      "epoch": 3.75552353506244,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003646684084001161,
      "loss": 4.5638,
      "mean_token_accuracy": 0.25495549738407136,
      "num_tokens": 89637933.0,
      "step": 39095
    },
    {
      "entropy": 5.079291534423828,
      "epoch": 3.7560038424591737,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003646369639590776,
      "loss": 4.5898,
      "mean_token_accuracy": 0.2505185127258301,
      "num_tokens": 89649111.0,
      "step": 39100
    },
    {
      "entropy": 5.087254381179809,
      "epoch": 3.7564841498559076,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003646055174369348,
      "loss": 4.6291,
      "mean_token_accuracy": 0.24979290515184402,
      "num_tokens": 89660445.0,
      "step": 39105
    },
    {
      "entropy": 5.043628311157226,
      "epoch": 3.756964457252642,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003645740688344181,
      "loss": 4.6076,
      "mean_token_accuracy": 0.2538358584046364,
      "num_tokens": 89672585.0,
      "step": 39110
    },
    {
      "entropy": 5.065124034881592,
      "epoch": 3.7574447646493754,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003645426181522575,
      "loss": 4.6773,
      "mean_token_accuracy": 0.252958832681179,
      "num_tokens": 89684117.0,
      "step": 39115
    },
    {
      "entropy": 5.035462999343872,
      "epoch": 3.7579250720461097,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036451116539118304,
      "loss": 4.5948,
      "mean_token_accuracy": 0.2578281283378601,
      "num_tokens": 89696679.0,
      "step": 39120
    },
    {
      "entropy": 5.089641141891479,
      "epoch": 3.7584053794428436,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036447971055192526,
      "loss": 4.6334,
      "mean_token_accuracy": 0.24802689701318742,
      "num_tokens": 89707896.0,
      "step": 39125
    },
    {
      "entropy": 5.107311058044433,
      "epoch": 3.7588856868395775,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003644482536352142,
      "loss": 4.6217,
      "mean_token_accuracy": 0.24821676164865494,
      "num_tokens": 89719796.0,
      "step": 39130
    },
    {
      "entropy": 5.091448879241943,
      "epoch": 3.7593659942363113,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036441679464178044,
      "loss": 4.6183,
      "mean_token_accuracy": 0.25557070821523664,
      "num_tokens": 89730986.0,
      "step": 39135
    },
    {
      "entropy": 5.102372550964356,
      "epoch": 3.7598463016330452,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003643853335723541,
      "loss": 4.6008,
      "mean_token_accuracy": 0.25904804170131684,
      "num_tokens": 89742923.0,
      "step": 39140
    },
    {
      "entropy": 5.072693300247193,
      "epoch": 3.760326609029779,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036435387042766575,
      "loss": 4.624,
      "mean_token_accuracy": 0.25536916106939317,
      "num_tokens": 89754883.0,
      "step": 39145
    },
    {
      "entropy": 5.054178714752197,
      "epoch": 3.760806916426513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003643224052084459,
      "loss": 4.6226,
      "mean_token_accuracy": 0.2511888727545738,
      "num_tokens": 89766872.0,
      "step": 39150
    },
    {
      "entropy": 5.164642477035523,
      "epoch": 3.761287223823247,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000364290937915425,
      "loss": 4.7602,
      "mean_token_accuracy": 0.24678073972463607,
      "num_tokens": 89777600.0,
      "step": 39155
    },
    {
      "entropy": 5.1246692657470705,
      "epoch": 3.7617675312199808,
      "grad_norm": 1.0,
      "learning_rate": 0.00036425946854933374,
      "loss": 4.6781,
      "mean_token_accuracy": 0.24802444726228715,
      "num_tokens": 89789760.0,
      "step": 39160
    },
    {
      "entropy": 5.004641723632813,
      "epoch": 3.7622478386167146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036422799711090253,
      "loss": 4.4796,
      "mean_token_accuracy": 0.2600270792841911,
      "num_tokens": 89801072.0,
      "step": 39165
    },
    {
      "entropy": 5.051788949966431,
      "epoch": 3.7627281460134485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003641965236008623,
      "loss": 4.6026,
      "mean_token_accuracy": 0.25869910418987274,
      "num_tokens": 89813503.0,
      "step": 39170
    },
    {
      "entropy": 5.02124547958374,
      "epoch": 3.7632084534101824,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003641650480199436,
      "loss": 4.4691,
      "mean_token_accuracy": 0.26380808651447296,
      "num_tokens": 89824741.0,
      "step": 39175
    },
    {
      "entropy": 5.172266721725464,
      "epoch": 3.7636887608069163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036413357036887723,
      "loss": 4.7444,
      "mean_token_accuracy": 0.24037015438079834,
      "num_tokens": 89835463.0,
      "step": 39180
    },
    {
      "entropy": 5.079380941390991,
      "epoch": 3.7641690682036506,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000364102090648394,
      "loss": 4.7005,
      "mean_token_accuracy": 0.24951838999986647,
      "num_tokens": 89848036.0,
      "step": 39185
    },
    {
      "entropy": 5.039254379272461,
      "epoch": 3.764649375600384,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003640706088592249,
      "loss": 4.5992,
      "mean_token_accuracy": 0.2562743604183197,
      "num_tokens": 89860373.0,
      "step": 39190
    },
    {
      "entropy": 5.109010124206543,
      "epoch": 3.7651296829971184,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036403912500210066,
      "loss": 4.7072,
      "mean_token_accuracy": 0.24658008962869643,
      "num_tokens": 89870767.0,
      "step": 39195
    },
    {
      "entropy": 5.143538141250611,
      "epoch": 3.765609990393852,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003640076390777524,
      "loss": 4.7077,
      "mean_token_accuracy": 0.24711437076330184,
      "num_tokens": 89882436.0,
      "step": 39200
    },
    {
      "entropy": 5.048177862167359,
      "epoch": 3.766090297790586,
      "grad_norm": 0.953125,
      "learning_rate": 0.000363976151086911,
      "loss": 4.5698,
      "mean_token_accuracy": 0.25365736335515976,
      "num_tokens": 89894223.0,
      "step": 39205
    },
    {
      "entropy": 5.119831705093384,
      "epoch": 3.76657060518732,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003639446610303076,
      "loss": 4.7123,
      "mean_token_accuracy": 0.24932771474123,
      "num_tokens": 89905183.0,
      "step": 39210
    },
    {
      "entropy": 5.144737911224365,
      "epoch": 3.767050912584054,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036391316890867334,
      "loss": 4.6776,
      "mean_token_accuracy": 0.2548335313796997,
      "num_tokens": 89916703.0,
      "step": 39215
    },
    {
      "entropy": 5.1132025718688965,
      "epoch": 3.7675312199807878,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036388167472273925,
      "loss": 4.7024,
      "mean_token_accuracy": 0.25457200706005095,
      "num_tokens": 89929015.0,
      "step": 39220
    },
    {
      "entropy": 5.036671781539917,
      "epoch": 3.7680115273775217,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003638501784732366,
      "loss": 4.5736,
      "mean_token_accuracy": 0.2659416854381561,
      "num_tokens": 89939925.0,
      "step": 39225
    },
    {
      "entropy": 5.166987752914428,
      "epoch": 3.7684918347742555,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003638186801608966,
      "loss": 4.7646,
      "mean_token_accuracy": 0.2427676036953926,
      "num_tokens": 89950866.0,
      "step": 39230
    },
    {
      "entropy": 5.081217622756958,
      "epoch": 3.7689721421709894,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036378717978645067,
      "loss": 4.6621,
      "mean_token_accuracy": 0.250364750623703,
      "num_tokens": 89961936.0,
      "step": 39235
    },
    {
      "entropy": 5.147131156921387,
      "epoch": 3.7694524495677233,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036375567735063,
      "loss": 4.7268,
      "mean_token_accuracy": 0.23876744359731675,
      "num_tokens": 89972589.0,
      "step": 39240
    },
    {
      "entropy": 5.104122591018677,
      "epoch": 3.769932756964457,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003637241728541661,
      "loss": 4.6618,
      "mean_token_accuracy": 0.249017596244812,
      "num_tokens": 89983702.0,
      "step": 39245
    },
    {
      "entropy": 5.115844821929931,
      "epoch": 3.770413064361191,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036369266629779037,
      "loss": 4.565,
      "mean_token_accuracy": 0.2568434402346611,
      "num_tokens": 89993399.0,
      "step": 39250
    },
    {
      "entropy": 5.066199398040771,
      "epoch": 3.770893371757925,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003636611576822342,
      "loss": 4.5962,
      "mean_token_accuracy": 0.2637181654572487,
      "num_tokens": 90005450.0,
      "step": 39255
    },
    {
      "entropy": 5.038927173614502,
      "epoch": 3.771373679154659,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003636296470082293,
      "loss": 4.57,
      "mean_token_accuracy": 0.25317212790250776,
      "num_tokens": 90016403.0,
      "step": 39260
    },
    {
      "entropy": 5.0935667037963865,
      "epoch": 3.7718539865513927,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036359813427650726,
      "loss": 4.6594,
      "mean_token_accuracy": 0.2520760104060173,
      "num_tokens": 90027598.0,
      "step": 39265
    },
    {
      "entropy": 5.02916784286499,
      "epoch": 3.772334293948127,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036356661948779945,
      "loss": 4.5509,
      "mean_token_accuracy": 0.26762392222881315,
      "num_tokens": 90037991.0,
      "step": 39270
    },
    {
      "entropy": 5.078156280517578,
      "epoch": 3.7728146013448605,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036353510264283776,
      "loss": 4.6845,
      "mean_token_accuracy": 0.254996557533741,
      "num_tokens": 90049351.0,
      "step": 39275
    },
    {
      "entropy": 5.0597320079803465,
      "epoch": 3.773294908741595,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003635035837423539,
      "loss": 4.6199,
      "mean_token_accuracy": 0.2583017244935036,
      "num_tokens": 90061208.0,
      "step": 39280
    },
    {
      "entropy": 5.149317264556885,
      "epoch": 3.7737752161383287,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003634720627870797,
      "loss": 4.6721,
      "mean_token_accuracy": 0.25239685028791425,
      "num_tokens": 90072622.0,
      "step": 39285
    },
    {
      "entropy": 5.106220817565918,
      "epoch": 3.7742555235350626,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036344053977774684,
      "loss": 4.6457,
      "mean_token_accuracy": 0.2518836915493011,
      "num_tokens": 90083217.0,
      "step": 39290
    },
    {
      "entropy": 5.076112222671509,
      "epoch": 3.7747358309317964,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003634090147150873,
      "loss": 4.6651,
      "mean_token_accuracy": 0.2524877190589905,
      "num_tokens": 90094733.0,
      "step": 39295
    },
    {
      "entropy": 5.014199256896973,
      "epoch": 3.7752161383285303,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003633774875998329,
      "loss": 4.5661,
      "mean_token_accuracy": 0.2622829332947731,
      "num_tokens": 90105088.0,
      "step": 39300
    },
    {
      "entropy": 5.070011377334595,
      "epoch": 3.775696445725264,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003633459584327157,
      "loss": 4.6622,
      "mean_token_accuracy": 0.2533065304160118,
      "num_tokens": 90115766.0,
      "step": 39305
    },
    {
      "entropy": 5.138667297363281,
      "epoch": 3.776176753121998,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036331442721446763,
      "loss": 4.6968,
      "mean_token_accuracy": 0.2537721902132034,
      "num_tokens": 90126799.0,
      "step": 39310
    },
    {
      "entropy": 5.094526481628418,
      "epoch": 3.776657060518732,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003632828939458209,
      "loss": 4.6318,
      "mean_token_accuracy": 0.2551018610596657,
      "num_tokens": 90139168.0,
      "step": 39315
    },
    {
      "entropy": 5.0414563655853275,
      "epoch": 3.777137367915466,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036325135862750745,
      "loss": 4.5612,
      "mean_token_accuracy": 0.25632727295160296,
      "num_tokens": 90150681.0,
      "step": 39320
    },
    {
      "entropy": 5.1383240699768065,
      "epoch": 3.7776176753121997,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036321982126025947,
      "loss": 4.7286,
      "mean_token_accuracy": 0.24603475630283356,
      "num_tokens": 90161866.0,
      "step": 39325
    },
    {
      "entropy": 5.090524435043335,
      "epoch": 3.7780979827089336,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003631882818448093,
      "loss": 4.714,
      "mean_token_accuracy": 0.24748684018850325,
      "num_tokens": 90174163.0,
      "step": 39330
    },
    {
      "entropy": 5.055618858337402,
      "epoch": 3.7785782901056675,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036315674038188905,
      "loss": 4.6208,
      "mean_token_accuracy": 0.25239803791046145,
      "num_tokens": 90186546.0,
      "step": 39335
    },
    {
      "entropy": 5.01078519821167,
      "epoch": 3.7790585975024014,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000363125196872231,
      "loss": 4.4906,
      "mean_token_accuracy": 0.270379501581192,
      "num_tokens": 90198191.0,
      "step": 39340
    },
    {
      "entropy": 5.047211408615112,
      "epoch": 3.7795389048991357,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003630936513165676,
      "loss": 4.6402,
      "mean_token_accuracy": 0.2577022179961205,
      "num_tokens": 90211407.0,
      "step": 39345
    },
    {
      "entropy": 5.092587852478028,
      "epoch": 3.780019212295869,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003630621037156312,
      "loss": 4.6093,
      "mean_token_accuracy": 0.2521964803338051,
      "num_tokens": 90222118.0,
      "step": 39350
    },
    {
      "entropy": 5.073549509048462,
      "epoch": 3.7804995196926034,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036303055407015437,
      "loss": 4.5956,
      "mean_token_accuracy": 0.2599742740392685,
      "num_tokens": 90232815.0,
      "step": 39355
    },
    {
      "entropy": 5.03845534324646,
      "epoch": 3.7809798270893373,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036299900238086924,
      "loss": 4.624,
      "mean_token_accuracy": 0.25485000014305115,
      "num_tokens": 90243687.0,
      "step": 39360
    },
    {
      "entropy": 5.078618907928467,
      "epoch": 3.781460134486071,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003629674486485087,
      "loss": 4.5985,
      "mean_token_accuracy": 0.2557645007967949,
      "num_tokens": 90254792.0,
      "step": 39365
    },
    {
      "entropy": 5.154414558410645,
      "epoch": 3.781940441882805,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003629358928738052,
      "loss": 4.6693,
      "mean_token_accuracy": 0.249586583673954,
      "num_tokens": 90266627.0,
      "step": 39370
    },
    {
      "entropy": 5.079439544677735,
      "epoch": 3.782420749279539,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036290433505749135,
      "loss": 4.5989,
      "mean_token_accuracy": 0.2588632717728615,
      "num_tokens": 90278171.0,
      "step": 39375
    },
    {
      "entropy": 5.033727025985717,
      "epoch": 3.782901056676273,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003628727752002999,
      "loss": 4.6075,
      "mean_token_accuracy": 0.2527764141559601,
      "num_tokens": 90290031.0,
      "step": 39380
    },
    {
      "entropy": 5.132905387878418,
      "epoch": 3.7833813640730067,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003628412133029636,
      "loss": 4.6525,
      "mean_token_accuracy": 0.25256804525852206,
      "num_tokens": 90301651.0,
      "step": 39385
    },
    {
      "entropy": 5.008990669250489,
      "epoch": 3.7838616714697406,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003628096493662151,
      "loss": 4.5228,
      "mean_token_accuracy": 0.2566218554973602,
      "num_tokens": 90311748.0,
      "step": 39390
    },
    {
      "entropy": 4.959102725982666,
      "epoch": 3.7843419788664745,
      "grad_norm": 1.0,
      "learning_rate": 0.0003627780833907873,
      "loss": 4.5489,
      "mean_token_accuracy": 0.26212636530399325,
      "num_tokens": 90322793.0,
      "step": 39395
    },
    {
      "entropy": 5.077256774902343,
      "epoch": 3.7848222862632084,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036274651537741305,
      "loss": 4.6382,
      "mean_token_accuracy": 0.2517705723643303,
      "num_tokens": 90333808.0,
      "step": 39400
    },
    {
      "entropy": 5.069723653793335,
      "epoch": 3.7853025936599423,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003627149453268252,
      "loss": 4.5752,
      "mean_token_accuracy": 0.25562125593423846,
      "num_tokens": 90345325.0,
      "step": 39405
    },
    {
      "entropy": 5.008271932601929,
      "epoch": 3.785782901056676,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036268337323975685,
      "loss": 4.6081,
      "mean_token_accuracy": 0.252145616710186,
      "num_tokens": 90356772.0,
      "step": 39410
    },
    {
      "entropy": 5.066097927093506,
      "epoch": 3.78626320845341,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036265179911694094,
      "loss": 4.6372,
      "mean_token_accuracy": 0.24930067509412765,
      "num_tokens": 90367642.0,
      "step": 39415
    },
    {
      "entropy": 5.103128862380982,
      "epoch": 3.7867435158501443,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003626202229591105,
      "loss": 4.6423,
      "mean_token_accuracy": 0.24930267333984374,
      "num_tokens": 90378517.0,
      "step": 39420
    },
    {
      "entropy": 5.101875925064087,
      "epoch": 3.787223823246878,
      "grad_norm": 0.875,
      "learning_rate": 0.00036258864476699864,
      "loss": 4.6183,
      "mean_token_accuracy": 0.25889441668987273,
      "num_tokens": 90389944.0,
      "step": 39425
    },
    {
      "entropy": 5.108898782730103,
      "epoch": 3.787704130643612,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003625570645413386,
      "loss": 4.64,
      "mean_token_accuracy": 0.24508608430624007,
      "num_tokens": 90401163.0,
      "step": 39430
    },
    {
      "entropy": 4.990303468704224,
      "epoch": 3.7881844380403455,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003625254822828635,
      "loss": 4.5051,
      "mean_token_accuracy": 0.2692083939909935,
      "num_tokens": 90411383.0,
      "step": 39435
    },
    {
      "entropy": 5.0442695140838625,
      "epoch": 3.78866474543708,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003624938979923065,
      "loss": 4.6215,
      "mean_token_accuracy": 0.25756385922431946,
      "num_tokens": 90424015.0,
      "step": 39440
    },
    {
      "entropy": 5.0853352546691895,
      "epoch": 3.7891450528338138,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036246231167040105,
      "loss": 4.639,
      "mean_token_accuracy": 0.25317939668893813,
      "num_tokens": 90435023.0,
      "step": 39445
    },
    {
      "entropy": 5.112827301025391,
      "epoch": 3.7896253602305476,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003624307233178805,
      "loss": 4.6134,
      "mean_token_accuracy": 0.25047609508037566,
      "num_tokens": 90445474.0,
      "step": 39450
    },
    {
      "entropy": 5.147011041641235,
      "epoch": 3.7901056676272815,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036239913293547807,
      "loss": 4.7506,
      "mean_token_accuracy": 0.240146404504776,
      "num_tokens": 90456568.0,
      "step": 39455
    },
    {
      "entropy": 5.159875440597534,
      "epoch": 3.7905859750240154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036236754052392737,
      "loss": 4.6641,
      "mean_token_accuracy": 0.2489015832543373,
      "num_tokens": 90468309.0,
      "step": 39460
    },
    {
      "entropy": 5.117067289352417,
      "epoch": 3.7910662824207493,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003623359460839618,
      "loss": 4.697,
      "mean_token_accuracy": 0.24726628214120866,
      "num_tokens": 90481014.0,
      "step": 39465
    },
    {
      "entropy": 5.07281723022461,
      "epoch": 3.791546589817483,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003623043496163149,
      "loss": 4.5696,
      "mean_token_accuracy": 0.2586320400238037,
      "num_tokens": 90492039.0,
      "step": 39470
    },
    {
      "entropy": 5.253846502304077,
      "epoch": 3.792026897214217,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003622727511217202,
      "loss": 4.7843,
      "mean_token_accuracy": 0.23554150760173798,
      "num_tokens": 90503680.0,
      "step": 39475
    },
    {
      "entropy": 5.06318507194519,
      "epoch": 3.792507204610951,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003622411506009113,
      "loss": 4.6134,
      "mean_token_accuracy": 0.26357082575559615,
      "num_tokens": 90515416.0,
      "step": 39480
    },
    {
      "entropy": 5.139085102081299,
      "epoch": 3.792987512007685,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003622095480546221,
      "loss": 4.6556,
      "mean_token_accuracy": 0.24769899994134903,
      "num_tokens": 90526374.0,
      "step": 39485
    },
    {
      "entropy": 5.149247026443481,
      "epoch": 3.7934678194044187,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036217794348358606,
      "loss": 4.7023,
      "mean_token_accuracy": 0.25111380219459534,
      "num_tokens": 90536809.0,
      "step": 39490
    },
    {
      "entropy": 5.195858955383301,
      "epoch": 3.793948126801153,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000362146336888537,
      "loss": 4.7275,
      "mean_token_accuracy": 0.2503852441906929,
      "num_tokens": 90548066.0,
      "step": 39495
    },
    {
      "entropy": 5.040206861495972,
      "epoch": 3.7944284341978864,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036211472827020886,
      "loss": 4.5041,
      "mean_token_accuracy": 0.2624347507953644,
      "num_tokens": 90560085.0,
      "step": 39500
    },
    {
      "entropy": 5.149334859848023,
      "epoch": 3.7949087415946208,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003620831176293354,
      "loss": 4.6991,
      "mean_token_accuracy": 0.2504510834813118,
      "num_tokens": 90571119.0,
      "step": 39505
    },
    {
      "entropy": 5.12492880821228,
      "epoch": 3.795389048991354,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036205150496665053,
      "loss": 4.6624,
      "mean_token_accuracy": 0.24360644519329072,
      "num_tokens": 90582294.0,
      "step": 39510
    },
    {
      "entropy": 5.069000005722046,
      "epoch": 3.7958693563880885,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036201989028288815,
      "loss": 4.6563,
      "mean_token_accuracy": 0.24892071336507798,
      "num_tokens": 90593681.0,
      "step": 39515
    },
    {
      "entropy": 5.103442144393921,
      "epoch": 3.7963496637848224,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036198827357878234,
      "loss": 4.6177,
      "mean_token_accuracy": 0.2518701612949371,
      "num_tokens": 90605202.0,
      "step": 39520
    },
    {
      "entropy": 5.088356256484985,
      "epoch": 3.7968299711815563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003619566548550671,
      "loss": 4.6532,
      "mean_token_accuracy": 0.2498650074005127,
      "num_tokens": 90617456.0,
      "step": 39525
    },
    {
      "entropy": 5.083259248733521,
      "epoch": 3.79731027857829,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003619250341124765,
      "loss": 4.6191,
      "mean_token_accuracy": 0.2560378283262253,
      "num_tokens": 90630631.0,
      "step": 39530
    },
    {
      "entropy": 5.242424488067627,
      "epoch": 3.797790585975024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003618934113517448,
      "loss": 4.8146,
      "mean_token_accuracy": 0.23918758630752562,
      "num_tokens": 90641996.0,
      "step": 39535
    },
    {
      "entropy": 5.127087783813477,
      "epoch": 3.798270893371758,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003618617865736061,
      "loss": 4.6013,
      "mean_token_accuracy": 0.24794234186410904,
      "num_tokens": 90654435.0,
      "step": 39540
    },
    {
      "entropy": 4.993952846527099,
      "epoch": 3.798751200768492,
      "grad_norm": 1.0,
      "learning_rate": 0.00036183015977879463,
      "loss": 4.5411,
      "mean_token_accuracy": 0.2574880555272102,
      "num_tokens": 90665569.0,
      "step": 39545
    },
    {
      "entropy": 5.041241550445557,
      "epoch": 3.7992315081652257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036179853096804474,
      "loss": 4.6058,
      "mean_token_accuracy": 0.2551989614963531,
      "num_tokens": 90676530.0,
      "step": 39550
    },
    {
      "entropy": 5.003446865081787,
      "epoch": 3.7997118155619596,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036176690014209055,
      "loss": 4.5502,
      "mean_token_accuracy": 0.261370949447155,
      "num_tokens": 90688206.0,
      "step": 39555
    },
    {
      "entropy": 5.106279706954956,
      "epoch": 3.8001921229586935,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003617352673016667,
      "loss": 4.6618,
      "mean_token_accuracy": 0.24849550276994706,
      "num_tokens": 90700683.0,
      "step": 39560
    },
    {
      "entropy": 5.0722497463226315,
      "epoch": 3.8006724303554273,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036170363244750756,
      "loss": 4.5709,
      "mean_token_accuracy": 0.2540323346853256,
      "num_tokens": 90711653.0,
      "step": 39565
    },
    {
      "entropy": 5.046086692810059,
      "epoch": 3.8011527377521612,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003616719955803474,
      "loss": 4.5571,
      "mean_token_accuracy": 0.2639791116118431,
      "num_tokens": 90722658.0,
      "step": 39570
    },
    {
      "entropy": 5.044203472137451,
      "epoch": 3.801633045148895,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003616403567009209,
      "loss": 4.5949,
      "mean_token_accuracy": 0.25656537860631945,
      "num_tokens": 90734564.0,
      "step": 39575
    },
    {
      "entropy": 5.065617465972901,
      "epoch": 3.8021133525456294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003616087158099626,
      "loss": 4.6725,
      "mean_token_accuracy": 0.25334840416908266,
      "num_tokens": 90747318.0,
      "step": 39580
    },
    {
      "entropy": 5.079186582565308,
      "epoch": 3.802593659942363,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003615770729082071,
      "loss": 4.6263,
      "mean_token_accuracy": 0.2564228191971779,
      "num_tokens": 90759157.0,
      "step": 39585
    },
    {
      "entropy": 5.09177451133728,
      "epoch": 3.803073967339097,
      "grad_norm": 1.09375,
      "learning_rate": 0.000361545427996389,
      "loss": 4.5794,
      "mean_token_accuracy": 0.2537792816758156,
      "num_tokens": 90768788.0,
      "step": 39590
    },
    {
      "entropy": 5.07339825630188,
      "epoch": 3.803554274735831,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003615137810752431,
      "loss": 4.5981,
      "mean_token_accuracy": 0.2572837248444557,
      "num_tokens": 90779873.0,
      "step": 39595
    },
    {
      "entropy": 5.118340873718262,
      "epoch": 3.804034582132565,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036148213214550406,
      "loss": 4.6844,
      "mean_token_accuracy": 0.2474487006664276,
      "num_tokens": 90792075.0,
      "step": 39600
    },
    {
      "entropy": 5.075357151031494,
      "epoch": 3.804514889529299,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003614504812079068,
      "loss": 4.5417,
      "mean_token_accuracy": 0.25427769869565964,
      "num_tokens": 90805362.0,
      "step": 39605
    },
    {
      "entropy": 5.106225109100341,
      "epoch": 3.8049951969260327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036141882826318597,
      "loss": 4.678,
      "mean_token_accuracy": 0.24600719660520554,
      "num_tokens": 90817316.0,
      "step": 39610
    },
    {
      "entropy": 5.151411867141723,
      "epoch": 3.8054755043227666,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003613871733120766,
      "loss": 4.7273,
      "mean_token_accuracy": 0.24300359934568405,
      "num_tokens": 90829688.0,
      "step": 39615
    },
    {
      "entropy": 5.114664745330811,
      "epoch": 3.8059558117195005,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036135551635531354,
      "loss": 4.6116,
      "mean_token_accuracy": 0.25943870544433595,
      "num_tokens": 90840154.0,
      "step": 39620
    },
    {
      "entropy": 5.030361604690552,
      "epoch": 3.8064361191162344,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003613238573936319,
      "loss": 4.622,
      "mean_token_accuracy": 0.25357009619474413,
      "num_tokens": 90852023.0,
      "step": 39625
    },
    {
      "entropy": 5.119562578201294,
      "epoch": 3.8069164265129682,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003612921964277665,
      "loss": 4.6588,
      "mean_token_accuracy": 0.24795475006103515,
      "num_tokens": 90864574.0,
      "step": 39630
    },
    {
      "entropy": 5.060334873199463,
      "epoch": 3.807396733909702,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003612605334584526,
      "loss": 4.5423,
      "mean_token_accuracy": 0.2681539088487625,
      "num_tokens": 90875810.0,
      "step": 39635
    },
    {
      "entropy": 5.099946403503418,
      "epoch": 3.807877041306436,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036122886848642536,
      "loss": 4.701,
      "mean_token_accuracy": 0.2514104098081589,
      "num_tokens": 90886946.0,
      "step": 39640
    },
    {
      "entropy": 5.072085762023926,
      "epoch": 3.80835734870317,
      "grad_norm": 1.125,
      "learning_rate": 0.0003611972015124197,
      "loss": 4.5578,
      "mean_token_accuracy": 0.25922838747501376,
      "num_tokens": 90897949.0,
      "step": 39645
    },
    {
      "entropy": 5.187338209152221,
      "epoch": 3.8088376560999038,
      "grad_norm": 0.953125,
      "learning_rate": 0.000361165532537171,
      "loss": 4.7051,
      "mean_token_accuracy": 0.24683507680892944,
      "num_tokens": 90910279.0,
      "step": 39650
    },
    {
      "entropy": 5.083806276321411,
      "epoch": 3.809317963496638,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003611338615614145,
      "loss": 4.6787,
      "mean_token_accuracy": 0.25227547287940977,
      "num_tokens": 90922917.0,
      "step": 39655
    },
    {
      "entropy": 5.023042058944702,
      "epoch": 3.8097982708933715,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036110218858588554,
      "loss": 4.5265,
      "mean_token_accuracy": 0.26236049085855484,
      "num_tokens": 90933943.0,
      "step": 39660
    },
    {
      "entropy": 5.043915224075318,
      "epoch": 3.810278578290106,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036107051361131934,
      "loss": 4.5556,
      "mean_token_accuracy": 0.26132817566394806,
      "num_tokens": 90945626.0,
      "step": 39665
    },
    {
      "entropy": 5.076431798934936,
      "epoch": 3.8107588856868397,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036103883663845155,
      "loss": 4.551,
      "mean_token_accuracy": 0.25937163829803467,
      "num_tokens": 90956969.0,
      "step": 39670
    },
    {
      "entropy": 5.067569351196289,
      "epoch": 3.8112391930835736,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003610071576680173,
      "loss": 4.6048,
      "mean_token_accuracy": 0.25469855666160585,
      "num_tokens": 90968916.0,
      "step": 39675
    },
    {
      "entropy": 5.127260541915893,
      "epoch": 3.8117195004803075,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036097547670075226,
      "loss": 4.6708,
      "mean_token_accuracy": 0.24502015560865403,
      "num_tokens": 90979242.0,
      "step": 39680
    },
    {
      "entropy": 5.032222700119019,
      "epoch": 3.8121998078770414,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000360943793737392,
      "loss": 4.6206,
      "mean_token_accuracy": 0.25419254004955294,
      "num_tokens": 90991374.0,
      "step": 39685
    },
    {
      "entropy": 5.009206295013428,
      "epoch": 3.8126801152737753,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000360912108778672,
      "loss": 4.5537,
      "mean_token_accuracy": 0.2640092447400093,
      "num_tokens": 91002648.0,
      "step": 39690
    },
    {
      "entropy": 5.110065126419068,
      "epoch": 3.813160422670509,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000360880421825328,
      "loss": 4.6059,
      "mean_token_accuracy": 0.2588018149137497,
      "num_tokens": 91013122.0,
      "step": 39695
    },
    {
      "entropy": 5.095313596725464,
      "epoch": 3.813640730067243,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003608487328780956,
      "loss": 4.6391,
      "mean_token_accuracy": 0.24331542253494262,
      "num_tokens": 91024638.0,
      "step": 39700
    },
    {
      "entropy": 5.096485996246338,
      "epoch": 3.814121037463977,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036081704193771056,
      "loss": 4.6521,
      "mean_token_accuracy": 0.24997472763061523,
      "num_tokens": 91036542.0,
      "step": 39705
    },
    {
      "entropy": 5.164034128189087,
      "epoch": 3.814601344860711,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003607853490049085,
      "loss": 4.6891,
      "mean_token_accuracy": 0.24839899837970733,
      "num_tokens": 91048188.0,
      "step": 39710
    },
    {
      "entropy": 5.066634178161621,
      "epoch": 3.8150816522574447,
      "grad_norm": 1.0,
      "learning_rate": 0.0003607536540804255,
      "loss": 4.5801,
      "mean_token_accuracy": 0.26041693687438966,
      "num_tokens": 91059761.0,
      "step": 39715
    },
    {
      "entropy": 5.154553937911987,
      "epoch": 3.8155619596541785,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036072195716499726,
      "loss": 4.712,
      "mean_token_accuracy": 0.2440507784485817,
      "num_tokens": 91071794.0,
      "step": 39720
    },
    {
      "entropy": 5.116303205490112,
      "epoch": 3.8160422670509124,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036069025825935974,
      "loss": 4.6719,
      "mean_token_accuracy": 0.2528825506567955,
      "num_tokens": 91083247.0,
      "step": 39725
    },
    {
      "entropy": 5.07110333442688,
      "epoch": 3.8165225744476468,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036065855736424886,
      "loss": 4.6038,
      "mean_token_accuracy": 0.24914286881685258,
      "num_tokens": 91096076.0,
      "step": 39730
    },
    {
      "entropy": 5.061686849594116,
      "epoch": 3.81700288184438,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003606268544804006,
      "loss": 4.5782,
      "mean_token_accuracy": 0.2528866395354271,
      "num_tokens": 91106959.0,
      "step": 39735
    },
    {
      "entropy": 5.126953887939453,
      "epoch": 3.8174831892411145,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036059514960855104,
      "loss": 4.6805,
      "mean_token_accuracy": 0.24847826063632966,
      "num_tokens": 91117259.0,
      "step": 39740
    },
    {
      "entropy": 5.13294620513916,
      "epoch": 3.817963496637848,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003605634427494363,
      "loss": 4.6612,
      "mean_token_accuracy": 0.25081645846366885,
      "num_tokens": 91129085.0,
      "step": 39745
    },
    {
      "entropy": 5.170076608657837,
      "epoch": 3.8184438040345823,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003605317339037925,
      "loss": 4.6919,
      "mean_token_accuracy": 0.2475058376789093,
      "num_tokens": 91139626.0,
      "step": 39750
    },
    {
      "entropy": 5.087055015563965,
      "epoch": 3.818924111431316,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036050002307235575,
      "loss": 4.6193,
      "mean_token_accuracy": 0.2540093719959259,
      "num_tokens": 91150802.0,
      "step": 39755
    },
    {
      "entropy": 5.187627458572388,
      "epoch": 3.81940441882805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003604683102558624,
      "loss": 4.6357,
      "mean_token_accuracy": 0.24811921268701553,
      "num_tokens": 91161789.0,
      "step": 39760
    },
    {
      "entropy": 5.108985328674317,
      "epoch": 3.819884726224784,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036043659545504866,
      "loss": 4.6614,
      "mean_token_accuracy": 0.25624125450849533,
      "num_tokens": 91173017.0,
      "step": 39765
    },
    {
      "entropy": 5.089702892303467,
      "epoch": 3.820365033621518,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003604048786706509,
      "loss": 4.6679,
      "mean_token_accuracy": 0.24779223650693893,
      "num_tokens": 91184915.0,
      "step": 39770
    },
    {
      "entropy": 5.159673547744751,
      "epoch": 3.8208453410182517,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003603731599034054,
      "loss": 4.6523,
      "mean_token_accuracy": 0.2547389566898346,
      "num_tokens": 91196711.0,
      "step": 39775
    },
    {
      "entropy": 5.017160701751709,
      "epoch": 3.8213256484149856,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00036034143915404865,
      "loss": 4.5307,
      "mean_token_accuracy": 0.26253381967544553,
      "num_tokens": 91208823.0,
      "step": 39780
    },
    {
      "entropy": 4.961542463302612,
      "epoch": 3.8218059558117194,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003603097164233171,
      "loss": 4.5203,
      "mean_token_accuracy": 0.26359042823314666,
      "num_tokens": 91220175.0,
      "step": 39785
    },
    {
      "entropy": 5.076615524291992,
      "epoch": 3.8222862632084533,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003602779917119473,
      "loss": 4.6289,
      "mean_token_accuracy": 0.2524866253137589,
      "num_tokens": 91232224.0,
      "step": 39790
    },
    {
      "entropy": 5.079014730453491,
      "epoch": 3.822766570605187,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036024626502067565,
      "loss": 4.5667,
      "mean_token_accuracy": 0.2609082728624344,
      "num_tokens": 91244626.0,
      "step": 39795
    },
    {
      "entropy": 5.111316776275634,
      "epoch": 3.823246878001921,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036021453635023894,
      "loss": 4.6102,
      "mean_token_accuracy": 0.25406000316143035,
      "num_tokens": 91257721.0,
      "step": 39800
    },
    {
      "entropy": 5.042025709152222,
      "epoch": 3.8237271853986554,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00036018280570137374,
      "loss": 4.5864,
      "mean_token_accuracy": 0.2593486472964287,
      "num_tokens": 91269681.0,
      "step": 39805
    },
    {
      "entropy": 5.013385391235351,
      "epoch": 3.824207492795389,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003601510730748167,
      "loss": 4.5503,
      "mean_token_accuracy": 0.2596085727214813,
      "num_tokens": 91280952.0,
      "step": 39810
    },
    {
      "entropy": 5.153445386886597,
      "epoch": 3.824687800192123,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003601193384713046,
      "loss": 4.7247,
      "mean_token_accuracy": 0.24508994966745376,
      "num_tokens": 91291668.0,
      "step": 39815
    },
    {
      "entropy": 5.062478113174438,
      "epoch": 3.8251681075888566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003600876018915742,
      "loss": 4.6565,
      "mean_token_accuracy": 0.24398599565029144,
      "num_tokens": 91304103.0,
      "step": 39820
    },
    {
      "entropy": 5.116948986053467,
      "epoch": 3.825648414985591,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003600558633363624,
      "loss": 4.6263,
      "mean_token_accuracy": 0.24790338575839996,
      "num_tokens": 91316269.0,
      "step": 39825
    },
    {
      "entropy": 5.077988719940185,
      "epoch": 3.826128722382325,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003600241228064059,
      "loss": 4.5405,
      "mean_token_accuracy": 0.2618341401219368,
      "num_tokens": 91327645.0,
      "step": 39830
    },
    {
      "entropy": 5.065535211563111,
      "epoch": 3.8266090297790587,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035999238030244183,
      "loss": 4.5526,
      "mean_token_accuracy": 0.2615657106041908,
      "num_tokens": 91339834.0,
      "step": 39835
    },
    {
      "entropy": 5.0859215259552,
      "epoch": 3.8270893371757926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035996063582520704,
      "loss": 4.5794,
      "mean_token_accuracy": 0.2538136810064316,
      "num_tokens": 91351636.0,
      "step": 39840
    },
    {
      "entropy": 5.0769429206848145,
      "epoch": 3.8275696445725265,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035992888937543856,
      "loss": 4.6403,
      "mean_token_accuracy": 0.2576439559459686,
      "num_tokens": 91363750.0,
      "step": 39845
    },
    {
      "entropy": 5.079017257690429,
      "epoch": 3.8280499519692603,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035989714095387344,
      "loss": 4.6212,
      "mean_token_accuracy": 0.25883476734161376,
      "num_tokens": 91374900.0,
      "step": 39850
    },
    {
      "entropy": 4.978415441513062,
      "epoch": 3.8285302593659942,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035986539056124877,
      "loss": 4.5773,
      "mean_token_accuracy": 0.25607665032148363,
      "num_tokens": 91385810.0,
      "step": 39855
    },
    {
      "entropy": 5.066156339645386,
      "epoch": 3.829010566762728,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003598336381983018,
      "loss": 4.5708,
      "mean_token_accuracy": 0.25148366391658783,
      "num_tokens": 91396209.0,
      "step": 39860
    },
    {
      "entropy": 5.072734785079956,
      "epoch": 3.829490874159462,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035980188386576967,
      "loss": 4.6326,
      "mean_token_accuracy": 0.2512514889240265,
      "num_tokens": 91406799.0,
      "step": 39865
    },
    {
      "entropy": 5.027131271362305,
      "epoch": 3.829971181556196,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003597701275643896,
      "loss": 4.5746,
      "mean_token_accuracy": 0.26071922332048414,
      "num_tokens": 91418868.0,
      "step": 39870
    },
    {
      "entropy": 5.074025392532349,
      "epoch": 3.8304514889529298,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003597383692948988,
      "loss": 4.6131,
      "mean_token_accuracy": 0.2532332733273506,
      "num_tokens": 91430864.0,
      "step": 39875
    },
    {
      "entropy": 5.028220987319946,
      "epoch": 3.8309317963496636,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003597066090580347,
      "loss": 4.5493,
      "mean_token_accuracy": 0.2567356154322624,
      "num_tokens": 91442107.0,
      "step": 39880
    },
    {
      "entropy": 5.068823003768921,
      "epoch": 3.8314121037463975,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003596748468545348,
      "loss": 4.5797,
      "mean_token_accuracy": 0.25467171221971513,
      "num_tokens": 91453029.0,
      "step": 39885
    },
    {
      "entropy": 5.040734338760376,
      "epoch": 3.831892411143132,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035964308268513623,
      "loss": 4.6123,
      "mean_token_accuracy": 0.25875951796770097,
      "num_tokens": 91463744.0,
      "step": 39890
    },
    {
      "entropy": 5.123314285278321,
      "epoch": 3.8323727185398653,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035961131655057674,
      "loss": 4.6501,
      "mean_token_accuracy": 0.2585961163043976,
      "num_tokens": 91474692.0,
      "step": 39895
    },
    {
      "entropy": 5.05369200706482,
      "epoch": 3.8328530259365996,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003595795484515936,
      "loss": 4.57,
      "mean_token_accuracy": 0.2590760841965675,
      "num_tokens": 91485641.0,
      "step": 39900
    },
    {
      "entropy": 5.027422857284546,
      "epoch": 3.8333333333333335,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035954777838892454,
      "loss": 4.5687,
      "mean_token_accuracy": 0.25732116848230363,
      "num_tokens": 91497516.0,
      "step": 39905
    },
    {
      "entropy": 5.162716436386108,
      "epoch": 3.8338136407300674,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003595160063633072,
      "loss": 4.7226,
      "mean_token_accuracy": 0.24991143345832825,
      "num_tokens": 91508254.0,
      "step": 39910
    },
    {
      "entropy": 5.092141914367676,
      "epoch": 3.8342939481268012,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003594842323754791,
      "loss": 4.617,
      "mean_token_accuracy": 0.24993027597665787,
      "num_tokens": 91519298.0,
      "step": 39915
    },
    {
      "entropy": 5.10280032157898,
      "epoch": 3.834774255523535,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035945245642617795,
      "loss": 4.6015,
      "mean_token_accuracy": 0.25232370793819425,
      "num_tokens": 91530184.0,
      "step": 39920
    },
    {
      "entropy": 5.053603744506836,
      "epoch": 3.835254562920269,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035942067851614165,
      "loss": 4.6602,
      "mean_token_accuracy": 0.2502201244235039,
      "num_tokens": 91541569.0,
      "step": 39925
    },
    {
      "entropy": 5.14578766822815,
      "epoch": 3.835734870317003,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035938889864610776,
      "loss": 4.6504,
      "mean_token_accuracy": 0.24795565158128738,
      "num_tokens": 91553090.0,
      "step": 39930
    },
    {
      "entropy": 5.096676826477051,
      "epoch": 3.8362151777137368,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035935711681681423,
      "loss": 4.5701,
      "mean_token_accuracy": 0.26698175370693206,
      "num_tokens": 91564343.0,
      "step": 39935
    },
    {
      "entropy": 5.083163690567017,
      "epoch": 3.8366954851104706,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000359325333028999,
      "loss": 4.63,
      "mean_token_accuracy": 0.2537864461541176,
      "num_tokens": 91575592.0,
      "step": 39940
    },
    {
      "entropy": 5.001988077163697,
      "epoch": 3.8371757925072045,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035929354728339986,
      "loss": 4.5899,
      "mean_token_accuracy": 0.24817996770143508,
      "num_tokens": 91587202.0,
      "step": 39945
    },
    {
      "entropy": 5.126279067993164,
      "epoch": 3.8376560999039384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003592617595807549,
      "loss": 4.6758,
      "mean_token_accuracy": 0.24941549003124236,
      "num_tokens": 91599753.0,
      "step": 39950
    },
    {
      "entropy": 5.160743713378906,
      "epoch": 3.8381364073006723,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035922996992180213,
      "loss": 4.7257,
      "mean_token_accuracy": 0.2501572549343109,
      "num_tokens": 91610999.0,
      "step": 39955
    },
    {
      "entropy": 5.077878999710083,
      "epoch": 3.838616714697406,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035919817830727945,
      "loss": 4.5847,
      "mean_token_accuracy": 0.2596787393093109,
      "num_tokens": 91624034.0,
      "step": 39960
    },
    {
      "entropy": 5.091542673110962,
      "epoch": 3.8390970220941405,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003591663847379252,
      "loss": 4.6071,
      "mean_token_accuracy": 0.25455514788627626,
      "num_tokens": 91635749.0,
      "step": 39965
    },
    {
      "entropy": 5.04342679977417,
      "epoch": 3.839577329490874,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003591345892144774,
      "loss": 4.6175,
      "mean_token_accuracy": 0.25704940110445024,
      "num_tokens": 91646680.0,
      "step": 39970
    },
    {
      "entropy": 5.040571975708008,
      "epoch": 3.8400576368876083,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035910279173767426,
      "loss": 4.5197,
      "mean_token_accuracy": 0.26381816864013674,
      "num_tokens": 91658329.0,
      "step": 39975
    },
    {
      "entropy": 5.032323789596558,
      "epoch": 3.840537944284342,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000359070992308254,
      "loss": 4.5754,
      "mean_token_accuracy": 0.25609817653894423,
      "num_tokens": 91669514.0,
      "step": 39980
    },
    {
      "entropy": 5.110657262802124,
      "epoch": 3.841018251681076,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000359039190926955,
      "loss": 4.6826,
      "mean_token_accuracy": 0.24916255623102188,
      "num_tokens": 91682073.0,
      "step": 39985
    },
    {
      "entropy": 5.06100435256958,
      "epoch": 3.84149855907781,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003590073875945154,
      "loss": 4.5845,
      "mean_token_accuracy": 0.25504290610551833,
      "num_tokens": 91692848.0,
      "step": 39990
    },
    {
      "entropy": 5.04906268119812,
      "epoch": 3.841978866474544,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00035897558231167375,
      "loss": 4.5595,
      "mean_token_accuracy": 0.25833175480365755,
      "num_tokens": 91704143.0,
      "step": 39995
    },
    {
      "entropy": 5.089449119567871,
      "epoch": 3.8424591738712777,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003589437750791685,
      "loss": 4.6298,
      "mean_token_accuracy": 0.25176094472408295,
      "num_tokens": 91714701.0,
      "step": 40000
    },
    {
      "entropy": 5.111348295211792,
      "epoch": 3.8429394812680115,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035891196589773805,
      "loss": 4.6507,
      "mean_token_accuracy": 0.25320031940937043,
      "num_tokens": 91727046.0,
      "step": 40005
    },
    {
      "entropy": 5.114036560058594,
      "epoch": 3.8434197886647454,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035888015476812085,
      "loss": 4.6506,
      "mean_token_accuracy": 0.2508819416165352,
      "num_tokens": 91738362.0,
      "step": 40010
    },
    {
      "entropy": 5.035728597640992,
      "epoch": 3.8439000960614793,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003588483416910557,
      "loss": 4.5618,
      "mean_token_accuracy": 0.257535058259964,
      "num_tokens": 91748048.0,
      "step": 40015
    },
    {
      "entropy": 5.129156589508057,
      "epoch": 3.844380403458213,
      "grad_norm": 1.0,
      "learning_rate": 0.0003588165266672809,
      "loss": 4.713,
      "mean_token_accuracy": 0.24709147363901138,
      "num_tokens": 91759376.0,
      "step": 40020
    },
    {
      "entropy": 5.0595849514007565,
      "epoch": 3.844860710854947,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035878470969753524,
      "loss": 4.6113,
      "mean_token_accuracy": 0.2522576987743378,
      "num_tokens": 91770327.0,
      "step": 40025
    },
    {
      "entropy": 5.008842897415161,
      "epoch": 3.845341018251681,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035875289078255747,
      "loss": 4.5903,
      "mean_token_accuracy": 0.2575593784451485,
      "num_tokens": 91782648.0,
      "step": 40030
    },
    {
      "entropy": 5.072422122955322,
      "epoch": 3.845821325648415,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035872106992308614,
      "loss": 4.6025,
      "mean_token_accuracy": 0.2606623977422714,
      "num_tokens": 91793514.0,
      "step": 40035
    },
    {
      "entropy": 5.089419174194336,
      "epoch": 3.846301633045149,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003586892471198603,
      "loss": 4.5831,
      "mean_token_accuracy": 0.25552795082330704,
      "num_tokens": 91804136.0,
      "step": 40040
    },
    {
      "entropy": 5.086842346191406,
      "epoch": 3.8467819404418826,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003586574223736186,
      "loss": 4.5877,
      "mean_token_accuracy": 0.2560537040233612,
      "num_tokens": 91814278.0,
      "step": 40045
    },
    {
      "entropy": 5.097049808502197,
      "epoch": 3.847262247838617,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035862559568509994,
      "loss": 4.6485,
      "mean_token_accuracy": 0.252693709731102,
      "num_tokens": 91826663.0,
      "step": 40050
    },
    {
      "entropy": 5.037154483795166,
      "epoch": 3.8477425552353504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003585937670550433,
      "loss": 4.4969,
      "mean_token_accuracy": 0.26508804261684416,
      "num_tokens": 91838577.0,
      "step": 40055
    },
    {
      "entropy": 5.006822109222412,
      "epoch": 3.8482228626320847,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035856193648418753,
      "loss": 4.5601,
      "mean_token_accuracy": 0.2625537753105164,
      "num_tokens": 91848321.0,
      "step": 40060
    },
    {
      "entropy": 5.0550071716308596,
      "epoch": 3.8487031700288186,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035853010397327186,
      "loss": 4.5842,
      "mean_token_accuracy": 0.25772023350000384,
      "num_tokens": 91859303.0,
      "step": 40065
    },
    {
      "entropy": 5.202198219299317,
      "epoch": 3.8491834774255524,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003584982695230351,
      "loss": 4.7837,
      "mean_token_accuracy": 0.24481260329484938,
      "num_tokens": 91871187.0,
      "step": 40070
    },
    {
      "entropy": 5.07805027961731,
      "epoch": 3.8496637848222863,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003584664331342166,
      "loss": 4.6389,
      "mean_token_accuracy": 0.2495470404624939,
      "num_tokens": 91883185.0,
      "step": 40075
    },
    {
      "entropy": 5.0855810165405275,
      "epoch": 3.85014409221902,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003584345948075552,
      "loss": 4.6316,
      "mean_token_accuracy": 0.25676064640283586,
      "num_tokens": 91894396.0,
      "step": 40080
    },
    {
      "entropy": 5.1563718795776365,
      "epoch": 3.850624399615754,
      "grad_norm": 1.0,
      "learning_rate": 0.0003584027545437903,
      "loss": 4.6788,
      "mean_token_accuracy": 0.2469526320695877,
      "num_tokens": 91906776.0,
      "step": 40085
    },
    {
      "entropy": 5.050556659698486,
      "epoch": 3.851104707012488,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035837091234366107,
      "loss": 4.6494,
      "mean_token_accuracy": 0.24630804657936095,
      "num_tokens": 91918341.0,
      "step": 40090
    },
    {
      "entropy": 5.047230195999146,
      "epoch": 3.851585014409222,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000358339068207907,
      "loss": 4.574,
      "mean_token_accuracy": 0.256154465675354,
      "num_tokens": 91931464.0,
      "step": 40095
    },
    {
      "entropy": 5.015864992141724,
      "epoch": 3.8520653218059557,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035830722213726695,
      "loss": 4.5111,
      "mean_token_accuracy": 0.2600944548845291,
      "num_tokens": 91941364.0,
      "step": 40100
    },
    {
      "entropy": 5.017479944229126,
      "epoch": 3.8525456292026896,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035827537413248077,
      "loss": 4.5452,
      "mean_token_accuracy": 0.25962784290313723,
      "num_tokens": 91951610.0,
      "step": 40105
    },
    {
      "entropy": 5.094094371795654,
      "epoch": 3.8530259365994235,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003582435241942877,
      "loss": 4.5994,
      "mean_token_accuracy": 0.25570460259914396,
      "num_tokens": 91962073.0,
      "step": 40110
    },
    {
      "entropy": 5.131180095672607,
      "epoch": 3.8535062439961574,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035821167232342706,
      "loss": 4.6983,
      "mean_token_accuracy": 0.24846826791763305,
      "num_tokens": 91974570.0,
      "step": 40115
    },
    {
      "entropy": 5.074732255935669,
      "epoch": 3.8539865513928913,
      "grad_norm": 1.0,
      "learning_rate": 0.0003581798185206385,
      "loss": 4.6507,
      "mean_token_accuracy": 0.25092553347349167,
      "num_tokens": 91986556.0,
      "step": 40120
    },
    {
      "entropy": 5.078423881530762,
      "epoch": 3.8544668587896256,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035814796278666166,
      "loss": 4.622,
      "mean_token_accuracy": 0.2547515958547592,
      "num_tokens": 91997855.0,
      "step": 40125
    },
    {
      "entropy": 5.0927704811096195,
      "epoch": 3.854947166186359,
      "grad_norm": 1.0,
      "learning_rate": 0.00035811610512223585,
      "loss": 4.6464,
      "mean_token_accuracy": 0.2530526384711266,
      "num_tokens": 92009291.0,
      "step": 40130
    },
    {
      "entropy": 5.073155403137207,
      "epoch": 3.8554274735830933,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000358084245528101,
      "loss": 4.5839,
      "mean_token_accuracy": 0.26004245281219485,
      "num_tokens": 92020436.0,
      "step": 40135
    },
    {
      "entropy": 5.146393966674805,
      "epoch": 3.8559077809798272,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035805238400499667,
      "loss": 4.6805,
      "mean_token_accuracy": 0.24784745872020722,
      "num_tokens": 92032153.0,
      "step": 40140
    },
    {
      "entropy": 5.15438666343689,
      "epoch": 3.856388088376561,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003580205205536626,
      "loss": 4.6653,
      "mean_token_accuracy": 0.2526441514492035,
      "num_tokens": 92042003.0,
      "step": 40145
    },
    {
      "entropy": 5.0652210235595705,
      "epoch": 3.856868395773295,
      "grad_norm": 1.0,
      "learning_rate": 0.0003579886551748385,
      "loss": 4.6456,
      "mean_token_accuracy": 0.2521187663078308,
      "num_tokens": 92052786.0,
      "step": 40150
    },
    {
      "entropy": 5.118894338607788,
      "epoch": 3.857348703170029,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00035795678786926434,
      "loss": 4.6861,
      "mean_token_accuracy": 0.24581236094236375,
      "num_tokens": 92064114.0,
      "step": 40155
    },
    {
      "entropy": 5.096407461166382,
      "epoch": 3.8578290105667628,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035792491863767984,
      "loss": 4.668,
      "mean_token_accuracy": 0.25234067291021345,
      "num_tokens": 92076429.0,
      "step": 40160
    },
    {
      "entropy": 5.085876417160034,
      "epoch": 3.8583093179634966,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035789304748082496,
      "loss": 4.6613,
      "mean_token_accuracy": 0.25047758370637896,
      "num_tokens": 92089121.0,
      "step": 40165
    },
    {
      "entropy": 5.119608974456787,
      "epoch": 3.8587896253602305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035786117439943967,
      "loss": 4.6779,
      "mean_token_accuracy": 0.2550880238413811,
      "num_tokens": 92099678.0,
      "step": 40170
    },
    {
      "entropy": 5.077173566818237,
      "epoch": 3.8592699327569644,
      "grad_norm": 1.0,
      "learning_rate": 0.00035782929939426393,
      "loss": 4.5507,
      "mean_token_accuracy": 0.25689494907855986,
      "num_tokens": 92110842.0,
      "step": 40175
    },
    {
      "entropy": 5.0632080078125,
      "epoch": 3.8597502401536983,
      "grad_norm": 1.0,
      "learning_rate": 0.0003577974224660379,
      "loss": 4.6372,
      "mean_token_accuracy": 0.2491772010922432,
      "num_tokens": 92121874.0,
      "step": 40180
    },
    {
      "entropy": 5.0459051609039305,
      "epoch": 3.860230547550432,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003577655436155015,
      "loss": 4.5508,
      "mean_token_accuracy": 0.26193372160196304,
      "num_tokens": 92133044.0,
      "step": 40185
    },
    {
      "entropy": 5.103451776504516,
      "epoch": 3.860710854947166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003577336628433949,
      "loss": 4.7146,
      "mean_token_accuracy": 0.24182055294513702,
      "num_tokens": 92144733.0,
      "step": 40190
    },
    {
      "entropy": 5.163102054595948,
      "epoch": 3.8611911623439,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035770178015045843,
      "loss": 4.6801,
      "mean_token_accuracy": 0.24717365056276322,
      "num_tokens": 92156307.0,
      "step": 40195
    },
    {
      "entropy": 5.22835955619812,
      "epoch": 3.8616714697406342,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003576698955374321,
      "loss": 4.7682,
      "mean_token_accuracy": 0.24265414327383042,
      "num_tokens": 92167347.0,
      "step": 40200
    },
    {
      "entropy": 5.046129846572876,
      "epoch": 3.8621517771373677,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003576380090050564,
      "loss": 4.6025,
      "mean_token_accuracy": 0.25760849714279177,
      "num_tokens": 92180735.0,
      "step": 40205
    },
    {
      "entropy": 5.1657397747039795,
      "epoch": 3.862632084534102,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003576061205540714,
      "loss": 4.7308,
      "mean_token_accuracy": 0.2437412291765213,
      "num_tokens": 92191449.0,
      "step": 40210
    },
    {
      "entropy": 5.144513416290283,
      "epoch": 3.863112391930836,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003575742301852176,
      "loss": 4.6938,
      "mean_token_accuracy": 0.2454361766576767,
      "num_tokens": 92203720.0,
      "step": 40215
    },
    {
      "entropy": 5.099646282196045,
      "epoch": 3.8635926993275698,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003575423378992354,
      "loss": 4.6278,
      "mean_token_accuracy": 0.24799599796533583,
      "num_tokens": 92215649.0,
      "step": 40220
    },
    {
      "entropy": 5.046434593200684,
      "epoch": 3.8640730067243036,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003575104436968652,
      "loss": 4.6307,
      "mean_token_accuracy": 0.2541769191622734,
      "num_tokens": 92227465.0,
      "step": 40225
    },
    {
      "entropy": 5.080897474288941,
      "epoch": 3.8645533141210375,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035747854757884744,
      "loss": 4.6287,
      "mean_token_accuracy": 0.2565134271979332,
      "num_tokens": 92238304.0,
      "step": 40230
    },
    {
      "entropy": 5.018829441070556,
      "epoch": 3.8650336215177714,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035744664954592283,
      "loss": 4.5762,
      "mean_token_accuracy": 0.26033739000558853,
      "num_tokens": 92248851.0,
      "step": 40235
    },
    {
      "entropy": 5.015884351730347,
      "epoch": 3.8655139289145053,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035741474959883175,
      "loss": 4.5503,
      "mean_token_accuracy": 0.25881310254335405,
      "num_tokens": 92261110.0,
      "step": 40240
    },
    {
      "entropy": 5.126636552810669,
      "epoch": 3.865994236311239,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000357382847738315,
      "loss": 4.661,
      "mean_token_accuracy": 0.2496726006269455,
      "num_tokens": 92272896.0,
      "step": 40245
    },
    {
      "entropy": 5.061757946014405,
      "epoch": 3.866474543707973,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035735094396511304,
      "loss": 4.6442,
      "mean_token_accuracy": 0.2589659318327904,
      "num_tokens": 92284564.0,
      "step": 40250
    },
    {
      "entropy": 5.068583393096924,
      "epoch": 3.866954851104707,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003573190382799667,
      "loss": 4.5758,
      "mean_token_accuracy": 0.26264801770448687,
      "num_tokens": 92294872.0,
      "step": 40255
    },
    {
      "entropy": 5.114331674575806,
      "epoch": 3.867435158501441,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035728713068361684,
      "loss": 4.6707,
      "mean_token_accuracy": 0.2529986709356308,
      "num_tokens": 92307291.0,
      "step": 40260
    },
    {
      "entropy": 5.076523113250732,
      "epoch": 3.8679154658981747,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035725522117680404,
      "loss": 4.6537,
      "mean_token_accuracy": 0.24759605079889296,
      "num_tokens": 92317903.0,
      "step": 40265
    },
    {
      "entropy": 5.143259334564209,
      "epoch": 3.8683957732949086,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003572233097602693,
      "loss": 4.6298,
      "mean_token_accuracy": 0.24862954914569854,
      "num_tokens": 92329664.0,
      "step": 40270
    },
    {
      "entropy": 5.054178476333618,
      "epoch": 3.868876080691643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035719139643475344,
      "loss": 4.5825,
      "mean_token_accuracy": 0.2629439800977707,
      "num_tokens": 92342709.0,
      "step": 40275
    },
    {
      "entropy": 5.097134923934936,
      "epoch": 3.8693563880883763,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035715948120099747,
      "loss": 4.6688,
      "mean_token_accuracy": 0.2513518050312996,
      "num_tokens": 92354254.0,
      "step": 40280
    },
    {
      "entropy": 5.084424877166748,
      "epoch": 3.8698366954851107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035712756405974216,
      "loss": 4.5783,
      "mean_token_accuracy": 0.25494205206632614,
      "num_tokens": 92365596.0,
      "step": 40285
    },
    {
      "entropy": 5.020837736129761,
      "epoch": 3.870317002881844,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003570956450117288,
      "loss": 4.5321,
      "mean_token_accuracy": 0.26179041862487795,
      "num_tokens": 92377789.0,
      "step": 40290
    },
    {
      "entropy": 5.050160503387451,
      "epoch": 3.8707973102785784,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003570637240576983,
      "loss": 4.6051,
      "mean_token_accuracy": 0.25544093251228334,
      "num_tokens": 92388310.0,
      "step": 40295
    },
    {
      "entropy": 4.975748538970947,
      "epoch": 3.8712776176753123,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035703180119839187,
      "loss": 4.5765,
      "mean_token_accuracy": 0.2578082919120789,
      "num_tokens": 92401378.0,
      "step": 40300
    },
    {
      "entropy": 5.105275583267212,
      "epoch": 3.871757925072046,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003569998764345506,
      "loss": 4.6155,
      "mean_token_accuracy": 0.2526064649224281,
      "num_tokens": 92411965.0,
      "step": 40305
    },
    {
      "entropy": 5.0094481945037845,
      "epoch": 3.87223823246878,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035696794976691564,
      "loss": 4.5703,
      "mean_token_accuracy": 0.262768542766571,
      "num_tokens": 92424265.0,
      "step": 40310
    },
    {
      "entropy": 5.1281468868255615,
      "epoch": 3.872718539865514,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035693602119622823,
      "loss": 4.6735,
      "mean_token_accuracy": 0.2578767165541649,
      "num_tokens": 92435938.0,
      "step": 40315
    },
    {
      "entropy": 5.072874307632446,
      "epoch": 3.873198847262248,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003569040907232298,
      "loss": 4.5526,
      "mean_token_accuracy": 0.2604093849658966,
      "num_tokens": 92447704.0,
      "step": 40320
    },
    {
      "entropy": 5.084598875045776,
      "epoch": 3.8736791546589817,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00035687215834866154,
      "loss": 4.5861,
      "mean_token_accuracy": 0.2576905727386475,
      "num_tokens": 92458681.0,
      "step": 40325
    },
    {
      "entropy": 5.062186574935913,
      "epoch": 3.8741594620557156,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035684022407326486,
      "loss": 4.5646,
      "mean_token_accuracy": 0.2534119561314583,
      "num_tokens": 92470423.0,
      "step": 40330
    },
    {
      "entropy": 5.088231468200684,
      "epoch": 3.8746397694524495,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035680828789778126,
      "loss": 4.65,
      "mean_token_accuracy": 0.2536809206008911,
      "num_tokens": 92481657.0,
      "step": 40335
    },
    {
      "entropy": 5.096616268157959,
      "epoch": 3.8751200768491834,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003567763498229521,
      "loss": 4.7322,
      "mean_token_accuracy": 0.25026182532310487,
      "num_tokens": 92494140.0,
      "step": 40340
    },
    {
      "entropy": 5.161847162246704,
      "epoch": 3.8756003842459172,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003567444098495189,
      "loss": 4.7229,
      "mean_token_accuracy": 0.2506111592054367,
      "num_tokens": 92506466.0,
      "step": 40345
    },
    {
      "entropy": 5.0817255020141605,
      "epoch": 3.8760806916426516,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035671246797822325,
      "loss": 4.6285,
      "mean_token_accuracy": 0.2585867241024971,
      "num_tokens": 92518619.0,
      "step": 40350
    },
    {
      "entropy": 5.1310343742370605,
      "epoch": 3.876560999039385,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003566805242098067,
      "loss": 4.7225,
      "mean_token_accuracy": 0.24233328551054,
      "num_tokens": 92530406.0,
      "step": 40355
    },
    {
      "entropy": 5.068937587738037,
      "epoch": 3.8770413064361193,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000356648578545011,
      "loss": 4.576,
      "mean_token_accuracy": 0.2548491835594177,
      "num_tokens": 92543010.0,
      "step": 40360
    },
    {
      "entropy": 4.984466791152954,
      "epoch": 3.8775216138328528,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003566166309845777,
      "loss": 4.5049,
      "mean_token_accuracy": 0.2629722625017166,
      "num_tokens": 92555672.0,
      "step": 40365
    },
    {
      "entropy": 5.052659320831299,
      "epoch": 3.878001921229587,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003565846815292486,
      "loss": 4.6556,
      "mean_token_accuracy": 0.2529388159513474,
      "num_tokens": 92567916.0,
      "step": 40370
    },
    {
      "entropy": 5.121433687210083,
      "epoch": 3.878482228626321,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003565527301797654,
      "loss": 4.7022,
      "mean_token_accuracy": 0.24571410417556763,
      "num_tokens": 92579048.0,
      "step": 40375
    },
    {
      "entropy": 5.113930320739746,
      "epoch": 3.878962536023055,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035652077693687,
      "loss": 4.6582,
      "mean_token_accuracy": 0.25884878635406494,
      "num_tokens": 92590283.0,
      "step": 40380
    },
    {
      "entropy": 5.13789849281311,
      "epoch": 3.8794428434197887,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003564888218013042,
      "loss": 4.74,
      "mean_token_accuracy": 0.24217734932899476,
      "num_tokens": 92602468.0,
      "step": 40385
    },
    {
      "entropy": 5.1647967338562015,
      "epoch": 3.8799231508165226,
      "grad_norm": 2.484375,
      "learning_rate": 0.00035645686477380996,
      "loss": 4.5714,
      "mean_token_accuracy": 0.2643877357244492,
      "num_tokens": 92614279.0,
      "step": 40390
    },
    {
      "entropy": 5.071643304824829,
      "epoch": 3.8804034582132565,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003564249058551293,
      "loss": 4.5648,
      "mean_token_accuracy": 0.2622404396533966,
      "num_tokens": 92626573.0,
      "step": 40395
    },
    {
      "entropy": 5.021797323226929,
      "epoch": 3.8808837656099904,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003563929450460039,
      "loss": 4.5454,
      "mean_token_accuracy": 0.26278137415647507,
      "num_tokens": 92637539.0,
      "step": 40400
    },
    {
      "entropy": 4.986843919754028,
      "epoch": 3.8813640730067243,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003563609823471761,
      "loss": 4.5691,
      "mean_token_accuracy": 0.26554004997015,
      "num_tokens": 92648538.0,
      "step": 40405
    },
    {
      "entropy": 5.074393367767334,
      "epoch": 3.881844380403458,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035632901775938794,
      "loss": 4.674,
      "mean_token_accuracy": 0.24258375614881517,
      "num_tokens": 92659815.0,
      "step": 40410
    },
    {
      "entropy": 5.163948202133179,
      "epoch": 3.882324687800192,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003562970512833814,
      "loss": 4.7166,
      "mean_token_accuracy": 0.23760873973369598,
      "num_tokens": 92671155.0,
      "step": 40415
    },
    {
      "entropy": 5.13849368095398,
      "epoch": 3.882804995196926,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003562650829198988,
      "loss": 4.6297,
      "mean_token_accuracy": 0.2531843736767769,
      "num_tokens": 92683397.0,
      "step": 40420
    },
    {
      "entropy": 5.071647262573242,
      "epoch": 3.88328530259366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035623311266968225,
      "loss": 4.692,
      "mean_token_accuracy": 0.25114868581295013,
      "num_tokens": 92696342.0,
      "step": 40425
    },
    {
      "entropy": 5.109956550598144,
      "epoch": 3.8837656099903937,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000356201140533474,
      "loss": 4.6005,
      "mean_token_accuracy": 0.2555091857910156,
      "num_tokens": 92707935.0,
      "step": 40430
    },
    {
      "entropy": 5.119651126861572,
      "epoch": 3.884245917387128,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003561691665120164,
      "loss": 4.6716,
      "mean_token_accuracy": 0.2514778509736061,
      "num_tokens": 92719112.0,
      "step": 40435
    },
    {
      "entropy": 5.089478015899658,
      "epoch": 3.8847262247838614,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003561371906060518,
      "loss": 4.598,
      "mean_token_accuracy": 0.2600146159529686,
      "num_tokens": 92730025.0,
      "step": 40440
    },
    {
      "entropy": 4.982234954833984,
      "epoch": 3.8852065321805958,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035610521281632257,
      "loss": 4.5637,
      "mean_token_accuracy": 0.26201850920915604,
      "num_tokens": 92741242.0,
      "step": 40445
    },
    {
      "entropy": 5.003654956817627,
      "epoch": 3.8856868395773296,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003560732331435711,
      "loss": 4.5646,
      "mean_token_accuracy": 0.2558830365538597,
      "num_tokens": 92752246.0,
      "step": 40450
    },
    {
      "entropy": 5.05890121459961,
      "epoch": 3.8861671469740635,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035604125158853996,
      "loss": 4.5208,
      "mean_token_accuracy": 0.2584674596786499,
      "num_tokens": 92763592.0,
      "step": 40455
    },
    {
      "entropy": 5.089093160629273,
      "epoch": 3.8866474543707974,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003560092681519715,
      "loss": 4.651,
      "mean_token_accuracy": 0.2528840497136116,
      "num_tokens": 92775379.0,
      "step": 40460
    },
    {
      "entropy": 5.0147946834564205,
      "epoch": 3.8871277617675313,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035597728283460843,
      "loss": 4.5561,
      "mean_token_accuracy": 0.25625451654195786,
      "num_tokens": 92786394.0,
      "step": 40465
    },
    {
      "entropy": 5.103244209289551,
      "epoch": 3.887608069164265,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035594529563719336,
      "loss": 4.6116,
      "mean_token_accuracy": 0.25290804356336594,
      "num_tokens": 92798447.0,
      "step": 40470
    },
    {
      "entropy": 5.082966613769531,
      "epoch": 3.888088376560999,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003559133065604688,
      "loss": 4.5802,
      "mean_token_accuracy": 0.26448908150196077,
      "num_tokens": 92810289.0,
      "step": 40475
    },
    {
      "entropy": 5.06098313331604,
      "epoch": 3.888568683957733,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003558813156051776,
      "loss": 4.6256,
      "mean_token_accuracy": 0.25831118375062945,
      "num_tokens": 92821483.0,
      "step": 40480
    },
    {
      "entropy": 4.977693843841553,
      "epoch": 3.889048991354467,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003558493227720624,
      "loss": 4.4953,
      "mean_token_accuracy": 0.25690812319517137,
      "num_tokens": 92832470.0,
      "step": 40485
    },
    {
      "entropy": 5.020063018798828,
      "epoch": 3.8895292987512007,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035581732806186595,
      "loss": 4.5418,
      "mean_token_accuracy": 0.26639296263456347,
      "num_tokens": 92843663.0,
      "step": 40490
    },
    {
      "entropy": 5.002882957458496,
      "epoch": 3.8900096061479346,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003557853314753311,
      "loss": 4.536,
      "mean_token_accuracy": 0.2612639844417572,
      "num_tokens": 92856070.0,
      "step": 40495
    },
    {
      "entropy": 5.139699935913086,
      "epoch": 3.8904899135446684,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035575333301320083,
      "loss": 4.6243,
      "mean_token_accuracy": 0.2565339237451553,
      "num_tokens": 92866499.0,
      "step": 40500
    },
    {
      "entropy": 5.068917989730835,
      "epoch": 3.8909702209414023,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035572133267621793,
      "loss": 4.5819,
      "mean_token_accuracy": 0.2569202482700348,
      "num_tokens": 92878621.0,
      "step": 40505
    },
    {
      "entropy": 5.07551589012146,
      "epoch": 3.8914505283381366,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003556893304651253,
      "loss": 4.6508,
      "mean_token_accuracy": 0.2517069086432457,
      "num_tokens": 92889781.0,
      "step": 40510
    },
    {
      "entropy": 5.1259129524230955,
      "epoch": 3.89193083573487,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003556573263806661,
      "loss": 4.6637,
      "mean_token_accuracy": 0.25380160212516784,
      "num_tokens": 92900296.0,
      "step": 40515
    },
    {
      "entropy": 4.993126058578492,
      "epoch": 3.8924111431316044,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003556253204235833,
      "loss": 4.5252,
      "mean_token_accuracy": 0.261714830994606,
      "num_tokens": 92912363.0,
      "step": 40520
    },
    {
      "entropy": 5.080163431167603,
      "epoch": 3.8928914505283383,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003555933125946199,
      "loss": 4.6449,
      "mean_token_accuracy": 0.26454389840364456,
      "num_tokens": 92922844.0,
      "step": 40525
    },
    {
      "entropy": 5.094967222213745,
      "epoch": 3.893371757925072,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003555613028945191,
      "loss": 4.6062,
      "mean_token_accuracy": 0.2547874704003334,
      "num_tokens": 92936210.0,
      "step": 40530
    },
    {
      "entropy": 5.059420537948609,
      "epoch": 3.893852065321806,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035552929132402414,
      "loss": 4.5437,
      "mean_token_accuracy": 0.26237587034702303,
      "num_tokens": 92948193.0,
      "step": 40535
    },
    {
      "entropy": 5.0113109111785885,
      "epoch": 3.89433237271854,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035549727788387805,
      "loss": 4.5568,
      "mean_token_accuracy": 0.25730744153261187,
      "num_tokens": 92961644.0,
      "step": 40540
    },
    {
      "entropy": 5.053181838989258,
      "epoch": 3.894812680115274,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035546526257482426,
      "loss": 4.595,
      "mean_token_accuracy": 0.25746293365955353,
      "num_tokens": 92971905.0,
      "step": 40545
    },
    {
      "entropy": 5.084030389785767,
      "epoch": 3.8952929875120077,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035543324539760604,
      "loss": 4.5637,
      "mean_token_accuracy": 0.2602913022041321,
      "num_tokens": 92982971.0,
      "step": 40550
    },
    {
      "entropy": 5.020782423019409,
      "epoch": 3.8957732949087416,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003554012263529666,
      "loss": 4.5626,
      "mean_token_accuracy": 0.2600758820772171,
      "num_tokens": 92993737.0,
      "step": 40555
    },
    {
      "entropy": 5.00920295715332,
      "epoch": 3.8962536023054755,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003553692054416495,
      "loss": 4.587,
      "mean_token_accuracy": 0.2603698492050171,
      "num_tokens": 93006063.0,
      "step": 40560
    },
    {
      "entropy": 5.016193580627442,
      "epoch": 3.8967339097022093,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003553371826643981,
      "loss": 4.5644,
      "mean_token_accuracy": 0.2687942057847977,
      "num_tokens": 93016790.0,
      "step": 40565
    },
    {
      "entropy": 5.080636978149414,
      "epoch": 3.8972142170989432,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003553051580219558,
      "loss": 4.7037,
      "mean_token_accuracy": 0.25370761901140215,
      "num_tokens": 93028275.0,
      "step": 40570
    },
    {
      "entropy": 5.119170904159546,
      "epoch": 3.897694524495677,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003552731315150662,
      "loss": 4.6341,
      "mean_token_accuracy": 0.2540642336010933,
      "num_tokens": 93039163.0,
      "step": 40575
    },
    {
      "entropy": 5.118839311599731,
      "epoch": 3.898174831892411,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035524110314447295,
      "loss": 4.6231,
      "mean_token_accuracy": 0.25485135316848756,
      "num_tokens": 93049867.0,
      "step": 40580
    },
    {
      "entropy": 5.165014171600342,
      "epoch": 3.8986551392891453,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003552090729109195,
      "loss": 4.707,
      "mean_token_accuracy": 0.2515792563557625,
      "num_tokens": 93061830.0,
      "step": 40585
    },
    {
      "entropy": 5.124035596847534,
      "epoch": 3.8991354466858787,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003551770408151496,
      "loss": 4.7247,
      "mean_token_accuracy": 0.24592090100049974,
      "num_tokens": 93073144.0,
      "step": 40590
    },
    {
      "entropy": 5.109860563278199,
      "epoch": 3.899615754082613,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035514500685790673,
      "loss": 4.6692,
      "mean_token_accuracy": 0.253959484398365,
      "num_tokens": 93083568.0,
      "step": 40595
    },
    {
      "entropy": 5.074950742721557,
      "epoch": 3.9000960614793465,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035511297103993486,
      "loss": 4.5955,
      "mean_token_accuracy": 0.2589622259140015,
      "num_tokens": 93094632.0,
      "step": 40600
    },
    {
      "entropy": 5.083727216720581,
      "epoch": 3.900576368876081,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003550809333619778,
      "loss": 4.5896,
      "mean_token_accuracy": 0.25701106041669847,
      "num_tokens": 93105759.0,
      "step": 40605
    },
    {
      "entropy": 5.091693449020386,
      "epoch": 3.9010566762728147,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003550488938247791,
      "loss": 4.6106,
      "mean_token_accuracy": 0.2591106057167053,
      "num_tokens": 93117101.0,
      "step": 40610
    },
    {
      "entropy": 5.022336959838867,
      "epoch": 3.9015369836695486,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003550168524290828,
      "loss": 4.586,
      "mean_token_accuracy": 0.2632342353463173,
      "num_tokens": 93127762.0,
      "step": 40615
    },
    {
      "entropy": 5.035040855407715,
      "epoch": 3.9020172910662825,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035498480917563294,
      "loss": 4.5313,
      "mean_token_accuracy": 0.2575222671031952,
      "num_tokens": 93139318.0,
      "step": 40620
    },
    {
      "entropy": 5.066061878204346,
      "epoch": 3.9024975984630164,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003549527640651732,
      "loss": 4.661,
      "mean_token_accuracy": 0.2527304172515869,
      "num_tokens": 93151544.0,
      "step": 40625
    },
    {
      "entropy": 5.079059791564942,
      "epoch": 3.9029779058597502,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003549207170984476,
      "loss": 4.6027,
      "mean_token_accuracy": 0.261785951256752,
      "num_tokens": 93162854.0,
      "step": 40630
    },
    {
      "entropy": 5.127342891693115,
      "epoch": 3.903458213256484,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035488866827620047,
      "loss": 4.724,
      "mean_token_accuracy": 0.2480955168604851,
      "num_tokens": 93174300.0,
      "step": 40635
    },
    {
      "entropy": 5.127477693557739,
      "epoch": 3.903938520653218,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003548566175991756,
      "loss": 4.6968,
      "mean_token_accuracy": 0.24737366586923598,
      "num_tokens": 93187284.0,
      "step": 40640
    },
    {
      "entropy": 5.012441873550415,
      "epoch": 3.904418828049952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035482456506811714,
      "loss": 4.5255,
      "mean_token_accuracy": 0.26372127830982206,
      "num_tokens": 93197840.0,
      "step": 40645
    },
    {
      "entropy": 5.090718221664429,
      "epoch": 3.9048991354466858,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035479251068376935,
      "loss": 4.6469,
      "mean_token_accuracy": 0.25379118621349334,
      "num_tokens": 93209605.0,
      "step": 40650
    },
    {
      "entropy": 5.096945190429688,
      "epoch": 3.9053794428434196,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035476045444687637,
      "loss": 4.6406,
      "mean_token_accuracy": 0.24774326980113984,
      "num_tokens": 93222125.0,
      "step": 40655
    },
    {
      "entropy": 5.032750225067138,
      "epoch": 3.905859750240154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003547283963581825,
      "loss": 4.5531,
      "mean_token_accuracy": 0.26204020380973814,
      "num_tokens": 93233487.0,
      "step": 40660
    },
    {
      "entropy": 5.023817539215088,
      "epoch": 3.9063400576368874,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000354696336418432,
      "loss": 4.5365,
      "mean_token_accuracy": 0.2597656726837158,
      "num_tokens": 93245358.0,
      "step": 40665
    },
    {
      "entropy": 5.085025882720947,
      "epoch": 3.9068203650336217,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035466427462836923,
      "loss": 4.6497,
      "mean_token_accuracy": 0.25930540859699247,
      "num_tokens": 93255902.0,
      "step": 40670
    },
    {
      "entropy": 5.059441566467285,
      "epoch": 3.907300672430355,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035463221098873854,
      "loss": 4.5873,
      "mean_token_accuracy": 0.26398222148418427,
      "num_tokens": 93267827.0,
      "step": 40675
    },
    {
      "entropy": 5.149115657806396,
      "epoch": 3.9077809798270895,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003546001455002844,
      "loss": 4.6614,
      "mean_token_accuracy": 0.24916008561849595,
      "num_tokens": 93280802.0,
      "step": 40680
    },
    {
      "entropy": 5.047484397888184,
      "epoch": 3.9082612872238234,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003545680781637512,
      "loss": 4.5576,
      "mean_token_accuracy": 0.2545773878693581,
      "num_tokens": 93292443.0,
      "step": 40685
    },
    {
      "entropy": 4.997695350646973,
      "epoch": 3.9087415946205573,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003545360089798836,
      "loss": 4.5416,
      "mean_token_accuracy": 0.26175117641687395,
      "num_tokens": 93304286.0,
      "step": 40690
    },
    {
      "entropy": 5.064384269714355,
      "epoch": 3.909221902017291,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003545039379494259,
      "loss": 4.5767,
      "mean_token_accuracy": 0.25795155167579653,
      "num_tokens": 93316211.0,
      "step": 40695
    },
    {
      "entropy": 5.069133138656616,
      "epoch": 3.909702209414025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000354471865073123,
      "loss": 4.6509,
      "mean_token_accuracy": 0.25377356857061384,
      "num_tokens": 93326472.0,
      "step": 40700
    },
    {
      "entropy": 5.068083333969116,
      "epoch": 3.910182516810759,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003544397903517192,
      "loss": 4.6802,
      "mean_token_accuracy": 0.24635239839553832,
      "num_tokens": 93337309.0,
      "step": 40705
    },
    {
      "entropy": 5.066747093200684,
      "epoch": 3.910662824207493,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003544077137859595,
      "loss": 4.635,
      "mean_token_accuracy": 0.25425519347190856,
      "num_tokens": 93349035.0,
      "step": 40710
    },
    {
      "entropy": 5.084029626846314,
      "epoch": 3.9111431316042267,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035437563537658835,
      "loss": 4.5906,
      "mean_token_accuracy": 0.26169452369213103,
      "num_tokens": 93361379.0,
      "step": 40715
    },
    {
      "entropy": 5.064356946945191,
      "epoch": 3.9116234390009605,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035434355512435075,
      "loss": 4.5584,
      "mean_token_accuracy": 0.2561794579029083,
      "num_tokens": 93371959.0,
      "step": 40720
    },
    {
      "entropy": 5.121482944488525,
      "epoch": 3.9121037463976944,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003543114730299913,
      "loss": 4.7404,
      "mean_token_accuracy": 0.2465062752366066,
      "num_tokens": 93384358.0,
      "step": 40725
    },
    {
      "entropy": 5.023046541213989,
      "epoch": 3.9125840537944283,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003542793890942551,
      "loss": 4.5388,
      "mean_token_accuracy": 0.2614253595471382,
      "num_tokens": 93395978.0,
      "step": 40730
    },
    {
      "entropy": 5.093103885650635,
      "epoch": 3.913064361191162,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003542473033178868,
      "loss": 4.6379,
      "mean_token_accuracy": 0.2658459573984146,
      "num_tokens": 93407268.0,
      "step": 40735
    },
    {
      "entropy": 5.084386110305786,
      "epoch": 3.913544668587896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003542152157016315,
      "loss": 4.5443,
      "mean_token_accuracy": 0.2598053365945816,
      "num_tokens": 93419559.0,
      "step": 40740
    },
    {
      "entropy": 5.151143360137939,
      "epoch": 3.9140249759846304,
      "grad_norm": 0.96875,
      "learning_rate": 0.000354183126246234,
      "loss": 4.6942,
      "mean_token_accuracy": 0.24472133368253707,
      "num_tokens": 93430873.0,
      "step": 40745
    },
    {
      "entropy": 5.101135158538819,
      "epoch": 3.914505283381364,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003541510349524395,
      "loss": 4.6629,
      "mean_token_accuracy": 0.25061183273792265,
      "num_tokens": 93442531.0,
      "step": 40750
    },
    {
      "entropy": 5.107471656799317,
      "epoch": 3.914985590778098,
      "grad_norm": 1.03125,
      "learning_rate": 0.000354118941820993,
      "loss": 4.6379,
      "mean_token_accuracy": 0.2519704610109329,
      "num_tokens": 93453901.0,
      "step": 40755
    },
    {
      "entropy": 5.106260824203491,
      "epoch": 3.915465898174832,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035408684685263954,
      "loss": 4.6416,
      "mean_token_accuracy": 0.25295367985963824,
      "num_tokens": 93466011.0,
      "step": 40760
    },
    {
      "entropy": 5.081088304519653,
      "epoch": 3.915946205571566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003540547500481244,
      "loss": 4.6572,
      "mean_token_accuracy": 0.24936035126447678,
      "num_tokens": 93476864.0,
      "step": 40765
    },
    {
      "entropy": 5.09660005569458,
      "epoch": 3.9164265129683,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003540226514081927,
      "loss": 4.6529,
      "mean_token_accuracy": 0.2567755326628685,
      "num_tokens": 93488933.0,
      "step": 40770
    },
    {
      "entropy": 5.123719358444214,
      "epoch": 3.9169068203650337,
      "grad_norm": 1.0,
      "learning_rate": 0.00035399055093358965,
      "loss": 4.6082,
      "mean_token_accuracy": 0.25368843227624893,
      "num_tokens": 93500738.0,
      "step": 40775
    },
    {
      "entropy": 5.055276584625244,
      "epoch": 3.9173871277617676,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035395844862506063,
      "loss": 4.5715,
      "mean_token_accuracy": 0.26078307032585146,
      "num_tokens": 93511522.0,
      "step": 40780
    },
    {
      "entropy": 5.086653089523315,
      "epoch": 3.9178674351585014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003539263444833509,
      "loss": 4.7024,
      "mean_token_accuracy": 0.2480623811483383,
      "num_tokens": 93522276.0,
      "step": 40785
    },
    {
      "entropy": 5.182757186889648,
      "epoch": 3.9183477425552353,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035389423850920576,
      "loss": 4.7141,
      "mean_token_accuracy": 0.24141571074724197,
      "num_tokens": 93533338.0,
      "step": 40790
    },
    {
      "entropy": 5.072907543182373,
      "epoch": 3.918828049951969,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035386213070337063,
      "loss": 4.5763,
      "mean_token_accuracy": 0.2590348184108734,
      "num_tokens": 93544151.0,
      "step": 40795
    },
    {
      "entropy": 5.037469863891602,
      "epoch": 3.919308357348703,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035383002106659114,
      "loss": 4.6013,
      "mean_token_accuracy": 0.26094929575920106,
      "num_tokens": 93554838.0,
      "step": 40800
    },
    {
      "entropy": 5.088932609558105,
      "epoch": 3.919788664745437,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035379790959961247,
      "loss": 4.6499,
      "mean_token_accuracy": 0.25361160188913345,
      "num_tokens": 93566058.0,
      "step": 40805
    },
    {
      "entropy": 5.0163496971130375,
      "epoch": 3.920268972142171,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003537657963031804,
      "loss": 4.5481,
      "mean_token_accuracy": 0.26223032027482984,
      "num_tokens": 93577763.0,
      "step": 40810
    },
    {
      "entropy": 5.012160110473633,
      "epoch": 3.9207492795389047,
      "grad_norm": 1.0,
      "learning_rate": 0.00035373368117804047,
      "loss": 4.5015,
      "mean_token_accuracy": 0.26639420092105864,
      "num_tokens": 93587793.0,
      "step": 40815
    },
    {
      "entropy": 5.038145685195923,
      "epoch": 3.921229586935639,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035370156422493814,
      "loss": 4.6501,
      "mean_token_accuracy": 0.24981000274419785,
      "num_tokens": 93599916.0,
      "step": 40820
    },
    {
      "entropy": 5.042733430862427,
      "epoch": 3.9217098943323725,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035366944544461926,
      "loss": 4.5544,
      "mean_token_accuracy": 0.2664510071277618,
      "num_tokens": 93611438.0,
      "step": 40825
    },
    {
      "entropy": 5.134717845916748,
      "epoch": 3.922190201729107,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003536373248378295,
      "loss": 4.6078,
      "mean_token_accuracy": 0.26042094379663466,
      "num_tokens": 93622413.0,
      "step": 40830
    },
    {
      "entropy": 5.09718885421753,
      "epoch": 3.9226705091258407,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035360520240531446,
      "loss": 4.6123,
      "mean_token_accuracy": 0.2531854659318924,
      "num_tokens": 93633667.0,
      "step": 40835
    },
    {
      "entropy": 5.044416809082032,
      "epoch": 3.9231508165225746,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035357307814782007,
      "loss": 4.6085,
      "mean_token_accuracy": 0.25936378389596937,
      "num_tokens": 93644531.0,
      "step": 40840
    },
    {
      "entropy": 5.053901433944702,
      "epoch": 3.9236311239193085,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003535409520660922,
      "loss": 4.5666,
      "mean_token_accuracy": 0.25754003375768664,
      "num_tokens": 93654912.0,
      "step": 40845
    },
    {
      "entropy": 5.081637334823609,
      "epoch": 3.9241114313160423,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003535088241608765,
      "loss": 4.602,
      "mean_token_accuracy": 0.2522063434123993,
      "num_tokens": 93666103.0,
      "step": 40850
    },
    {
      "entropy": 5.133281564712524,
      "epoch": 3.9245917387127762,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003534766944329191,
      "loss": 4.7043,
      "mean_token_accuracy": 0.2515334501862526,
      "num_tokens": 93677448.0,
      "step": 40855
    },
    {
      "entropy": 5.0951653003692625,
      "epoch": 3.92507204610951,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035344456288296595,
      "loss": 4.6314,
      "mean_token_accuracy": 0.25407138615846636,
      "num_tokens": 93689468.0,
      "step": 40860
    },
    {
      "entropy": 5.0588303565979,
      "epoch": 3.925552353506244,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003534124295117629,
      "loss": 4.6501,
      "mean_token_accuracy": 0.25435151904821396,
      "num_tokens": 93702160.0,
      "step": 40865
    },
    {
      "entropy": 5.080163240432739,
      "epoch": 3.926032660902978,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003533802943200561,
      "loss": 4.5613,
      "mean_token_accuracy": 0.2634582996368408,
      "num_tokens": 93713244.0,
      "step": 40870
    },
    {
      "entropy": 5.1356954097747805,
      "epoch": 3.9265129682997117,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003533481573085916,
      "loss": 4.6607,
      "mean_token_accuracy": 0.24502820670604705,
      "num_tokens": 93724788.0,
      "step": 40875
    },
    {
      "entropy": 5.105747509002685,
      "epoch": 3.9269932756964456,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003533160184781155,
      "loss": 4.6027,
      "mean_token_accuracy": 0.2472915843129158,
      "num_tokens": 93736290.0,
      "step": 40880
    },
    {
      "entropy": 5.154068231582642,
      "epoch": 3.9274735830931795,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000353283877829374,
      "loss": 4.6892,
      "mean_token_accuracy": 0.25373566448688506,
      "num_tokens": 93748841.0,
      "step": 40885
    },
    {
      "entropy": 5.0155782222747805,
      "epoch": 3.9279538904899134,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003532517353631134,
      "loss": 4.5426,
      "mean_token_accuracy": 0.2576776295900345,
      "num_tokens": 93760072.0,
      "step": 40890
    },
    {
      "entropy": 5.085140609741211,
      "epoch": 3.9284341978866477,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003532195910800798,
      "loss": 4.6739,
      "mean_token_accuracy": 0.2549238383769989,
      "num_tokens": 93769700.0,
      "step": 40895
    },
    {
      "entropy": 5.159902858734131,
      "epoch": 3.928914505283381,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003531874449810195,
      "loss": 4.73,
      "mean_token_accuracy": 0.2427578628063202,
      "num_tokens": 93781959.0,
      "step": 40900
    },
    {
      "entropy": 5.155364274978638,
      "epoch": 3.9293948126801155,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035315529706667905,
      "loss": 4.6287,
      "mean_token_accuracy": 0.25301305651664735,
      "num_tokens": 93794379.0,
      "step": 40905
    },
    {
      "entropy": 5.0328058242797855,
      "epoch": 3.929875120076849,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003531231473378046,
      "loss": 4.5838,
      "mean_token_accuracy": 0.25933928191661837,
      "num_tokens": 93807664.0,
      "step": 40910
    },
    {
      "entropy": 5.05617036819458,
      "epoch": 3.9303554274735832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035309099579514255,
      "loss": 4.5993,
      "mean_token_accuracy": 0.2570549637079239,
      "num_tokens": 93817758.0,
      "step": 40915
    },
    {
      "entropy": 5.095176315307617,
      "epoch": 3.930835734870317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003530588424394396,
      "loss": 4.6588,
      "mean_token_accuracy": 0.24943972676992415,
      "num_tokens": 93829742.0,
      "step": 40920
    },
    {
      "entropy": 5.0871867656707765,
      "epoch": 3.931316042267051,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035302668727144206,
      "loss": 4.5751,
      "mean_token_accuracy": 0.2572619795799255,
      "num_tokens": 93840807.0,
      "step": 40925
    },
    {
      "entropy": 5.060864400863648,
      "epoch": 3.931796349663785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035299453029189647,
      "loss": 4.5776,
      "mean_token_accuracy": 0.2667940482497215,
      "num_tokens": 93851206.0,
      "step": 40930
    },
    {
      "entropy": 5.073400592803955,
      "epoch": 3.9322766570605188,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035296237150154964,
      "loss": 4.6075,
      "mean_token_accuracy": 0.2533204212784767,
      "num_tokens": 93862730.0,
      "step": 40935
    },
    {
      "entropy": 5.060735893249512,
      "epoch": 3.9327569644572526,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003529302109011479,
      "loss": 4.576,
      "mean_token_accuracy": 0.25618217289447787,
      "num_tokens": 93874917.0,
      "step": 40940
    },
    {
      "entropy": 5.107678651809692,
      "epoch": 3.9332372718539865,
      "grad_norm": 1.0,
      "learning_rate": 0.00035289804849143806,
      "loss": 4.5832,
      "mean_token_accuracy": 0.25925260186195376,
      "num_tokens": 93886694.0,
      "step": 40945
    },
    {
      "entropy": 5.029679918289185,
      "epoch": 3.9337175792507204,
      "grad_norm": 1.109375,
      "learning_rate": 0.000352865884273167,
      "loss": 4.5855,
      "mean_token_accuracy": 0.26062726229429245,
      "num_tokens": 93898176.0,
      "step": 40950
    },
    {
      "entropy": 5.050363826751709,
      "epoch": 3.9341978866474543,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003528337182470812,
      "loss": 4.6135,
      "mean_token_accuracy": 0.25678150802850724,
      "num_tokens": 93910493.0,
      "step": 40955
    },
    {
      "entropy": 5.105652761459351,
      "epoch": 3.934678194044188,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035280155041392757,
      "loss": 4.5951,
      "mean_token_accuracy": 0.26132183969020845,
      "num_tokens": 93921625.0,
      "step": 40960
    },
    {
      "entropy": 5.0239458084106445,
      "epoch": 3.935158501440922,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003527693807744531,
      "loss": 4.5679,
      "mean_token_accuracy": 0.25856156647205353,
      "num_tokens": 93932884.0,
      "step": 40965
    },
    {
      "entropy": 5.023355293273926,
      "epoch": 3.9356388088376564,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003527372093294044,
      "loss": 4.5289,
      "mean_token_accuracy": 0.2618435904383659,
      "num_tokens": 93943372.0,
      "step": 40970
    },
    {
      "entropy": 5.027084541320801,
      "epoch": 3.93611911623439,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003527050360795285,
      "loss": 4.5636,
      "mean_token_accuracy": 0.2580923542380333,
      "num_tokens": 93954432.0,
      "step": 40975
    },
    {
      "entropy": 5.100531578063965,
      "epoch": 3.936599423631124,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003526728610255725,
      "loss": 4.5988,
      "mean_token_accuracy": 0.2506100133061409,
      "num_tokens": 93965441.0,
      "step": 40980
    },
    {
      "entropy": 5.145212316513062,
      "epoch": 3.9370797310278576,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035264068416828326,
      "loss": 4.6528,
      "mean_token_accuracy": 0.2566868826746941,
      "num_tokens": 93976309.0,
      "step": 40985
    },
    {
      "entropy": 5.0250404357910154,
      "epoch": 3.937560038424592,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003526085055084079,
      "loss": 4.5744,
      "mean_token_accuracy": 0.24905963540077208,
      "num_tokens": 93989491.0,
      "step": 40990
    },
    {
      "entropy": 5.106499290466308,
      "epoch": 3.938040345821326,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003525763250466935,
      "loss": 4.6555,
      "mean_token_accuracy": 0.24998115003108978,
      "num_tokens": 94001577.0,
      "step": 40995
    },
    {
      "entropy": 5.076009082794189,
      "epoch": 3.9385206532180597,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003525441427838872,
      "loss": 4.592,
      "mean_token_accuracy": 0.255885474383831,
      "num_tokens": 94012176.0,
      "step": 41000
    },
    {
      "entropy": 4.993652248382569,
      "epoch": 3.9390009606147935,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003525119587207361,
      "loss": 4.6129,
      "mean_token_accuracy": 0.2550539806485176,
      "num_tokens": 94025147.0,
      "step": 41005
    },
    {
      "entropy": 5.084108877182007,
      "epoch": 3.9394812680115274,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003524797728579875,
      "loss": 4.6207,
      "mean_token_accuracy": 0.2602296218276024,
      "num_tokens": 94035802.0,
      "step": 41010
    },
    {
      "entropy": 5.152504205703735,
      "epoch": 3.9399615754082613,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003524475851963887,
      "loss": 4.7237,
      "mean_token_accuracy": 0.255805641412735,
      "num_tokens": 94046438.0,
      "step": 41015
    },
    {
      "entropy": 5.046103811264038,
      "epoch": 3.940441882804995,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035241539573668693,
      "loss": 4.5398,
      "mean_token_accuracy": 0.2663401022553444,
      "num_tokens": 94058794.0,
      "step": 41020
    },
    {
      "entropy": 5.059650421142578,
      "epoch": 3.940922190201729,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035238320447962947,
      "loss": 4.5881,
      "mean_token_accuracy": 0.2564546689391136,
      "num_tokens": 94070562.0,
      "step": 41025
    },
    {
      "entropy": 5.015465593338012,
      "epoch": 3.941402497598463,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003523510114259638,
      "loss": 4.5757,
      "mean_token_accuracy": 0.2654047966003418,
      "num_tokens": 94081493.0,
      "step": 41030
    },
    {
      "entropy": 5.003022480010986,
      "epoch": 3.941882804995197,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003523188165764374,
      "loss": 4.5379,
      "mean_token_accuracy": 0.26234046667814254,
      "num_tokens": 94091896.0,
      "step": 41035
    },
    {
      "entropy": 5.133801603317261,
      "epoch": 3.9423631123919307,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035228661993179757,
      "loss": 4.6768,
      "mean_token_accuracy": 0.2500840097665787,
      "num_tokens": 94103743.0,
      "step": 41040
    },
    {
      "entropy": 5.068476009368896,
      "epoch": 3.9428434197886646,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000352254421492792,
      "loss": 4.5671,
      "mean_token_accuracy": 0.2599003121256828,
      "num_tokens": 94113923.0,
      "step": 41045
    },
    {
      "entropy": 5.077379894256592,
      "epoch": 3.9433237271853985,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035222222126016814,
      "loss": 4.5255,
      "mean_token_accuracy": 0.25662740916013715,
      "num_tokens": 94125053.0,
      "step": 41050
    },
    {
      "entropy": 5.063521671295166,
      "epoch": 3.943804034582133,
      "grad_norm": 1.0,
      "learning_rate": 0.0003521900192346736,
      "loss": 4.5852,
      "mean_token_accuracy": 0.2524002268910408,
      "num_tokens": 94137745.0,
      "step": 41055
    },
    {
      "entropy": 4.986479806900024,
      "epoch": 3.9442843419788662,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000352157815417056,
      "loss": 4.5004,
      "mean_token_accuracy": 0.27520544826984406,
      "num_tokens": 94150604.0,
      "step": 41060
    },
    {
      "entropy": 5.072677564620972,
      "epoch": 3.9447646493756006,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035212560980806305,
      "loss": 4.5579,
      "mean_token_accuracy": 0.2595095098018646,
      "num_tokens": 94161055.0,
      "step": 41065
    },
    {
      "entropy": 5.112888669967651,
      "epoch": 3.9452449567723344,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003520934024084425,
      "loss": 4.6464,
      "mean_token_accuracy": 0.259583680331707,
      "num_tokens": 94171243.0,
      "step": 41070
    },
    {
      "entropy": 5.015179920196533,
      "epoch": 3.9457252641690683,
      "grad_norm": 1.0,
      "learning_rate": 0.000352061193218942,
      "loss": 4.5665,
      "mean_token_accuracy": 0.2521196812391281,
      "num_tokens": 94182826.0,
      "step": 41075
    },
    {
      "entropy": 5.035417985916138,
      "epoch": 3.946205571565802,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003520289822403095,
      "loss": 4.6579,
      "mean_token_accuracy": 0.2565024450421333,
      "num_tokens": 94195298.0,
      "step": 41080
    },
    {
      "entropy": 5.141892766952514,
      "epoch": 3.946685878962536,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035199676947329274,
      "loss": 4.6827,
      "mean_token_accuracy": 0.25469382256269457,
      "num_tokens": 94205993.0,
      "step": 41085
    },
    {
      "entropy": 5.0452343940734865,
      "epoch": 3.94716618635927,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035196455491863954,
      "loss": 4.6017,
      "mean_token_accuracy": 0.25371981859207154,
      "num_tokens": 94218241.0,
      "step": 41090
    },
    {
      "entropy": 5.186172389984131,
      "epoch": 3.947646493756004,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000351932338577098,
      "loss": 4.7416,
      "mean_token_accuracy": 0.24053913503885269,
      "num_tokens": 94230351.0,
      "step": 41095
    },
    {
      "entropy": 5.096252393722534,
      "epoch": 3.9481268011527377,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035190012044941605,
      "loss": 4.5953,
      "mean_token_accuracy": 0.2556596025824547,
      "num_tokens": 94241478.0,
      "step": 41100
    },
    {
      "entropy": 5.145805883407593,
      "epoch": 3.9486071085494716,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003518679005363415,
      "loss": 4.6585,
      "mean_token_accuracy": 0.2508416876196861,
      "num_tokens": 94254206.0,
      "step": 41105
    },
    {
      "entropy": 5.1305646896362305,
      "epoch": 3.9490874159462055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003518356788386226,
      "loss": 4.6799,
      "mean_token_accuracy": 0.24513860642910004,
      "num_tokens": 94264842.0,
      "step": 41110
    },
    {
      "entropy": 5.086434412002563,
      "epoch": 3.9495677233429394,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035180345535700745,
      "loss": 4.6055,
      "mean_token_accuracy": 0.25246614366769793,
      "num_tokens": 94277392.0,
      "step": 41115
    },
    {
      "entropy": 5.13049087524414,
      "epoch": 3.9500480307396733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003517712300922441,
      "loss": 4.6345,
      "mean_token_accuracy": 0.24871613085269928,
      "num_tokens": 94287581.0,
      "step": 41120
    },
    {
      "entropy": 5.145023488998413,
      "epoch": 3.950528338136407,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035173900304508074,
      "loss": 4.6755,
      "mean_token_accuracy": 0.24903128445148467,
      "num_tokens": 94299733.0,
      "step": 41125
    },
    {
      "entropy": 5.12058744430542,
      "epoch": 3.9510086455331415,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003517067742162656,
      "loss": 4.683,
      "mean_token_accuracy": 0.24742978066205978,
      "num_tokens": 94311306.0,
      "step": 41130
    },
    {
      "entropy": 5.03211989402771,
      "epoch": 3.951488952929875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000351674543606547,
      "loss": 4.6122,
      "mean_token_accuracy": 0.2521358415484428,
      "num_tokens": 94322998.0,
      "step": 41135
    },
    {
      "entropy": 5.109110021591187,
      "epoch": 3.9519692603266092,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003516423112166731,
      "loss": 4.5746,
      "mean_token_accuracy": 0.25169509947299956,
      "num_tokens": 94334059.0,
      "step": 41140
    },
    {
      "entropy": 5.1447210788726805,
      "epoch": 3.952449567723343,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035161007704739226,
      "loss": 4.6964,
      "mean_token_accuracy": 0.2496648445725441,
      "num_tokens": 94345288.0,
      "step": 41145
    },
    {
      "entropy": 5.085406637191772,
      "epoch": 3.952929875120077,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000351577841099453,
      "loss": 4.6182,
      "mean_token_accuracy": 0.25234234184026716,
      "num_tokens": 94356484.0,
      "step": 41150
    },
    {
      "entropy": 5.131757736206055,
      "epoch": 3.953410182516811,
      "grad_norm": 1.0,
      "learning_rate": 0.0003515456033736036,
      "loss": 4.7054,
      "mean_token_accuracy": 0.25121617168188093,
      "num_tokens": 94368312.0,
      "step": 41155
    },
    {
      "entropy": 5.097722434997559,
      "epoch": 3.9538904899135447,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035151336387059275,
      "loss": 4.5889,
      "mean_token_accuracy": 0.266692678630352,
      "num_tokens": 94379325.0,
      "step": 41160
    },
    {
      "entropy": 5.130922126770019,
      "epoch": 3.9543707973102786,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003514811225911686,
      "loss": 4.64,
      "mean_token_accuracy": 0.2437990203499794,
      "num_tokens": 94389714.0,
      "step": 41165
    },
    {
      "entropy": 5.009433937072754,
      "epoch": 3.9548511047070125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035144887953607995,
      "loss": 4.5946,
      "mean_token_accuracy": 0.2579566717147827,
      "num_tokens": 94401764.0,
      "step": 41170
    },
    {
      "entropy": 5.083831882476806,
      "epoch": 3.9553314121037464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003514166347060753,
      "loss": 4.6673,
      "mean_token_accuracy": 0.2503815248608589,
      "num_tokens": 94413318.0,
      "step": 41175
    },
    {
      "entropy": 5.11995644569397,
      "epoch": 3.9558117195004803,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003513843881019033,
      "loss": 4.6216,
      "mean_token_accuracy": 0.2593636468052864,
      "num_tokens": 94425004.0,
      "step": 41180
    },
    {
      "entropy": 5.089215230941773,
      "epoch": 3.956292026897214,
      "grad_norm": 1.0,
      "learning_rate": 0.00035135213972431267,
      "loss": 4.5934,
      "mean_token_accuracy": 0.2593241587281227,
      "num_tokens": 94436546.0,
      "step": 41185
    },
    {
      "entropy": 5.103637552261352,
      "epoch": 3.956772334293948,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035131988957405206,
      "loss": 4.6371,
      "mean_token_accuracy": 0.2552954494953156,
      "num_tokens": 94447763.0,
      "step": 41190
    },
    {
      "entropy": 5.051658058166504,
      "epoch": 3.957252641690682,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035128763765187026,
      "loss": 4.5905,
      "mean_token_accuracy": 0.2485773265361786,
      "num_tokens": 94458809.0,
      "step": 41195
    },
    {
      "entropy": 5.079994535446167,
      "epoch": 3.957732949087416,
      "grad_norm": 0.953125,
      "learning_rate": 0.000351255383958516,
      "loss": 4.6312,
      "mean_token_accuracy": 0.258722747862339,
      "num_tokens": 94470615.0,
      "step": 41200
    },
    {
      "entropy": 5.0791820049285885,
      "epoch": 3.95821325648415,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003512231284947382,
      "loss": 4.6111,
      "mean_token_accuracy": 0.25821816772222517,
      "num_tokens": 94481674.0,
      "step": 41205
    },
    {
      "entropy": 5.071080160140991,
      "epoch": 3.9586935638808836,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003511908712612857,
      "loss": 4.6545,
      "mean_token_accuracy": 0.2519149392843246,
      "num_tokens": 94493584.0,
      "step": 41210
    },
    {
      "entropy": 5.0546629428863525,
      "epoch": 3.959173871277618,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003511586122589073,
      "loss": 4.5797,
      "mean_token_accuracy": 0.2575056731700897,
      "num_tokens": 94504761.0,
      "step": 41215
    },
    {
      "entropy": 5.044349336624146,
      "epoch": 3.9596541786743513,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003511263514883522,
      "loss": 4.5897,
      "mean_token_accuracy": 0.26002790182828905,
      "num_tokens": 94515552.0,
      "step": 41220
    },
    {
      "entropy": 4.982811164855957,
      "epoch": 3.9601344860710856,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003510940889503692,
      "loss": 4.544,
      "mean_token_accuracy": 0.2589930683374405,
      "num_tokens": 94527149.0,
      "step": 41225
    },
    {
      "entropy": 5.061422729492188,
      "epoch": 3.9606147934678195,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035106182464570736,
      "loss": 4.6272,
      "mean_token_accuracy": 0.2534879371523857,
      "num_tokens": 94538388.0,
      "step": 41230
    },
    {
      "entropy": 5.214080333709717,
      "epoch": 3.9610951008645534,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003510295585751159,
      "loss": 4.7458,
      "mean_token_accuracy": 0.24469696134328842,
      "num_tokens": 94551333.0,
      "step": 41235
    },
    {
      "entropy": 5.0667914867401125,
      "epoch": 3.9615754082612873,
      "grad_norm": 1.125,
      "learning_rate": 0.00035099729073934376,
      "loss": 4.6541,
      "mean_token_accuracy": 0.25020589083433153,
      "num_tokens": 94562729.0,
      "step": 41240
    },
    {
      "entropy": 5.0769459247589115,
      "epoch": 3.962055715658021,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003509650211391403,
      "loss": 4.6011,
      "mean_token_accuracy": 0.2562673017382622,
      "num_tokens": 94573738.0,
      "step": 41245
    },
    {
      "entropy": 5.045137166976929,
      "epoch": 3.962536023054755,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035093274977525456,
      "loss": 4.6338,
      "mean_token_accuracy": 0.25302416235208514,
      "num_tokens": 94585308.0,
      "step": 41250
    },
    {
      "entropy": 5.0201092720031735,
      "epoch": 3.963016330451489,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003509004766484359,
      "loss": 4.558,
      "mean_token_accuracy": 0.25837168246507647,
      "num_tokens": 94595840.0,
      "step": 41255
    },
    {
      "entropy": 5.1076109409332275,
      "epoch": 3.963496637848223,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035086820175943344,
      "loss": 4.6229,
      "mean_token_accuracy": 0.24612971395254135,
      "num_tokens": 94607062.0,
      "step": 41260
    },
    {
      "entropy": 5.083880138397217,
      "epoch": 3.9639769452449567,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003508359251089967,
      "loss": 4.6326,
      "mean_token_accuracy": 0.26021112948656083,
      "num_tokens": 94618498.0,
      "step": 41265
    },
    {
      "entropy": 5.053261470794678,
      "epoch": 3.9644572526416906,
      "grad_norm": 0.875,
      "learning_rate": 0.00035080364669787495,
      "loss": 4.5817,
      "mean_token_accuracy": 0.25683428198099134,
      "num_tokens": 94630727.0,
      "step": 41270
    },
    {
      "entropy": 5.08069224357605,
      "epoch": 3.9649375600384245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035077136652681757,
      "loss": 4.5464,
      "mean_token_accuracy": 0.256886300444603,
      "num_tokens": 94642237.0,
      "step": 41275
    },
    {
      "entropy": 5.090951156616211,
      "epoch": 3.9654178674351583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035073908459657413,
      "loss": 4.671,
      "mean_token_accuracy": 0.24369477182626725,
      "num_tokens": 94652856.0,
      "step": 41280
    },
    {
      "entropy": 5.01417350769043,
      "epoch": 3.965898174831892,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035070680090789404,
      "loss": 4.5979,
      "mean_token_accuracy": 0.252669720351696,
      "num_tokens": 94664575.0,
      "step": 41285
    },
    {
      "entropy": 5.080107164382935,
      "epoch": 3.9663784822286265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003506745154615268,
      "loss": 4.6216,
      "mean_token_accuracy": 0.2582400843501091,
      "num_tokens": 94675619.0,
      "step": 41290
    },
    {
      "entropy": 5.077891731262207,
      "epoch": 3.96685878962536,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035064222825822197,
      "loss": 4.5584,
      "mean_token_accuracy": 0.2580797880887985,
      "num_tokens": 94687791.0,
      "step": 41295
    },
    {
      "entropy": 5.155251598358154,
      "epoch": 3.9673390970220943,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035060993929872923,
      "loss": 4.6616,
      "mean_token_accuracy": 0.24733202457427977,
      "num_tokens": 94698823.0,
      "step": 41300
    },
    {
      "entropy": 5.091981220245361,
      "epoch": 3.967819404418828,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003505776485837983,
      "loss": 4.6108,
      "mean_token_accuracy": 0.2558497428894043,
      "num_tokens": 94709451.0,
      "step": 41305
    },
    {
      "entropy": 5.052384281158448,
      "epoch": 3.968299711815562,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003505453561141787,
      "loss": 4.6187,
      "mean_token_accuracy": 0.25215230584144593,
      "num_tokens": 94720598.0,
      "step": 41310
    },
    {
      "entropy": 5.068251132965088,
      "epoch": 3.968780019212296,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003505130618906203,
      "loss": 4.5601,
      "mean_token_accuracy": 0.2702736184000969,
      "num_tokens": 94730329.0,
      "step": 41315
    },
    {
      "entropy": 5.079864835739135,
      "epoch": 3.96926032660903,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003504807659138728,
      "loss": 4.6057,
      "mean_token_accuracy": 0.2529628723859787,
      "num_tokens": 94742548.0,
      "step": 41320
    },
    {
      "entropy": 5.061963748931885,
      "epoch": 3.9697406340057637,
      "grad_norm": 1.0,
      "learning_rate": 0.0003504484681846861,
      "loss": 4.6116,
      "mean_token_accuracy": 0.25548112243413923,
      "num_tokens": 94753157.0,
      "step": 41325
    },
    {
      "entropy": 5.103848457336426,
      "epoch": 3.9702209414024976,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035041616870380997,
      "loss": 4.6491,
      "mean_token_accuracy": 0.25185677856206895,
      "num_tokens": 94765800.0,
      "step": 41330
    },
    {
      "entropy": 5.04579176902771,
      "epoch": 3.9707012487992315,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035038386747199437,
      "loss": 4.5569,
      "mean_token_accuracy": 0.25663802176713946,
      "num_tokens": 94776774.0,
      "step": 41335
    },
    {
      "entropy": 5.058940744400024,
      "epoch": 3.9711815561959654,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003503515644899892,
      "loss": 4.5712,
      "mean_token_accuracy": 0.26109920293092725,
      "num_tokens": 94787645.0,
      "step": 41340
    },
    {
      "entropy": 5.064389753341675,
      "epoch": 3.9716618635926992,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003503192597585444,
      "loss": 4.5482,
      "mean_token_accuracy": 0.2517098903656006,
      "num_tokens": 94798925.0,
      "step": 41345
    },
    {
      "entropy": 5.04087462425232,
      "epoch": 3.972142170989433,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003502869532784101,
      "loss": 4.5847,
      "mean_token_accuracy": 0.26403249204158785,
      "num_tokens": 94810399.0,
      "step": 41350
    },
    {
      "entropy": 5.065904712677002,
      "epoch": 3.972622478386167,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035025464505033626,
      "loss": 4.5815,
      "mean_token_accuracy": 0.2593311756849289,
      "num_tokens": 94821278.0,
      "step": 41355
    },
    {
      "entropy": 5.046607732772827,
      "epoch": 3.973102785782901,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000350222335075073,
      "loss": 4.5474,
      "mean_token_accuracy": 0.2612125352025032,
      "num_tokens": 94832585.0,
      "step": 41360
    },
    {
      "entropy": 5.1815454959869385,
      "epoch": 3.973583093179635,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035019002335337057,
      "loss": 4.7588,
      "mean_token_accuracy": 0.24615631401538848,
      "num_tokens": 94843237.0,
      "step": 41365
    },
    {
      "entropy": 4.997013998031616,
      "epoch": 3.9740634005763686,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003501577098859791,
      "loss": 4.5297,
      "mean_token_accuracy": 0.26441447585821154,
      "num_tokens": 94855300.0,
      "step": 41370
    },
    {
      "entropy": 5.011654996871949,
      "epoch": 3.974543707973103,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035012539467364865,
      "loss": 4.5897,
      "mean_token_accuracy": 0.26112214624881747,
      "num_tokens": 94867749.0,
      "step": 41375
    },
    {
      "entropy": 5.083395624160767,
      "epoch": 3.975024015369837,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003500930777171296,
      "loss": 4.6865,
      "mean_token_accuracy": 0.25132421106100084,
      "num_tokens": 94880631.0,
      "step": 41380
    },
    {
      "entropy": 5.118611907958984,
      "epoch": 3.9755043227665707,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003500607590171724,
      "loss": 4.6605,
      "mean_token_accuracy": 0.2525247484445572,
      "num_tokens": 94893506.0,
      "step": 41385
    },
    {
      "entropy": 5.047426462173462,
      "epoch": 3.9759846301633046,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003500284385745272,
      "loss": 4.5715,
      "mean_token_accuracy": 0.25847800374031066,
      "num_tokens": 94905128.0,
      "step": 41390
    },
    {
      "entropy": 5.120534658432007,
      "epoch": 3.9764649375600385,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003499961163899444,
      "loss": 4.6734,
      "mean_token_accuracy": 0.2578121840953827,
      "num_tokens": 94916725.0,
      "step": 41395
    },
    {
      "entropy": 5.100362777709961,
      "epoch": 3.9769452449567724,
      "grad_norm": 1.0,
      "learning_rate": 0.00034996379246417453,
      "loss": 4.6677,
      "mean_token_accuracy": 0.25026853382587433,
      "num_tokens": 94928486.0,
      "step": 41400
    },
    {
      "entropy": 5.113478565216065,
      "epoch": 3.9774255523535063,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034993146679796795,
      "loss": 4.6494,
      "mean_token_accuracy": 0.247309410572052,
      "num_tokens": 94940688.0,
      "step": 41405
    },
    {
      "entropy": 5.079276847839355,
      "epoch": 3.97790585975024,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003498991393920752,
      "loss": 4.581,
      "mean_token_accuracy": 0.25917048901319506,
      "num_tokens": 94951821.0,
      "step": 41410
    },
    {
      "entropy": 5.152861452102661,
      "epoch": 3.978386167146974,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034986681024724696,
      "loss": 4.7184,
      "mean_token_accuracy": 0.248225399851799,
      "num_tokens": 94963850.0,
      "step": 41415
    },
    {
      "entropy": 5.110031700134277,
      "epoch": 3.978866474543708,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00034983447936423357,
      "loss": 4.6265,
      "mean_token_accuracy": 0.25048366338014605,
      "num_tokens": 94975428.0,
      "step": 41420
    },
    {
      "entropy": 5.12569227218628,
      "epoch": 3.979346781940442,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003498021467437858,
      "loss": 4.6441,
      "mean_token_accuracy": 0.2591647759079933,
      "num_tokens": 94987314.0,
      "step": 41425
    },
    {
      "entropy": 5.098045206069946,
      "epoch": 3.9798270893371757,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003497698123866544,
      "loss": 4.5942,
      "mean_token_accuracy": 0.25820251256227494,
      "num_tokens": 94998852.0,
      "step": 41430
    },
    {
      "entropy": 5.171065855026245,
      "epoch": 3.9803073967339095,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034973747629358983,
      "loss": 4.6989,
      "mean_token_accuracy": 0.2523294657468796,
      "num_tokens": 95009901.0,
      "step": 41435
    },
    {
      "entropy": 5.140763568878174,
      "epoch": 3.980787704130644,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034970513846534305,
      "loss": 4.5744,
      "mean_token_accuracy": 0.2606427401304245,
      "num_tokens": 95020705.0,
      "step": 41440
    },
    {
      "entropy": 5.101607847213745,
      "epoch": 3.9812680115273773,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003496727989026648,
      "loss": 4.5732,
      "mean_token_accuracy": 0.25488118827342987,
      "num_tokens": 95032249.0,
      "step": 41445
    },
    {
      "entropy": 4.952258062362671,
      "epoch": 3.9817483189241116,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034964045760630584,
      "loss": 4.5159,
      "mean_token_accuracy": 0.25989279597997667,
      "num_tokens": 95044462.0,
      "step": 41450
    },
    {
      "entropy": 5.1300427436828615,
      "epoch": 3.982228626320845,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034960811457701715,
      "loss": 4.644,
      "mean_token_accuracy": 0.2529966622591019,
      "num_tokens": 95054335.0,
      "step": 41455
    },
    {
      "entropy": 5.101235485076904,
      "epoch": 3.9827089337175794,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034957576981554955,
      "loss": 4.6078,
      "mean_token_accuracy": 0.2503061830997467,
      "num_tokens": 95064809.0,
      "step": 41460
    },
    {
      "entropy": 5.042921638488769,
      "epoch": 3.9831892411143133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034954342332265407,
      "loss": 4.616,
      "mean_token_accuracy": 0.25179589688777926,
      "num_tokens": 95076167.0,
      "step": 41465
    },
    {
      "entropy": 5.194021606445313,
      "epoch": 3.983669548511047,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034951107509908157,
      "loss": 4.8083,
      "mean_token_accuracy": 0.24156892001628877,
      "num_tokens": 95087165.0,
      "step": 41470
    },
    {
      "entropy": 5.003713893890381,
      "epoch": 3.984149855907781,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003494787251455832,
      "loss": 4.4861,
      "mean_token_accuracy": 0.265624038875103,
      "num_tokens": 95098516.0,
      "step": 41475
    },
    {
      "entropy": 5.036776924133301,
      "epoch": 3.984630163304515,
      "grad_norm": 1.0,
      "learning_rate": 0.00034944637346291,
      "loss": 4.5476,
      "mean_token_accuracy": 0.25928640365600586,
      "num_tokens": 95110102.0,
      "step": 41480
    },
    {
      "entropy": 5.044084215164185,
      "epoch": 3.985110470701249,
      "grad_norm": 1.0625,
      "learning_rate": 0.000349414020051813,
      "loss": 4.5622,
      "mean_token_accuracy": 0.2619197010993958,
      "num_tokens": 95121029.0,
      "step": 41485
    },
    {
      "entropy": 5.138277769088745,
      "epoch": 3.9855907780979827,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003493816649130435,
      "loss": 4.6056,
      "mean_token_accuracy": 0.25867910087108614,
      "num_tokens": 95132564.0,
      "step": 41490
    },
    {
      "entropy": 5.165099668502807,
      "epoch": 3.9860710854947166,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003493493080473526,
      "loss": 4.6986,
      "mean_token_accuracy": 0.2517673909664154,
      "num_tokens": 95144595.0,
      "step": 41495
    },
    {
      "entropy": 5.146123933792114,
      "epoch": 3.9865513928914504,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034931694945549146,
      "loss": 4.669,
      "mean_token_accuracy": 0.25137288570404054,
      "num_tokens": 95155625.0,
      "step": 41500
    },
    {
      "entropy": 5.062015056610107,
      "epoch": 3.9870317002881843,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003492845891382115,
      "loss": 4.6624,
      "mean_token_accuracy": 0.24912654459476472,
      "num_tokens": 95166770.0,
      "step": 41505
    },
    {
      "entropy": 5.105222749710083,
      "epoch": 3.987512007684918,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000349252227096264,
      "loss": 4.6297,
      "mean_token_accuracy": 0.25725183039903643,
      "num_tokens": 95177853.0,
      "step": 41510
    },
    {
      "entropy": 5.041123342514038,
      "epoch": 3.9879923150816525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003492198633304002,
      "loss": 4.5212,
      "mean_token_accuracy": 0.26650853604078295,
      "num_tokens": 95189925.0,
      "step": 41515
    },
    {
      "entropy": 5.128666973114013,
      "epoch": 3.988472622478386,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003491874978413716,
      "loss": 4.6975,
      "mean_token_accuracy": 0.2415354445576668,
      "num_tokens": 95202670.0,
      "step": 41520
    },
    {
      "entropy": 5.066181182861328,
      "epoch": 3.9889529298751203,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003491551306299296,
      "loss": 4.5524,
      "mean_token_accuracy": 0.25217062383890154,
      "num_tokens": 95213739.0,
      "step": 41525
    },
    {
      "entropy": 5.1016778469085695,
      "epoch": 3.9894332372718537,
      "grad_norm": 0.921875,
      "learning_rate": 0.00034912276169682556,
      "loss": 4.661,
      "mean_token_accuracy": 0.2425233915448189,
      "num_tokens": 95225325.0,
      "step": 41530
    },
    {
      "entropy": 5.113547992706299,
      "epoch": 3.989913544668588,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034909039104281123,
      "loss": 4.6448,
      "mean_token_accuracy": 0.2544423758983612,
      "num_tokens": 95237177.0,
      "step": 41535
    },
    {
      "entropy": 5.105944585800171,
      "epoch": 3.990393852065322,
      "grad_norm": 1.125,
      "learning_rate": 0.0003490580186686379,
      "loss": 4.6226,
      "mean_token_accuracy": 0.2505847379565239,
      "num_tokens": 95248801.0,
      "step": 41540
    },
    {
      "entropy": 5.188764953613282,
      "epoch": 3.990874159462056,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003490256445750573,
      "loss": 4.719,
      "mean_token_accuracy": 0.2462164580821991,
      "num_tokens": 95261233.0,
      "step": 41545
    },
    {
      "entropy": 5.020900726318359,
      "epoch": 3.9913544668587897,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003489932687628211,
      "loss": 4.4921,
      "mean_token_accuracy": 0.25836329609155656,
      "num_tokens": 95270953.0,
      "step": 41550
    },
    {
      "entropy": 5.094780588150025,
      "epoch": 3.9918347742555236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034896089123268096,
      "loss": 4.6482,
      "mean_token_accuracy": 0.25105644166469576,
      "num_tokens": 95281076.0,
      "step": 41555
    },
    {
      "entropy": 5.073063373565674,
      "epoch": 3.9923150816522575,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003489285119853884,
      "loss": 4.6232,
      "mean_token_accuracy": 0.25233477503061297,
      "num_tokens": 95292953.0,
      "step": 41560
    },
    {
      "entropy": 5.026347875595093,
      "epoch": 3.9927953890489913,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003488961310216954,
      "loss": 4.5733,
      "mean_token_accuracy": 0.25977848321199415,
      "num_tokens": 95304875.0,
      "step": 41565
    },
    {
      "entropy": 5.118353080749512,
      "epoch": 3.993275696445725,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003488637483423537,
      "loss": 4.6709,
      "mean_token_accuracy": 0.2497504323720932,
      "num_tokens": 95316965.0,
      "step": 41570
    },
    {
      "entropy": 5.130131530761719,
      "epoch": 3.993756003842459,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034883136394811505,
      "loss": 4.6155,
      "mean_token_accuracy": 0.2520702019333839,
      "num_tokens": 95327580.0,
      "step": 41575
    },
    {
      "entropy": 5.078052091598511,
      "epoch": 3.994236311239193,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034879897783973143,
      "loss": 4.5747,
      "mean_token_accuracy": 0.2637300506234169,
      "num_tokens": 95339796.0,
      "step": 41580
    },
    {
      "entropy": 5.010869932174683,
      "epoch": 3.994716618635927,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003487665900179545,
      "loss": 4.5621,
      "mean_token_accuracy": 0.26534209549427035,
      "num_tokens": 95351113.0,
      "step": 41585
    },
    {
      "entropy": 4.989310073852539,
      "epoch": 3.9951969260326607,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003487342004835366,
      "loss": 4.4847,
      "mean_token_accuracy": 0.2653703704476357,
      "num_tokens": 95363459.0,
      "step": 41590
    },
    {
      "entropy": 5.034049844741821,
      "epoch": 3.9956772334293946,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003487018092372294,
      "loss": 4.5613,
      "mean_token_accuracy": 0.26119697391986846,
      "num_tokens": 95374015.0,
      "step": 41595
    },
    {
      "entropy": 5.037086868286133,
      "epoch": 3.996157540826129,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000348669416279785,
      "loss": 4.503,
      "mean_token_accuracy": 0.2699758753180504,
      "num_tokens": 95384335.0,
      "step": 41600
    },
    {
      "entropy": 4.970576572418213,
      "epoch": 3.9966378482228624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034863702161195566,
      "loss": 4.5827,
      "mean_token_accuracy": 0.25566050261259077,
      "num_tokens": 95397278.0,
      "step": 41605
    },
    {
      "entropy": 5.1230302333831785,
      "epoch": 3.9971181556195967,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003486046252344933,
      "loss": 4.6427,
      "mean_token_accuracy": 0.25485625863075256,
      "num_tokens": 95408165.0,
      "step": 41610
    },
    {
      "entropy": 5.130393075942993,
      "epoch": 3.9975984630163306,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034857222714814996,
      "loss": 4.7092,
      "mean_token_accuracy": 0.24931791424751282,
      "num_tokens": 95418789.0,
      "step": 41615
    },
    {
      "entropy": 5.067824268341065,
      "epoch": 3.9980787704130645,
      "grad_norm": 1.0,
      "learning_rate": 0.000348539827353678,
      "loss": 4.5869,
      "mean_token_accuracy": 0.25763340294361115,
      "num_tokens": 95428964.0,
      "step": 41620
    },
    {
      "entropy": 5.116991138458252,
      "epoch": 3.9985590778097984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003485074258518298,
      "loss": 4.6243,
      "mean_token_accuracy": 0.2628925502300262,
      "num_tokens": 95439732.0,
      "step": 41625
    },
    {
      "entropy": 5.128873586654663,
      "epoch": 3.9990393852065322,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003484750226433573,
      "loss": 4.6808,
      "mean_token_accuracy": 0.2517023295164108,
      "num_tokens": 95451267.0,
      "step": 41630
    },
    {
      "entropy": 5.075998973846436,
      "epoch": 3.999519692603266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034844261772901293,
      "loss": 4.6686,
      "mean_token_accuracy": 0.2564258113503456,
      "num_tokens": 95463379.0,
      "step": 41635
    },
    {
      "entropy": 5.1885018825531,
      "epoch": 4.0,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003484102111095492,
      "loss": 4.7733,
      "mean_token_accuracy": 0.24711171835660933,
      "num_tokens": 95474144.0,
      "step": 41640
    },
    {
      "entropy": 5.1357502937316895,
      "epoch": 4.000480307396734,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034837780278571814,
      "loss": 4.5486,
      "mean_token_accuracy": 0.2608508110046387,
      "num_tokens": 95485013.0,
      "step": 41645
    },
    {
      "entropy": 5.097986698150635,
      "epoch": 4.000960614793468,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003483453927582725,
      "loss": 4.5643,
      "mean_token_accuracy": 0.2564060777425766,
      "num_tokens": 95495866.0,
      "step": 41650
    },
    {
      "entropy": 5.1033697605133055,
      "epoch": 4.001440922190202,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003483129810279647,
      "loss": 4.5843,
      "mean_token_accuracy": 0.26487377732992173,
      "num_tokens": 95507675.0,
      "step": 41655
    },
    {
      "entropy": 5.103724575042724,
      "epoch": 4.0019212295869355,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000348280567595547,
      "loss": 4.5631,
      "mean_token_accuracy": 0.26090475022792814,
      "num_tokens": 95520000.0,
      "step": 41660
    },
    {
      "entropy": 5.2137939453125,
      "epoch": 4.00240153698367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034824815246177227,
      "loss": 4.6649,
      "mean_token_accuracy": 0.25210620611906054,
      "num_tokens": 95531010.0,
      "step": 41665
    },
    {
      "entropy": 5.100002765655518,
      "epoch": 4.002881844380403,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003482157356273929,
      "loss": 4.6113,
      "mean_token_accuracy": 0.258753077685833,
      "num_tokens": 95542662.0,
      "step": 41670
    },
    {
      "entropy": 5.139379644393921,
      "epoch": 4.003362151777138,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003481833170931615,
      "loss": 4.5908,
      "mean_token_accuracy": 0.25734581649303434,
      "num_tokens": 95553471.0,
      "step": 41675
    },
    {
      "entropy": 5.030312442779541,
      "epoch": 4.003842459173871,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034815089685983085,
      "loss": 4.4849,
      "mean_token_accuracy": 0.26884002834558485,
      "num_tokens": 95563935.0,
      "step": 41680
    },
    {
      "entropy": 4.956342983245849,
      "epoch": 4.004322766570605,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003481184749281536,
      "loss": 4.4648,
      "mean_token_accuracy": 0.2649632468819618,
      "num_tokens": 95576516.0,
      "step": 41685
    },
    {
      "entropy": 5.0513513565063475,
      "epoch": 4.004803073967339,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034808605129888246,
      "loss": 4.5482,
      "mean_token_accuracy": 0.2679751381278038,
      "num_tokens": 95587877.0,
      "step": 41690
    },
    {
      "entropy": 5.107216167449951,
      "epoch": 4.005283381364073,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034805362597277017,
      "loss": 4.5849,
      "mean_token_accuracy": 0.25523918867111206,
      "num_tokens": 95599759.0,
      "step": 41695
    },
    {
      "entropy": 5.073794889450073,
      "epoch": 4.005763688760807,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034802119895056973,
      "loss": 4.5133,
      "mean_token_accuracy": 0.2628215357661247,
      "num_tokens": 95612137.0,
      "step": 41700
    },
    {
      "entropy": 5.047576665878296,
      "epoch": 4.006243996157541,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034798877023303385,
      "loss": 4.5689,
      "mean_token_accuracy": 0.26148719489574435,
      "num_tokens": 95623922.0,
      "step": 41705
    },
    {
      "entropy": 5.194438552856445,
      "epoch": 4.006724303554274,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034795633982091534,
      "loss": 4.6647,
      "mean_token_accuracy": 0.24399442225694656,
      "num_tokens": 95634998.0,
      "step": 41710
    },
    {
      "entropy": 5.113803482055664,
      "epoch": 4.007204610951009,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034792390771496735,
      "loss": 4.5742,
      "mean_token_accuracy": 0.25507057905197145,
      "num_tokens": 95646261.0,
      "step": 41715
    },
    {
      "entropy": 5.078635787963867,
      "epoch": 4.007684918347743,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034789147391594275,
      "loss": 4.5541,
      "mean_token_accuracy": 0.2659521609544754,
      "num_tokens": 95657606.0,
      "step": 41720
    },
    {
      "entropy": 5.1151121139526365,
      "epoch": 4.008165225744476,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003478590384245945,
      "loss": 4.5905,
      "mean_token_accuracy": 0.25365942120552065,
      "num_tokens": 95668854.0,
      "step": 41725
    },
    {
      "entropy": 5.042076063156128,
      "epoch": 4.008645533141211,
      "grad_norm": 1.0,
      "learning_rate": 0.00034782660124167583,
      "loss": 4.559,
      "mean_token_accuracy": 0.25987450033426285,
      "num_tokens": 95680785.0,
      "step": 41730
    },
    {
      "entropy": 5.0800395011901855,
      "epoch": 4.009125840537944,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003477941623679397,
      "loss": 4.5279,
      "mean_token_accuracy": 0.264495487511158,
      "num_tokens": 95692660.0,
      "step": 41735
    },
    {
      "entropy": 5.096365690231323,
      "epoch": 4.0096061479346785,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003477617218041393,
      "loss": 4.578,
      "mean_token_accuracy": 0.2583979621529579,
      "num_tokens": 95703972.0,
      "step": 41740
    },
    {
      "entropy": 5.088635110855103,
      "epoch": 4.010086455331412,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003477292795510277,
      "loss": 4.5534,
      "mean_token_accuracy": 0.26257616877555845,
      "num_tokens": 95716086.0,
      "step": 41745
    },
    {
      "entropy": 5.1117973804473875,
      "epoch": 4.010566762728146,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003476968356093582,
      "loss": 4.5874,
      "mean_token_accuracy": 0.25425514578819275,
      "num_tokens": 95728759.0,
      "step": 41750
    },
    {
      "entropy": 5.033574151992798,
      "epoch": 4.01104707012488,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003476643899798841,
      "loss": 4.4875,
      "mean_token_accuracy": 0.2667245015501976,
      "num_tokens": 95739370.0,
      "step": 41755
    },
    {
      "entropy": 5.119418668746948,
      "epoch": 4.011527377521614,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034763194266335855,
      "loss": 4.6573,
      "mean_token_accuracy": 0.2514065846800804,
      "num_tokens": 95750999.0,
      "step": 41760
    },
    {
      "entropy": 5.130194759368896,
      "epoch": 4.0120076849183475,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003475994936605351,
      "loss": 4.5833,
      "mean_token_accuracy": 0.2573777914047241,
      "num_tokens": 95761982.0,
      "step": 41765
    },
    {
      "entropy": 5.173987102508545,
      "epoch": 4.012487992315082,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034756704297216686,
      "loss": 4.582,
      "mean_token_accuracy": 0.25709699988365176,
      "num_tokens": 95773492.0,
      "step": 41770
    },
    {
      "entropy": 5.050680685043335,
      "epoch": 4.012968299711815,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034753459059900745,
      "loss": 4.5327,
      "mean_token_accuracy": 0.2603677734732628,
      "num_tokens": 95786446.0,
      "step": 41775
    },
    {
      "entropy": 5.118686532974243,
      "epoch": 4.01344860710855,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034750213654181016,
      "loss": 4.6275,
      "mean_token_accuracy": 0.25261824429035185,
      "num_tokens": 95797164.0,
      "step": 41780
    },
    {
      "entropy": 5.057374477386475,
      "epoch": 4.013928914505283,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034746968080132855,
      "loss": 4.5674,
      "mean_token_accuracy": 0.26391961723566054,
      "num_tokens": 95809832.0,
      "step": 41785
    },
    {
      "entropy": 5.1393838882446286,
      "epoch": 4.014409221902017,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003474372233783162,
      "loss": 4.601,
      "mean_token_accuracy": 0.2679472863674164,
      "num_tokens": 95821117.0,
      "step": 41790
    },
    {
      "entropy": 5.002946233749389,
      "epoch": 4.014889529298751,
      "grad_norm": 1.0,
      "learning_rate": 0.00034740476427352657,
      "loss": 4.4112,
      "mean_token_accuracy": 0.2727580890059471,
      "num_tokens": 95831795.0,
      "step": 41795
    },
    {
      "entropy": 5.042566394805908,
      "epoch": 4.015369836695485,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003473723034877133,
      "loss": 4.5538,
      "mean_token_accuracy": 0.2646966964006424,
      "num_tokens": 95842701.0,
      "step": 41800
    },
    {
      "entropy": 5.088676881790161,
      "epoch": 4.015850144092219,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034733984102163005,
      "loss": 4.5716,
      "mean_token_accuracy": 0.2555494025349617,
      "num_tokens": 95853616.0,
      "step": 41805
    },
    {
      "entropy": 5.049332904815674,
      "epoch": 4.016330451488953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003473073768760306,
      "loss": 4.5033,
      "mean_token_accuracy": 0.26390389949083326,
      "num_tokens": 95865423.0,
      "step": 41810
    },
    {
      "entropy": 5.166158056259155,
      "epoch": 4.016810758885687,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034727491105166847,
      "loss": 4.5812,
      "mean_token_accuracy": 0.25769004225730896,
      "num_tokens": 95876058.0,
      "step": 41815
    },
    {
      "entropy": 5.060488939285278,
      "epoch": 4.017291066282421,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003472424435492975,
      "loss": 4.5748,
      "mean_token_accuracy": 0.2616377010941505,
      "num_tokens": 95887777.0,
      "step": 41820
    },
    {
      "entropy": 5.069109869003296,
      "epoch": 4.017771373679155,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003472099743696716,
      "loss": 4.535,
      "mean_token_accuracy": 0.2650677964091301,
      "num_tokens": 95899003.0,
      "step": 41825
    },
    {
      "entropy": 5.098663091659546,
      "epoch": 4.018251681075888,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003471775035135444,
      "loss": 4.5688,
      "mean_token_accuracy": 0.26519678086042403,
      "num_tokens": 95910409.0,
      "step": 41830
    },
    {
      "entropy": 5.0915576934814455,
      "epoch": 4.018731988472623,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034714503098166996,
      "loss": 4.5349,
      "mean_token_accuracy": 0.26004113256931305,
      "num_tokens": 95920923.0,
      "step": 41835
    },
    {
      "entropy": 5.058418273925781,
      "epoch": 4.019212295869356,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034711255677480216,
      "loss": 4.588,
      "mean_token_accuracy": 0.2598757892847061,
      "num_tokens": 95933029.0,
      "step": 41840
    },
    {
      "entropy": 5.119459867477417,
      "epoch": 4.0196926032660905,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003470800808936948,
      "loss": 4.5724,
      "mean_token_accuracy": 0.2620282843708992,
      "num_tokens": 95945617.0,
      "step": 41845
    },
    {
      "entropy": 5.070283651351929,
      "epoch": 4.020172910662824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034704760333910205,
      "loss": 4.5508,
      "mean_token_accuracy": 0.26018992364406585,
      "num_tokens": 95956522.0,
      "step": 41850
    },
    {
      "entropy": 5.106172943115235,
      "epoch": 4.020653218059558,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003470151241117779,
      "loss": 4.6087,
      "mean_token_accuracy": 0.2567663684487343,
      "num_tokens": 95967666.0,
      "step": 41855
    },
    {
      "entropy": 5.053364610671997,
      "epoch": 4.021133525456292,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003469826432124764,
      "loss": 4.447,
      "mean_token_accuracy": 0.26287035942077636,
      "num_tokens": 95978822.0,
      "step": 41860
    },
    {
      "entropy": 5.09239616394043,
      "epoch": 4.021613832853026,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003469501606419517,
      "loss": 4.5817,
      "mean_token_accuracy": 0.25613055676221846,
      "num_tokens": 95989695.0,
      "step": 41865
    },
    {
      "entropy": 5.054267406463623,
      "epoch": 4.022094140249759,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003469176764009579,
      "loss": 4.484,
      "mean_token_accuracy": 0.26939487606287005,
      "num_tokens": 96000373.0,
      "step": 41870
    },
    {
      "entropy": 5.075977754592896,
      "epoch": 4.022574447646494,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034688519049024924,
      "loss": 4.5655,
      "mean_token_accuracy": 0.2664172753691673,
      "num_tokens": 96011495.0,
      "step": 41875
    },
    {
      "entropy": 5.026505661010742,
      "epoch": 4.023054755043228,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034685270291057987,
      "loss": 4.4884,
      "mean_token_accuracy": 0.2676610141992569,
      "num_tokens": 96022558.0,
      "step": 41880
    },
    {
      "entropy": 5.013860559463501,
      "epoch": 4.0235350624399615,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034682021366270403,
      "loss": 4.4603,
      "mean_token_accuracy": 0.26711770296096804,
      "num_tokens": 96033534.0,
      "step": 41885
    },
    {
      "entropy": 5.071307277679443,
      "epoch": 4.024015369836696,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034678772274737626,
      "loss": 4.6271,
      "mean_token_accuracy": 0.25526183247566225,
      "num_tokens": 96046846.0,
      "step": 41890
    },
    {
      "entropy": 5.0328206539154055,
      "epoch": 4.024495677233429,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034675523016535067,
      "loss": 4.5397,
      "mean_token_accuracy": 0.25673589408397673,
      "num_tokens": 96058034.0,
      "step": 41895
    },
    {
      "entropy": 5.100937080383301,
      "epoch": 4.024975984630164,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034672273591738155,
      "loss": 4.586,
      "mean_token_accuracy": 0.2583854109048843,
      "num_tokens": 96070467.0,
      "step": 41900
    },
    {
      "entropy": 5.0912140846252445,
      "epoch": 4.025456292026897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003466902400042236,
      "loss": 4.6018,
      "mean_token_accuracy": 0.2607292667031288,
      "num_tokens": 96082019.0,
      "step": 41905
    },
    {
      "entropy": 5.125153732299805,
      "epoch": 4.025936599423631,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003466577424266312,
      "loss": 4.5694,
      "mean_token_accuracy": 0.25791977643966674,
      "num_tokens": 96093976.0,
      "step": 41910
    },
    {
      "entropy": 5.170462274551392,
      "epoch": 4.026416906820365,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034662524318535877,
      "loss": 4.6327,
      "mean_token_accuracy": 0.25480874329805375,
      "num_tokens": 96105882.0,
      "step": 41915
    },
    {
      "entropy": 5.057155513763428,
      "epoch": 4.026897214217099,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034659274228116086,
      "loss": 4.5585,
      "mean_token_accuracy": 0.26065066605806353,
      "num_tokens": 96116154.0,
      "step": 41920
    },
    {
      "entropy": 5.055847120285034,
      "epoch": 4.027377521613833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034656023971479206,
      "loss": 4.5664,
      "mean_token_accuracy": 0.26071166098117826,
      "num_tokens": 96128626.0,
      "step": 41925
    },
    {
      "entropy": 5.046073627471924,
      "epoch": 4.027857829010567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003465277354870069,
      "loss": 4.4954,
      "mean_token_accuracy": 0.263053597509861,
      "num_tokens": 96140205.0,
      "step": 41930
    },
    {
      "entropy": 5.113253879547119,
      "epoch": 4.0283381364073,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034649522959856027,
      "loss": 4.5911,
      "mean_token_accuracy": 0.2658152520656586,
      "num_tokens": 96152577.0,
      "step": 41935
    },
    {
      "entropy": 4.967879724502564,
      "epoch": 4.028818443804035,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034646272205020664,
      "loss": 4.4122,
      "mean_token_accuracy": 0.27329352796077727,
      "num_tokens": 96163500.0,
      "step": 41940
    },
    {
      "entropy": 5.112587976455688,
      "epoch": 4.029298751200768,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003464302128427008,
      "loss": 4.6095,
      "mean_token_accuracy": 0.25425254106521605,
      "num_tokens": 96174259.0,
      "step": 41945
    },
    {
      "entropy": 5.098554420471191,
      "epoch": 4.029779058597502,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003463977019767975,
      "loss": 4.5721,
      "mean_token_accuracy": 0.2580135941505432,
      "num_tokens": 96186823.0,
      "step": 41950
    },
    {
      "entropy": 5.10051007270813,
      "epoch": 4.030259365994237,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003463651894532516,
      "loss": 4.5952,
      "mean_token_accuracy": 0.25465919077396393,
      "num_tokens": 96197685.0,
      "step": 41955
    },
    {
      "entropy": 5.067636394500733,
      "epoch": 4.03073967339097,
      "grad_norm": 1.0625,
      "learning_rate": 0.000346332675272818,
      "loss": 4.5357,
      "mean_token_accuracy": 0.26268841326236725,
      "num_tokens": 96208307.0,
      "step": 41960
    },
    {
      "entropy": 5.1000203609466555,
      "epoch": 4.0312199807877045,
      "grad_norm": 1.0,
      "learning_rate": 0.0003463001594362513,
      "loss": 4.5667,
      "mean_token_accuracy": 0.26624458432197573,
      "num_tokens": 96219703.0,
      "step": 41965
    },
    {
      "entropy": 5.1693230152130125,
      "epoch": 4.031700288184438,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003462676419443067,
      "loss": 4.6453,
      "mean_token_accuracy": 0.2546707794070244,
      "num_tokens": 96231013.0,
      "step": 41970
    },
    {
      "entropy": 5.116321563720703,
      "epoch": 4.032180595581172,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003462351227977391,
      "loss": 4.6226,
      "mean_token_accuracy": 0.25643702000379565,
      "num_tokens": 96242542.0,
      "step": 41975
    },
    {
      "entropy": 5.197872972488403,
      "epoch": 4.032660902977906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034620260199730345,
      "loss": 4.6622,
      "mean_token_accuracy": 0.24939454942941666,
      "num_tokens": 96252829.0,
      "step": 41980
    },
    {
      "entropy": 5.064321279525757,
      "epoch": 4.03314121037464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034617007954375486,
      "loss": 4.4438,
      "mean_token_accuracy": 0.2701556831598282,
      "num_tokens": 96263306.0,
      "step": 41985
    },
    {
      "entropy": 5.144461011886596,
      "epoch": 4.0336215177713735,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003461375554378483,
      "loss": 4.6989,
      "mean_token_accuracy": 0.2504366472363472,
      "num_tokens": 96275369.0,
      "step": 41990
    },
    {
      "entropy": 5.022068691253662,
      "epoch": 4.034101825168108,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034610502968033895,
      "loss": 4.4155,
      "mean_token_accuracy": 0.2743390053510666,
      "num_tokens": 96285332.0,
      "step": 41995
    },
    {
      "entropy": 5.019009780883789,
      "epoch": 4.034582132564841,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000346072502271982,
      "loss": 4.5413,
      "mean_token_accuracy": 0.2548611700534821,
      "num_tokens": 96295819.0,
      "step": 42000
    },
    {
      "epoch": 4.034582132564841,
      "eval_entropy": 4.848282634194801,
      "eval_loss": 4.736080169677734,
      "eval_mean_token_accuracy": 0.2572913851919536,
      "eval_num_tokens": 96295819.0,
      "eval_runtime": 26.4839,
      "eval_samples_per_second": 1239.053,
      "eval_steps_per_second": 154.886,
      "step": 42000
    },
    {
      "entropy": 5.068308973312378,
      "epoch": 4.0350624399615755,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003460399732135326,
      "loss": 4.5959,
      "mean_token_accuracy": 0.2658592686057091,
      "num_tokens": 96308585.0,
      "step": 42005
    },
    {
      "entropy": 5.136622762680053,
      "epoch": 4.035542747358309,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003460074425057458,
      "loss": 4.59,
      "mean_token_accuracy": 0.2514837637543678,
      "num_tokens": 96320122.0,
      "step": 42010
    },
    {
      "entropy": 5.063159990310669,
      "epoch": 4.036023054755043,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034597491014937723,
      "loss": 4.5623,
      "mean_token_accuracy": 0.2572839468717575,
      "num_tokens": 96331878.0,
      "step": 42015
    },
    {
      "entropy": 5.110752964019776,
      "epoch": 4.036503362151777,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003459423761451819,
      "loss": 4.5709,
      "mean_token_accuracy": 0.26069018095731733,
      "num_tokens": 96344286.0,
      "step": 42020
    },
    {
      "entropy": 5.085515403747559,
      "epoch": 4.036983669548511,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003459098404939153,
      "loss": 4.5572,
      "mean_token_accuracy": 0.2651988387107849,
      "num_tokens": 96357894.0,
      "step": 42025
    },
    {
      "entropy": 5.023877573013306,
      "epoch": 4.037463976945245,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034587730319633265,
      "loss": 4.5034,
      "mean_token_accuracy": 0.2704067438840866,
      "num_tokens": 96370213.0,
      "step": 42030
    },
    {
      "entropy": 5.08345046043396,
      "epoch": 4.037944284341979,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034584476425318964,
      "loss": 4.5857,
      "mean_token_accuracy": 0.2528646171092987,
      "num_tokens": 96381135.0,
      "step": 42035
    },
    {
      "entropy": 5.077811288833618,
      "epoch": 4.038424591738713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034581222366524147,
      "loss": 4.5533,
      "mean_token_accuracy": 0.2600689798593521,
      "num_tokens": 96392279.0,
      "step": 42040
    },
    {
      "entropy": 5.064612770080567,
      "epoch": 4.038904899135447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034577968143324365,
      "loss": 4.5191,
      "mean_token_accuracy": 0.26835009306669233,
      "num_tokens": 96403225.0,
      "step": 42045
    },
    {
      "entropy": 5.091956758499146,
      "epoch": 4.039385206532181,
      "grad_norm": 1.0625,
      "learning_rate": 0.000345747137557952,
      "loss": 4.5344,
      "mean_token_accuracy": 0.25958815813064573,
      "num_tokens": 96413062.0,
      "step": 42050
    },
    {
      "entropy": 5.10804591178894,
      "epoch": 4.039865513928914,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003457145920401218,
      "loss": 4.6086,
      "mean_token_accuracy": 0.2590765044093132,
      "num_tokens": 96424828.0,
      "step": 42055
    },
    {
      "entropy": 5.175853490829468,
      "epoch": 4.040345821325649,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034568204488050875,
      "loss": 4.6192,
      "mean_token_accuracy": 0.24854864776134492,
      "num_tokens": 96436476.0,
      "step": 42060
    },
    {
      "entropy": 5.0329824924469,
      "epoch": 4.040826128722382,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003456494960798685,
      "loss": 4.485,
      "mean_token_accuracy": 0.2675792142748833,
      "num_tokens": 96448812.0,
      "step": 42065
    },
    {
      "entropy": 5.132245111465454,
      "epoch": 4.041306436119116,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003456169456389567,
      "loss": 4.6666,
      "mean_token_accuracy": 0.2519511103630066,
      "num_tokens": 96461272.0,
      "step": 42070
    },
    {
      "entropy": 5.143880081176758,
      "epoch": 4.04178674351585,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034558439355852915,
      "loss": 4.5667,
      "mean_token_accuracy": 0.2625057533383369,
      "num_tokens": 96472212.0,
      "step": 42075
    },
    {
      "entropy": 5.091010999679566,
      "epoch": 4.042267050912584,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003455518398393416,
      "loss": 4.5608,
      "mean_token_accuracy": 0.2639792487025261,
      "num_tokens": 96484866.0,
      "step": 42080
    },
    {
      "entropy": 5.171709775924683,
      "epoch": 4.042747358309318,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034551928448214983,
      "loss": 4.6102,
      "mean_token_accuracy": 0.25324823409318925,
      "num_tokens": 96495745.0,
      "step": 42085
    },
    {
      "entropy": 5.03561635017395,
      "epoch": 4.043227665706052,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003454867274877096,
      "loss": 4.5268,
      "mean_token_accuracy": 0.2590127155184746,
      "num_tokens": 96507428.0,
      "step": 42090
    },
    {
      "entropy": 5.116614866256714,
      "epoch": 4.043707973102785,
      "grad_norm": 1.125,
      "learning_rate": 0.00034545416885677695,
      "loss": 4.5891,
      "mean_token_accuracy": 0.26098636239767076,
      "num_tokens": 96517700.0,
      "step": 42095
    },
    {
      "entropy": 4.997502326965332,
      "epoch": 4.04418828049952,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003454216085901076,
      "loss": 4.4397,
      "mean_token_accuracy": 0.27768113613128664,
      "num_tokens": 96528977.0,
      "step": 42100
    },
    {
      "entropy": 5.011810827255249,
      "epoch": 4.044668587896253,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00034538904668845766,
      "loss": 4.4958,
      "mean_token_accuracy": 0.2703707188367844,
      "num_tokens": 96540062.0,
      "step": 42105
    },
    {
      "entropy": 5.042691612243653,
      "epoch": 4.0451488952929875,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003453564831525831,
      "loss": 4.5054,
      "mean_token_accuracy": 0.2624609813094139,
      "num_tokens": 96551087.0,
      "step": 42110
    },
    {
      "entropy": 5.069543933868408,
      "epoch": 4.045629202689722,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003453239179832399,
      "loss": 4.5723,
      "mean_token_accuracy": 0.264744833111763,
      "num_tokens": 96562260.0,
      "step": 42115
    },
    {
      "entropy": 5.00747013092041,
      "epoch": 4.046109510086455,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003452913511811841,
      "loss": 4.4252,
      "mean_token_accuracy": 0.2698741897940636,
      "num_tokens": 96572912.0,
      "step": 42120
    },
    {
      "entropy": 5.074120616912841,
      "epoch": 4.04658981748319,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003452587827471719,
      "loss": 4.5802,
      "mean_token_accuracy": 0.26134087443351744,
      "num_tokens": 96585292.0,
      "step": 42125
    },
    {
      "entropy": 5.050881814956665,
      "epoch": 4.047070124879923,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034522621268195936,
      "loss": 4.5122,
      "mean_token_accuracy": 0.2722261056303978,
      "num_tokens": 96597061.0,
      "step": 42130
    },
    {
      "entropy": 5.024364805221557,
      "epoch": 4.047550432276657,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003451936409863027,
      "loss": 4.5038,
      "mean_token_accuracy": 0.26334591805934904,
      "num_tokens": 96609289.0,
      "step": 42135
    },
    {
      "entropy": 5.032188081741333,
      "epoch": 4.048030739673391,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034516106766095813,
      "loss": 4.4829,
      "mean_token_accuracy": 0.27068856805562974,
      "num_tokens": 96620304.0,
      "step": 42140
    },
    {
      "entropy": 5.035274648666382,
      "epoch": 4.048511047070125,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003451284927066819,
      "loss": 4.5573,
      "mean_token_accuracy": 0.2618005365133286,
      "num_tokens": 96631229.0,
      "step": 42145
    },
    {
      "entropy": 5.182892370223999,
      "epoch": 4.0489913544668585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003450959161242303,
      "loss": 4.6497,
      "mean_token_accuracy": 0.2598159149289131,
      "num_tokens": 96643534.0,
      "step": 42150
    },
    {
      "entropy": 5.0772521018981935,
      "epoch": 4.049471661863593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003450633379143597,
      "loss": 4.5442,
      "mean_token_accuracy": 0.26151061356067656,
      "num_tokens": 96654111.0,
      "step": 42155
    },
    {
      "entropy": 5.085466194152832,
      "epoch": 4.049951969260326,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00034503075807782634,
      "loss": 4.5196,
      "mean_token_accuracy": 0.2698402628302574,
      "num_tokens": 96666676.0,
      "step": 42160
    },
    {
      "entropy": 5.007355642318726,
      "epoch": 4.050432276657061,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003449981766153869,
      "loss": 4.4214,
      "mean_token_accuracy": 0.2790235370397568,
      "num_tokens": 96677903.0,
      "step": 42165
    },
    {
      "entropy": 4.952964448928833,
      "epoch": 4.050912584053794,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003449655935277976,
      "loss": 4.4403,
      "mean_token_accuracy": 0.2680338308215141,
      "num_tokens": 96689582.0,
      "step": 42170
    },
    {
      "entropy": 5.1532494068145756,
      "epoch": 4.051392891450528,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034493300881581484,
      "loss": 4.6229,
      "mean_token_accuracy": 0.2534977585077286,
      "num_tokens": 96700195.0,
      "step": 42175
    },
    {
      "entropy": 5.003329133987426,
      "epoch": 4.051873198847262,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003449004224801954,
      "loss": 4.4456,
      "mean_token_accuracy": 0.2693677693605423,
      "num_tokens": 96713171.0,
      "step": 42180
    },
    {
      "entropy": 5.0415503025054935,
      "epoch": 4.052353506243996,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003448678345216957,
      "loss": 4.5941,
      "mean_token_accuracy": 0.25642693787813187,
      "num_tokens": 96726239.0,
      "step": 42185
    },
    {
      "entropy": 4.9465868949890135,
      "epoch": 4.0528338136407305,
      "grad_norm": 0.921875,
      "learning_rate": 0.00034483524494107235,
      "loss": 4.3938,
      "mean_token_accuracy": 0.27224671095609665,
      "num_tokens": 96738112.0,
      "step": 42190
    },
    {
      "entropy": 5.0951423168182375,
      "epoch": 4.053314121037464,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034480265373908187,
      "loss": 4.5517,
      "mean_token_accuracy": 0.26858493536710737,
      "num_tokens": 96749485.0,
      "step": 42195
    },
    {
      "entropy": 5.114979362487793,
      "epoch": 4.053794428434198,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034477006091648116,
      "loss": 4.5912,
      "mean_token_accuracy": 0.2614215314388275,
      "num_tokens": 96760368.0,
      "step": 42200
    },
    {
      "entropy": 5.002750635147095,
      "epoch": 4.054274735830932,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034473746647402674,
      "loss": 4.444,
      "mean_token_accuracy": 0.27048413306474683,
      "num_tokens": 96771071.0,
      "step": 42205
    },
    {
      "entropy": 5.145899868011474,
      "epoch": 4.054755043227666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003447048704124754,
      "loss": 4.6509,
      "mean_token_accuracy": 0.25357774645090103,
      "num_tokens": 96781880.0,
      "step": 42210
    },
    {
      "entropy": 5.150488710403442,
      "epoch": 4.055235350624399,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003446722727325841,
      "loss": 4.593,
      "mean_token_accuracy": 0.2611670151352882,
      "num_tokens": 96793524.0,
      "step": 42215
    },
    {
      "entropy": 5.022638177871704,
      "epoch": 4.055715658021134,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003446396734351094,
      "loss": 4.471,
      "mean_token_accuracy": 0.2653254434466362,
      "num_tokens": 96805848.0,
      "step": 42220
    },
    {
      "entropy": 5.008848190307617,
      "epoch": 4.056195965417867,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003446070725208083,
      "loss": 4.5287,
      "mean_token_accuracy": 0.2651268929243088,
      "num_tokens": 96817585.0,
      "step": 42225
    },
    {
      "entropy": 5.160595655441284,
      "epoch": 4.0566762728146015,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003445744699904377,
      "loss": 4.6992,
      "mean_token_accuracy": 0.24660615772008895,
      "num_tokens": 96829447.0,
      "step": 42230
    },
    {
      "entropy": 5.026753234863281,
      "epoch": 4.057156580211335,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003445418658447544,
      "loss": 4.4947,
      "mean_token_accuracy": 0.27224435806274416,
      "num_tokens": 96842785.0,
      "step": 42235
    },
    {
      "entropy": 5.011379814147949,
      "epoch": 4.057636887608069,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034450926008451556,
      "loss": 4.4276,
      "mean_token_accuracy": 0.2721185445785522,
      "num_tokens": 96853526.0,
      "step": 42240
    },
    {
      "entropy": 5.014807462692261,
      "epoch": 4.058117195004803,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003444766527104781,
      "loss": 4.5009,
      "mean_token_accuracy": 0.2743007466197014,
      "num_tokens": 96864958.0,
      "step": 42245
    },
    {
      "entropy": 5.084926271438599,
      "epoch": 4.058597502401537,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000344444043723399,
      "loss": 4.5069,
      "mean_token_accuracy": 0.2642006456851959,
      "num_tokens": 96875908.0,
      "step": 42250
    },
    {
      "entropy": 5.083907794952393,
      "epoch": 4.0590778097982705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003444114331240355,
      "loss": 4.5308,
      "mean_token_accuracy": 0.25714174956083297,
      "num_tokens": 96887378.0,
      "step": 42255
    },
    {
      "entropy": 5.054579544067383,
      "epoch": 4.059558117195005,
      "grad_norm": 1.0,
      "learning_rate": 0.0003443788209131446,
      "loss": 4.5355,
      "mean_token_accuracy": 0.2611877009272575,
      "num_tokens": 96898183.0,
      "step": 42260
    },
    {
      "entropy": 5.090785217285156,
      "epoch": 4.060038424591739,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034434620709148343,
      "loss": 4.5049,
      "mean_token_accuracy": 0.2623579457402229,
      "num_tokens": 96909977.0,
      "step": 42265
    },
    {
      "entropy": 5.103012800216675,
      "epoch": 4.060518731988473,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003443135916598093,
      "loss": 4.6323,
      "mean_token_accuracy": 0.24946069866418838,
      "num_tokens": 96921044.0,
      "step": 42270
    },
    {
      "entropy": 4.918801736831665,
      "epoch": 4.060999039385207,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034428097461887934,
      "loss": 4.4642,
      "mean_token_accuracy": 0.27226513922214507,
      "num_tokens": 96931831.0,
      "step": 42275
    },
    {
      "entropy": 5.108166313171386,
      "epoch": 4.06147934678194,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003442483559694509,
      "loss": 4.602,
      "mean_token_accuracy": 0.2650128871202469,
      "num_tokens": 96943639.0,
      "step": 42280
    },
    {
      "entropy": 5.018485116958618,
      "epoch": 4.061959654178675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034421573571228134,
      "loss": 4.5065,
      "mean_token_accuracy": 0.26419263035058976,
      "num_tokens": 96955792.0,
      "step": 42285
    },
    {
      "entropy": 5.177859306335449,
      "epoch": 4.062439961575408,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003441831138481279,
      "loss": 4.6436,
      "mean_token_accuracy": 0.25357927531003954,
      "num_tokens": 96967610.0,
      "step": 42290
    },
    {
      "entropy": 5.1155870914459225,
      "epoch": 4.062920268972142,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003441504903777479,
      "loss": 4.5485,
      "mean_token_accuracy": 0.26758098751306536,
      "num_tokens": 96979548.0,
      "step": 42295
    },
    {
      "entropy": 5.083858489990234,
      "epoch": 4.063400576368876,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034411786530189894,
      "loss": 4.5597,
      "mean_token_accuracy": 0.2580718770623207,
      "num_tokens": 96991517.0,
      "step": 42300
    },
    {
      "entropy": 5.107615947723389,
      "epoch": 4.06388088376561,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034408523862133834,
      "loss": 4.5545,
      "mean_token_accuracy": 0.2620858162641525,
      "num_tokens": 97002320.0,
      "step": 42305
    },
    {
      "entropy": 5.039994812011718,
      "epoch": 4.064361191162344,
      "grad_norm": 1.0,
      "learning_rate": 0.00034405261033682363,
      "loss": 4.5519,
      "mean_token_accuracy": 0.26358502805233003,
      "num_tokens": 97013596.0,
      "step": 42310
    },
    {
      "entropy": 5.063442802429199,
      "epoch": 4.064841498559078,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003440199804491123,
      "loss": 4.5467,
      "mean_token_accuracy": 0.26262364685535433,
      "num_tokens": 97025933.0,
      "step": 42315
    },
    {
      "entropy": 5.05333309173584,
      "epoch": 4.065321805955811,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003439873489589621,
      "loss": 4.5602,
      "mean_token_accuracy": 0.2560948595404625,
      "num_tokens": 97036239.0,
      "step": 42320
    },
    {
      "entropy": 5.100766611099243,
      "epoch": 4.065802113352546,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003439547158671304,
      "loss": 4.5785,
      "mean_token_accuracy": 0.2590340808033943,
      "num_tokens": 97047085.0,
      "step": 42325
    },
    {
      "entropy": 5.092358064651489,
      "epoch": 4.066282420749279,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034392208117437503,
      "loss": 4.5389,
      "mean_token_accuracy": 0.26140434890985487,
      "num_tokens": 97059080.0,
      "step": 42330
    },
    {
      "entropy": 5.028905534744263,
      "epoch": 4.0667627281460135,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034388944488145356,
      "loss": 4.532,
      "mean_token_accuracy": 0.2595806270837784,
      "num_tokens": 97070760.0,
      "step": 42335
    },
    {
      "entropy": 5.057321214675904,
      "epoch": 4.067243035542747,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034385680698912377,
      "loss": 4.6027,
      "mean_token_accuracy": 0.261161045730114,
      "num_tokens": 97081818.0,
      "step": 42340
    },
    {
      "entropy": 4.971514129638672,
      "epoch": 4.067723342939481,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034382416749814335,
      "loss": 4.454,
      "mean_token_accuracy": 0.2684491977095604,
      "num_tokens": 97093611.0,
      "step": 42345
    },
    {
      "entropy": 5.101085901260376,
      "epoch": 4.068203650336216,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003437915264092701,
      "loss": 4.5411,
      "mean_token_accuracy": 0.26548954993486407,
      "num_tokens": 97104963.0,
      "step": 42350
    },
    {
      "entropy": 5.030142068862915,
      "epoch": 4.068683957732949,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003437588837232619,
      "loss": 4.4545,
      "mean_token_accuracy": 0.26882091760635374,
      "num_tokens": 97116232.0,
      "step": 42355
    },
    {
      "entropy": 5.044739294052124,
      "epoch": 4.069164265129683,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034372623944087656,
      "loss": 4.5435,
      "mean_token_accuracy": 0.2641688033938408,
      "num_tokens": 97126871.0,
      "step": 42360
    },
    {
      "entropy": 5.121575832366943,
      "epoch": 4.069644572526417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003436935935628721,
      "loss": 4.587,
      "mean_token_accuracy": 0.26563057899475095,
      "num_tokens": 97138341.0,
      "step": 42365
    },
    {
      "entropy": 5.034147930145264,
      "epoch": 4.070124879923151,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034366094609000633,
      "loss": 4.4805,
      "mean_token_accuracy": 0.2694145143032074,
      "num_tokens": 97149566.0,
      "step": 42370
    },
    {
      "entropy": 5.024968957901001,
      "epoch": 4.0706051873198845,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003436282970230372,
      "loss": 4.5081,
      "mean_token_accuracy": 0.26412221491336824,
      "num_tokens": 97159668.0,
      "step": 42375
    },
    {
      "entropy": 5.07321605682373,
      "epoch": 4.071085494716619,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034359564636272287,
      "loss": 4.5733,
      "mean_token_accuracy": 0.25377473831176756,
      "num_tokens": 97171502.0,
      "step": 42380
    },
    {
      "entropy": 5.1445252895355225,
      "epoch": 4.071565802113352,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003435629941098213,
      "loss": 4.5756,
      "mean_token_accuracy": 0.25695735067129133,
      "num_tokens": 97182758.0,
      "step": 42385
    },
    {
      "entropy": 5.064072465896606,
      "epoch": 4.072046109510087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003435303402650906,
      "loss": 4.5613,
      "mean_token_accuracy": 0.2640271961688995,
      "num_tokens": 97193534.0,
      "step": 42390
    },
    {
      "entropy": 5.028741121292114,
      "epoch": 4.07252641690682,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000343497684829289,
      "loss": 4.5391,
      "mean_token_accuracy": 0.2617288649082184,
      "num_tokens": 97204877.0,
      "step": 42395
    },
    {
      "entropy": 5.105880403518677,
      "epoch": 4.073006724303554,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003434650278031744,
      "loss": 4.6132,
      "mean_token_accuracy": 0.2584720656275749,
      "num_tokens": 97217132.0,
      "step": 42400
    },
    {
      "entropy": 5.141275072097779,
      "epoch": 4.073487031700288,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034343236918750523,
      "loss": 4.6472,
      "mean_token_accuracy": 0.2508707821369171,
      "num_tokens": 97228549.0,
      "step": 42405
    },
    {
      "entropy": 5.13331241607666,
      "epoch": 4.073967339097022,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034339970898303963,
      "loss": 4.6138,
      "mean_token_accuracy": 0.2580729529261589,
      "num_tokens": 97239852.0,
      "step": 42410
    },
    {
      "entropy": 5.030359268188477,
      "epoch": 4.074447646493756,
      "grad_norm": 0.90625,
      "learning_rate": 0.00034336704719053584,
      "loss": 4.5358,
      "mean_token_accuracy": 0.26769891530275347,
      "num_tokens": 97251537.0,
      "step": 42415
    },
    {
      "entropy": 5.128170728683472,
      "epoch": 4.07492795389049,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003433343838107523,
      "loss": 4.6532,
      "mean_token_accuracy": 0.2567164972424507,
      "num_tokens": 97263076.0,
      "step": 42420
    },
    {
      "entropy": 5.0553446292877195,
      "epoch": 4.075408261287224,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034330171884444725,
      "loss": 4.466,
      "mean_token_accuracy": 0.2724316596984863,
      "num_tokens": 97273323.0,
      "step": 42425
    },
    {
      "entropy": 5.052630662918091,
      "epoch": 4.075888568683958,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034326905229237914,
      "loss": 4.4791,
      "mean_token_accuracy": 0.26889060735702514,
      "num_tokens": 97284076.0,
      "step": 42430
    },
    {
      "entropy": 5.114502954483032,
      "epoch": 4.076368876080692,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034323638415530627,
      "loss": 4.5736,
      "mean_token_accuracy": 0.2512586131691933,
      "num_tokens": 97295930.0,
      "step": 42435
    },
    {
      "entropy": 5.123907804489136,
      "epoch": 4.076849183477425,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003432037144339872,
      "loss": 4.6115,
      "mean_token_accuracy": 0.2600629165768623,
      "num_tokens": 97306908.0,
      "step": 42440
    },
    {
      "entropy": 5.037136030197144,
      "epoch": 4.07732949087416,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00034317104312918054,
      "loss": 4.582,
      "mean_token_accuracy": 0.2608344316482544,
      "num_tokens": 97319475.0,
      "step": 42445
    },
    {
      "entropy": 5.033924388885498,
      "epoch": 4.077809798270893,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003431383702416445,
      "loss": 4.5324,
      "mean_token_accuracy": 0.2627349510788918,
      "num_tokens": 97330057.0,
      "step": 42450
    },
    {
      "entropy": 5.06527361869812,
      "epoch": 4.0782901056676275,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034310569577213795,
      "loss": 4.5129,
      "mean_token_accuracy": 0.26514274924993514,
      "num_tokens": 97341666.0,
      "step": 42455
    },
    {
      "entropy": 5.0942771434783936,
      "epoch": 4.078770413064361,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003430730197214193,
      "loss": 4.5779,
      "mean_token_accuracy": 0.261427041888237,
      "num_tokens": 97353490.0,
      "step": 42460
    },
    {
      "entropy": 5.043814563751221,
      "epoch": 4.079250720461095,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003430403420902473,
      "loss": 4.515,
      "mean_token_accuracy": 0.2626552924513817,
      "num_tokens": 97366109.0,
      "step": 42465
    },
    {
      "entropy": 5.054933404922485,
      "epoch": 4.079731027857829,
      "grad_norm": 1.0,
      "learning_rate": 0.00034300766287938066,
      "loss": 4.5469,
      "mean_token_accuracy": 0.26228996813297273,
      "num_tokens": 97378134.0,
      "step": 42470
    },
    {
      "entropy": 5.142154264450073,
      "epoch": 4.080211335254563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034297498208957797,
      "loss": 4.6262,
      "mean_token_accuracy": 0.25263641178607943,
      "num_tokens": 97389494.0,
      "step": 42475
    },
    {
      "entropy": 5.11395525932312,
      "epoch": 4.0806916426512965,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000342942299721598,
      "loss": 4.6047,
      "mean_token_accuracy": 0.2533433198928833,
      "num_tokens": 97400950.0,
      "step": 42480
    },
    {
      "entropy": 4.988726282119751,
      "epoch": 4.081171950048031,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003429096157761997,
      "loss": 4.4179,
      "mean_token_accuracy": 0.2714580178260803,
      "num_tokens": 97411704.0,
      "step": 42485
    },
    {
      "entropy": 5.087080621719361,
      "epoch": 4.081652257444764,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003428769302541416,
      "loss": 4.5851,
      "mean_token_accuracy": 0.2556873202323914,
      "num_tokens": 97423539.0,
      "step": 42490
    },
    {
      "entropy": 5.102409887313843,
      "epoch": 4.082132564841499,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003428442431561828,
      "loss": 4.5644,
      "mean_token_accuracy": 0.2635495200753212,
      "num_tokens": 97435443.0,
      "step": 42495
    },
    {
      "entropy": 5.039914751052857,
      "epoch": 4.082612872238233,
      "grad_norm": 1.0,
      "learning_rate": 0.0003428115544830823,
      "loss": 4.4754,
      "mean_token_accuracy": 0.2587674006819725,
      "num_tokens": 97447129.0,
      "step": 42500
    },
    {
      "entropy": 5.058708667755127,
      "epoch": 4.083093179634966,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034277886423559863,
      "loss": 4.513,
      "mean_token_accuracy": 0.26581926047801974,
      "num_tokens": 97458696.0,
      "step": 42505
    },
    {
      "entropy": 5.025904321670533,
      "epoch": 4.083573487031701,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034274617241449105,
      "loss": 4.5571,
      "mean_token_accuracy": 0.26562837809324263,
      "num_tokens": 97470779.0,
      "step": 42510
    },
    {
      "entropy": 5.03430209159851,
      "epoch": 4.084053794428434,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034271347902051864,
      "loss": 4.4998,
      "mean_token_accuracy": 0.2681331723928452,
      "num_tokens": 97482222.0,
      "step": 42515
    },
    {
      "entropy": 5.022475385665894,
      "epoch": 4.084534101825168,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003426807840544401,
      "loss": 4.5057,
      "mean_token_accuracy": 0.2668172612786293,
      "num_tokens": 97493753.0,
      "step": 42520
    },
    {
      "entropy": 5.1058611392974855,
      "epoch": 4.085014409221902,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003426480875170149,
      "loss": 4.5959,
      "mean_token_accuracy": 0.25577609688043595,
      "num_tokens": 97505710.0,
      "step": 42525
    },
    {
      "entropy": 5.053561687469482,
      "epoch": 4.085494716618636,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034261538940900195,
      "loss": 4.4857,
      "mean_token_accuracy": 0.2694190442562103,
      "num_tokens": 97518294.0,
      "step": 42530
    },
    {
      "entropy": 5.07573184967041,
      "epoch": 4.08597502401537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003425826897311604,
      "loss": 4.5233,
      "mean_token_accuracy": 0.2652104407548904,
      "num_tokens": 97529473.0,
      "step": 42535
    },
    {
      "entropy": 5.069411993026733,
      "epoch": 4.086455331412104,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034254998848424947,
      "loss": 4.5108,
      "mean_token_accuracy": 0.2611373573541641,
      "num_tokens": 97540144.0,
      "step": 42540
    },
    {
      "entropy": 5.072084236145019,
      "epoch": 4.086935638808837,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003425172856690284,
      "loss": 4.563,
      "mean_token_accuracy": 0.26258373707532884,
      "num_tokens": 97552413.0,
      "step": 42545
    },
    {
      "entropy": 5.019681358337403,
      "epoch": 4.087415946205572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034248458128625643,
      "loss": 4.5009,
      "mean_token_accuracy": 0.26509013772010803,
      "num_tokens": 97563223.0,
      "step": 42550
    },
    {
      "entropy": 5.098736715316773,
      "epoch": 4.087896253602305,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034245187533669285,
      "loss": 4.5869,
      "mean_token_accuracy": 0.2628322243690491,
      "num_tokens": 97574360.0,
      "step": 42555
    },
    {
      "entropy": 5.198687124252319,
      "epoch": 4.0883765609990395,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034241916782109714,
      "loss": 4.7,
      "mean_token_accuracy": 0.2469273954629898,
      "num_tokens": 97585403.0,
      "step": 42560
    },
    {
      "entropy": 5.09079155921936,
      "epoch": 4.088856868395773,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034238645874022845,
      "loss": 4.5299,
      "mean_token_accuracy": 0.26516222953796387,
      "num_tokens": 97597237.0,
      "step": 42565
    },
    {
      "entropy": 5.015631151199341,
      "epoch": 4.089337175792507,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034235374809484626,
      "loss": 4.5474,
      "mean_token_accuracy": 0.2572324648499489,
      "num_tokens": 97608645.0,
      "step": 42570
    },
    {
      "entropy": 5.030273151397705,
      "epoch": 4.0898174831892415,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003423210358857101,
      "loss": 4.5105,
      "mean_token_accuracy": 0.25498391687870026,
      "num_tokens": 97619138.0,
      "step": 42575
    },
    {
      "entropy": 5.1894371032714846,
      "epoch": 4.090297790585975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034228832211357935,
      "loss": 4.6874,
      "mean_token_accuracy": 0.24818929582834243,
      "num_tokens": 97630793.0,
      "step": 42580
    },
    {
      "entropy": 5.072873163223266,
      "epoch": 4.090778097982709,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034225560677921354,
      "loss": 4.4963,
      "mean_token_accuracy": 0.2707873106002808,
      "num_tokens": 97642680.0,
      "step": 42585
    },
    {
      "entropy": 5.067024755477905,
      "epoch": 4.091258405379443,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034222288988337233,
      "loss": 4.5775,
      "mean_token_accuracy": 0.26021796464920044,
      "num_tokens": 97654247.0,
      "step": 42590
    },
    {
      "entropy": 5.046070003509522,
      "epoch": 4.091738712776177,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003421901714268152,
      "loss": 4.5091,
      "mean_token_accuracy": 0.2614078938961029,
      "num_tokens": 97665460.0,
      "step": 42595
    },
    {
      "entropy": 5.077012157440185,
      "epoch": 4.0922190201729105,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034215745141030177,
      "loss": 4.5832,
      "mean_token_accuracy": 0.26471482068300245,
      "num_tokens": 97678111.0,
      "step": 42600
    },
    {
      "entropy": 5.018032264709473,
      "epoch": 4.092699327569645,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034212472983459176,
      "loss": 4.4943,
      "mean_token_accuracy": 0.25888332426548005,
      "num_tokens": 97689107.0,
      "step": 42605
    },
    {
      "entropy": 5.091714191436767,
      "epoch": 4.093179634966378,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003420920067004448,
      "loss": 4.6064,
      "mean_token_accuracy": 0.2605071559548378,
      "num_tokens": 97700890.0,
      "step": 42610
    },
    {
      "entropy": 5.168421983718872,
      "epoch": 4.093659942363113,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003420592820086207,
      "loss": 4.6535,
      "mean_token_accuracy": 0.24671294987201692,
      "num_tokens": 97711952.0,
      "step": 42615
    },
    {
      "entropy": 5.060731506347656,
      "epoch": 4.094140249759846,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034202655575987925,
      "loss": 4.5444,
      "mean_token_accuracy": 0.26990751922130585,
      "num_tokens": 97723570.0,
      "step": 42620
    },
    {
      "entropy": 5.052746725082398,
      "epoch": 4.09462055715658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034199382795498015,
      "loss": 4.5231,
      "mean_token_accuracy": 0.264261919260025,
      "num_tokens": 97736052.0,
      "step": 42625
    },
    {
      "entropy": 5.088389825820923,
      "epoch": 4.095100864553314,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034196109859468325,
      "loss": 4.5429,
      "mean_token_accuracy": 0.26266554594039915,
      "num_tokens": 97747248.0,
      "step": 42630
    },
    {
      "entropy": 5.131372737884521,
      "epoch": 4.095581171950048,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003419283676797485,
      "loss": 4.6111,
      "mean_token_accuracy": 0.2561540067195892,
      "num_tokens": 97757906.0,
      "step": 42635
    },
    {
      "entropy": 5.0563514709472654,
      "epoch": 4.0960614793467816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034189563521093587,
      "loss": 4.5286,
      "mean_token_accuracy": 0.2656946375966072,
      "num_tokens": 97768490.0,
      "step": 42640
    },
    {
      "entropy": 5.057745695114136,
      "epoch": 4.096541786743516,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003418629011890051,
      "loss": 4.5387,
      "mean_token_accuracy": 0.26577805876731875,
      "num_tokens": 97780949.0,
      "step": 42645
    },
    {
      "entropy": 5.117424440383911,
      "epoch": 4.09702209414025,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034183016561471644,
      "loss": 4.5567,
      "mean_token_accuracy": 0.25758529603481295,
      "num_tokens": 97792327.0,
      "step": 42650
    },
    {
      "entropy": 5.012349462509155,
      "epoch": 4.097502401536984,
      "grad_norm": 1.0,
      "learning_rate": 0.00034179742848882967,
      "loss": 4.5003,
      "mean_token_accuracy": 0.2674847304821014,
      "num_tokens": 97804225.0,
      "step": 42655
    },
    {
      "entropy": 4.947874927520752,
      "epoch": 4.097982708933718,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034176468981210494,
      "loss": 4.5161,
      "mean_token_accuracy": 0.2657447725534439,
      "num_tokens": 97816663.0,
      "step": 42660
    },
    {
      "entropy": 5.014748191833496,
      "epoch": 4.098463016330451,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003417319495853025,
      "loss": 4.4988,
      "mean_token_accuracy": 0.2664140984416008,
      "num_tokens": 97829350.0,
      "step": 42665
    },
    {
      "entropy": 5.139909362792968,
      "epoch": 4.098943323727186,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003416992078091823,
      "loss": 4.5871,
      "mean_token_accuracy": 0.2584946468472481,
      "num_tokens": 97840596.0,
      "step": 42670
    },
    {
      "entropy": 5.0004112243652346,
      "epoch": 4.099423631123919,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003416664644845044,
      "loss": 4.4298,
      "mean_token_accuracy": 0.2728751912713051,
      "num_tokens": 97851496.0,
      "step": 42675
    },
    {
      "entropy": 5.002013301849365,
      "epoch": 4.0999039385206535,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00034163371961202935,
      "loss": 4.4918,
      "mean_token_accuracy": 0.2724514678120613,
      "num_tokens": 97862739.0,
      "step": 42680
    },
    {
      "entropy": 5.053995275497437,
      "epoch": 4.100384245917387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034160097319251714,
      "loss": 4.4721,
      "mean_token_accuracy": 0.26620172411203386,
      "num_tokens": 97874453.0,
      "step": 42685
    },
    {
      "entropy": 5.078334188461303,
      "epoch": 4.100864553314121,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034156822522672805,
      "loss": 4.584,
      "mean_token_accuracy": 0.2596738949418068,
      "num_tokens": 97885578.0,
      "step": 42690
    },
    {
      "entropy": 5.040371513366699,
      "epoch": 4.101344860710855,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003415354757154225,
      "loss": 4.4936,
      "mean_token_accuracy": 0.2650184080004692,
      "num_tokens": 97895801.0,
      "step": 42695
    },
    {
      "entropy": 5.0540306091308596,
      "epoch": 4.101825168107589,
      "grad_norm": 1.0,
      "learning_rate": 0.00034150272465936066,
      "loss": 4.4714,
      "mean_token_accuracy": 0.2631304159760475,
      "num_tokens": 97906948.0,
      "step": 42700
    },
    {
      "entropy": 4.989506435394287,
      "epoch": 4.1023054755043225,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034146997205930307,
      "loss": 4.4947,
      "mean_token_accuracy": 0.267454594373703,
      "num_tokens": 97917785.0,
      "step": 42705
    },
    {
      "entropy": 5.056091928482056,
      "epoch": 4.102785782901057,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003414372179160101,
      "loss": 4.5699,
      "mean_token_accuracy": 0.26440054923295975,
      "num_tokens": 97929098.0,
      "step": 42710
    },
    {
      "entropy": 5.141332149505615,
      "epoch": 4.10326609029779,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003414044622302424,
      "loss": 4.5665,
      "mean_token_accuracy": 0.25599194765090943,
      "num_tokens": 97940375.0,
      "step": 42715
    },
    {
      "entropy": 5.070445775985718,
      "epoch": 4.1037463976945245,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034137170500276004,
      "loss": 4.5935,
      "mean_token_accuracy": 0.26362370103597643,
      "num_tokens": 97951151.0,
      "step": 42720
    },
    {
      "entropy": 5.126795387268066,
      "epoch": 4.104226705091258,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003413389462343238,
      "loss": 4.5835,
      "mean_token_accuracy": 0.26329986155033114,
      "num_tokens": 97961746.0,
      "step": 42725
    },
    {
      "entropy": 5.059521341323853,
      "epoch": 4.104707012487992,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003413061859256943,
      "loss": 4.4564,
      "mean_token_accuracy": 0.26622880846261976,
      "num_tokens": 97972332.0,
      "step": 42730
    },
    {
      "entropy": 5.049635505676269,
      "epoch": 4.105187319884727,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034127342407763196,
      "loss": 4.6206,
      "mean_token_accuracy": 0.2607726246118546,
      "num_tokens": 97983294.0,
      "step": 42735
    },
    {
      "entropy": 5.076466464996338,
      "epoch": 4.10566762728146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003412406606908976,
      "loss": 4.5475,
      "mean_token_accuracy": 0.2624126210808754,
      "num_tokens": 97995617.0,
      "step": 42740
    },
    {
      "entropy": 5.111734819412232,
      "epoch": 4.106147934678194,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003412078957662517,
      "loss": 4.6264,
      "mean_token_accuracy": 0.2609729677438736,
      "num_tokens": 98007182.0,
      "step": 42745
    },
    {
      "entropy": 5.047837543487549,
      "epoch": 4.106628242074928,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034117512930445504,
      "loss": 4.5325,
      "mean_token_accuracy": 0.25957913100719454,
      "num_tokens": 98019005.0,
      "step": 42750
    },
    {
      "entropy": 5.107386589050293,
      "epoch": 4.107108549471662,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034114236130626855,
      "loss": 4.5984,
      "mean_token_accuracy": 0.2537531167268753,
      "num_tokens": 98030377.0,
      "step": 42755
    },
    {
      "entropy": 5.068209075927735,
      "epoch": 4.107588856868396,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003411095917724527,
      "loss": 4.5407,
      "mean_token_accuracy": 0.26670600175857545,
      "num_tokens": 98042327.0,
      "step": 42760
    },
    {
      "entropy": 5.012467575073242,
      "epoch": 4.10806916426513,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034107682070376845,
      "loss": 4.4944,
      "mean_token_accuracy": 0.26568332612514495,
      "num_tokens": 98054014.0,
      "step": 42765
    },
    {
      "entropy": 5.023444938659668,
      "epoch": 4.108549471661863,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034104404810097663,
      "loss": 4.5225,
      "mean_token_accuracy": 0.2683302938938141,
      "num_tokens": 98064398.0,
      "step": 42770
    },
    {
      "entropy": 5.007653284072876,
      "epoch": 4.109029779058598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003410112739648381,
      "loss": 4.5066,
      "mean_token_accuracy": 0.27455252408981323,
      "num_tokens": 98075061.0,
      "step": 42775
    },
    {
      "entropy": 5.140065479278564,
      "epoch": 4.109510086455331,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034097849829611387,
      "loss": 4.619,
      "mean_token_accuracy": 0.2514939472079277,
      "num_tokens": 98086761.0,
      "step": 42780
    },
    {
      "entropy": 5.136370229721069,
      "epoch": 4.109990393852065,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00034094572109556477,
      "loss": 4.59,
      "mean_token_accuracy": 0.2603756576776505,
      "num_tokens": 98098683.0,
      "step": 42785
    },
    {
      "entropy": 5.111892747879028,
      "epoch": 4.110470701248799,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034091294236395186,
      "loss": 4.568,
      "mean_token_accuracy": 0.26500487625598906,
      "num_tokens": 98110312.0,
      "step": 42790
    },
    {
      "entropy": 4.994075059890747,
      "epoch": 4.110951008645533,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034088016210203617,
      "loss": 4.5125,
      "mean_token_accuracy": 0.25991438925266264,
      "num_tokens": 98122351.0,
      "step": 42795
    },
    {
      "entropy": 5.030589056015015,
      "epoch": 4.111431316042267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003408473803105788,
      "loss": 4.5038,
      "mean_token_accuracy": 0.2710335448384285,
      "num_tokens": 98132938.0,
      "step": 42800
    },
    {
      "entropy": 5.054779958724976,
      "epoch": 4.111911623439001,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003408145969903406,
      "loss": 4.5784,
      "mean_token_accuracy": 0.26171737909317017,
      "num_tokens": 98143785.0,
      "step": 42805
    },
    {
      "entropy": 5.110040140151978,
      "epoch": 4.112391930835735,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034078181214208304,
      "loss": 4.5616,
      "mean_token_accuracy": 0.24994735717773436,
      "num_tokens": 98154539.0,
      "step": 42810
    },
    {
      "entropy": 5.052552843093872,
      "epoch": 4.112872238232469,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003407490257665671,
      "loss": 4.4963,
      "mean_token_accuracy": 0.275269603729248,
      "num_tokens": 98166332.0,
      "step": 42815
    },
    {
      "entropy": 5.186036729812622,
      "epoch": 4.113352545629203,
      "grad_norm": 1.015625,
      "learning_rate": 0.000340716237864554,
      "loss": 4.6712,
      "mean_token_accuracy": 0.2568313792347908,
      "num_tokens": 98177536.0,
      "step": 42820
    },
    {
      "entropy": 5.158657264709473,
      "epoch": 4.1138328530259365,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034068344843680493,
      "loss": 4.6815,
      "mean_token_accuracy": 0.2532579228281975,
      "num_tokens": 98188482.0,
      "step": 42825
    },
    {
      "entropy": 5.145281505584717,
      "epoch": 4.114313160422671,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034065065748408135,
      "loss": 4.5979,
      "mean_token_accuracy": 0.25885251313447954,
      "num_tokens": 98200615.0,
      "step": 42830
    },
    {
      "entropy": 5.093455791473389,
      "epoch": 4.114793467819404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034061786500714434,
      "loss": 4.545,
      "mean_token_accuracy": 0.26340720653533933,
      "num_tokens": 98212984.0,
      "step": 42835
    },
    {
      "entropy": 5.008515119552612,
      "epoch": 4.115273775216139,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034058507100675545,
      "loss": 4.4352,
      "mean_token_accuracy": 0.26866900473833083,
      "num_tokens": 98224389.0,
      "step": 42840
    },
    {
      "entropy": 5.0310094356536865,
      "epoch": 4.115754082612872,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034055227548367595,
      "loss": 4.5919,
      "mean_token_accuracy": 0.2594778284430504,
      "num_tokens": 98235266.0,
      "step": 42845
    },
    {
      "entropy": 5.028916883468628,
      "epoch": 4.116234390009606,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003405194784386672,
      "loss": 4.51,
      "mean_token_accuracy": 0.2687856301665306,
      "num_tokens": 98246127.0,
      "step": 42850
    },
    {
      "entropy": 5.061490726470947,
      "epoch": 4.11671469740634,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003404866798724907,
      "loss": 4.5399,
      "mean_token_accuracy": 0.25840146839618683,
      "num_tokens": 98258642.0,
      "step": 42855
    },
    {
      "entropy": 5.123851776123047,
      "epoch": 4.117195004803074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000340453879785908,
      "loss": 4.5725,
      "mean_token_accuracy": 0.266925060749054,
      "num_tokens": 98270194.0,
      "step": 42860
    },
    {
      "entropy": 5.121744585037232,
      "epoch": 4.1176753121998075,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034042107817968054,
      "loss": 4.6191,
      "mean_token_accuracy": 0.2621067687869072,
      "num_tokens": 98281512.0,
      "step": 42865
    },
    {
      "entropy": 5.092948293685913,
      "epoch": 4.118155619596542,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034038827505456995,
      "loss": 4.5875,
      "mean_token_accuracy": 0.26315184980630874,
      "num_tokens": 98294097.0,
      "step": 42870
    },
    {
      "entropy": 5.122455883026123,
      "epoch": 4.118635926993275,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003403554704113378,
      "loss": 4.573,
      "mean_token_accuracy": 0.2523643270134926,
      "num_tokens": 98305897.0,
      "step": 42875
    },
    {
      "entropy": 5.028709888458252,
      "epoch": 4.11911623439001,
      "grad_norm": 1.0,
      "learning_rate": 0.0003403226642507456,
      "loss": 4.5225,
      "mean_token_accuracy": 0.26227116137742995,
      "num_tokens": 98316119.0,
      "step": 42880
    },
    {
      "entropy": 5.071589183807373,
      "epoch": 4.119596541786744,
      "grad_norm": 0.921875,
      "learning_rate": 0.00034028985657355516,
      "loss": 4.5961,
      "mean_token_accuracy": 0.2587287798523903,
      "num_tokens": 98326304.0,
      "step": 42885
    },
    {
      "entropy": 5.078380489349366,
      "epoch": 4.120076849183477,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003402570473805281,
      "loss": 4.532,
      "mean_token_accuracy": 0.2605093687772751,
      "num_tokens": 98337482.0,
      "step": 42890
    },
    {
      "entropy": 5.048219919204712,
      "epoch": 4.120557156580212,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003402242366724262,
      "loss": 4.4707,
      "mean_token_accuracy": 0.27054533213376997,
      "num_tokens": 98348391.0,
      "step": 42895
    },
    {
      "entropy": 5.058544492721557,
      "epoch": 4.121037463976945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034019142445001116,
      "loss": 4.5416,
      "mean_token_accuracy": 0.25993458181619644,
      "num_tokens": 98359361.0,
      "step": 42900
    },
    {
      "entropy": 5.030623960494995,
      "epoch": 4.1215177713736795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034015861071404484,
      "loss": 4.5056,
      "mean_token_accuracy": 0.2656349927186966,
      "num_tokens": 98370674.0,
      "step": 42905
    },
    {
      "entropy": 5.065752649307251,
      "epoch": 4.121998078770413,
      "grad_norm": 1.0,
      "learning_rate": 0.0003401257954652891,
      "loss": 4.5286,
      "mean_token_accuracy": 0.26361051946878433,
      "num_tokens": 98382669.0,
      "step": 42910
    },
    {
      "entropy": 5.007788610458374,
      "epoch": 4.122478386167147,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034009297870450574,
      "loss": 4.4683,
      "mean_token_accuracy": 0.2686392918229103,
      "num_tokens": 98393306.0,
      "step": 42915
    },
    {
      "entropy": 5.086452865600586,
      "epoch": 4.122958693563881,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003400601604324567,
      "loss": 4.5644,
      "mean_token_accuracy": 0.2702146634459496,
      "num_tokens": 98404570.0,
      "step": 42920
    },
    {
      "entropy": 5.057223463058472,
      "epoch": 4.123439000960615,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034002734064990386,
      "loss": 4.5385,
      "mean_token_accuracy": 0.2589630573987961,
      "num_tokens": 98415064.0,
      "step": 42925
    },
    {
      "entropy": 5.099735927581787,
      "epoch": 4.123919308357348,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003399945193576093,
      "loss": 4.5573,
      "mean_token_accuracy": 0.26265600621700286,
      "num_tokens": 98426035.0,
      "step": 42930
    },
    {
      "entropy": 5.133914947509766,
      "epoch": 4.124399615754083,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003399616965563351,
      "loss": 4.6626,
      "mean_token_accuracy": 0.2516149714589119,
      "num_tokens": 98437361.0,
      "step": 42935
    },
    {
      "entropy": 5.16773157119751,
      "epoch": 4.124879923150816,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003399288722468431,
      "loss": 4.6472,
      "mean_token_accuracy": 0.24617961943149566,
      "num_tokens": 98450703.0,
      "step": 42940
    },
    {
      "entropy": 5.148586845397949,
      "epoch": 4.1253602305475505,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003398960464298955,
      "loss": 4.6125,
      "mean_token_accuracy": 0.2589780956506729,
      "num_tokens": 98462625.0,
      "step": 42945
    },
    {
      "entropy": 5.089141893386841,
      "epoch": 4.125840537944284,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003398632191062544,
      "loss": 4.6263,
      "mean_token_accuracy": 0.25592027604579926,
      "num_tokens": 98474541.0,
      "step": 42950
    },
    {
      "entropy": 5.098961305618286,
      "epoch": 4.126320845341018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000339830390276682,
      "loss": 4.6001,
      "mean_token_accuracy": 0.2624612390995026,
      "num_tokens": 98486200.0,
      "step": 42955
    },
    {
      "entropy": 5.1144898414611815,
      "epoch": 4.126801152737752,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033979755994194033,
      "loss": 4.5709,
      "mean_token_accuracy": 0.264832529425621,
      "num_tokens": 98498792.0,
      "step": 42960
    },
    {
      "entropy": 5.051899814605713,
      "epoch": 4.127281460134486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003397647281027919,
      "loss": 4.545,
      "mean_token_accuracy": 0.2561636880040169,
      "num_tokens": 98511188.0,
      "step": 42965
    },
    {
      "entropy": 5.10940375328064,
      "epoch": 4.12776176753122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003397318947599987,
      "loss": 4.5234,
      "mean_token_accuracy": 0.26445336639881134,
      "num_tokens": 98522265.0,
      "step": 42970
    },
    {
      "entropy": 5.028534078598023,
      "epoch": 4.128242074927954,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033969905991432307,
      "loss": 4.4818,
      "mean_token_accuracy": 0.2706770047545433,
      "num_tokens": 98533595.0,
      "step": 42975
    },
    {
      "entropy": 5.075533723831176,
      "epoch": 4.128722382324688,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033966622356652746,
      "loss": 4.5739,
      "mean_token_accuracy": 0.2648451551795006,
      "num_tokens": 98544879.0,
      "step": 42980
    },
    {
      "entropy": 5.0021100521087645,
      "epoch": 4.129202689721422,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003396333857173741,
      "loss": 4.4788,
      "mean_token_accuracy": 0.2695299476385117,
      "num_tokens": 98556865.0,
      "step": 42985
    },
    {
      "entropy": 5.0372556209564205,
      "epoch": 4.129682997118156,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003396005463676255,
      "loss": 4.5583,
      "mean_token_accuracy": 0.2594573274254799,
      "num_tokens": 98569093.0,
      "step": 42990
    },
    {
      "entropy": 5.012969875335694,
      "epoch": 4.130163304514889,
      "grad_norm": 1.03125,
      "learning_rate": 0.000339567705518044,
      "loss": 4.3953,
      "mean_token_accuracy": 0.2778189301490784,
      "num_tokens": 98581011.0,
      "step": 42995
    },
    {
      "entropy": 5.014597415924072,
      "epoch": 4.130643611911624,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033953486316939214,
      "loss": 4.5314,
      "mean_token_accuracy": 0.26381932348012926,
      "num_tokens": 98592696.0,
      "step": 43000
    },
    {
      "entropy": 5.059126329421997,
      "epoch": 4.131123919308357,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033950201932243237,
      "loss": 4.6221,
      "mean_token_accuracy": 0.2482265681028366,
      "num_tokens": 98604763.0,
      "step": 43005
    },
    {
      "entropy": 5.15454249382019,
      "epoch": 4.131604226705091,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033946917397792724,
      "loss": 4.6461,
      "mean_token_accuracy": 0.25058896094560623,
      "num_tokens": 98616795.0,
      "step": 43010
    },
    {
      "entropy": 5.037417888641357,
      "epoch": 4.132084534101825,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003394363271366393,
      "loss": 4.4298,
      "mean_token_accuracy": 0.2731252357363701,
      "num_tokens": 98626535.0,
      "step": 43015
    },
    {
      "entropy": 4.981300115585327,
      "epoch": 4.132564841498559,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003394034787993312,
      "loss": 4.4765,
      "mean_token_accuracy": 0.26763445436954497,
      "num_tokens": 98637401.0,
      "step": 43020
    },
    {
      "entropy": 5.080830812454224,
      "epoch": 4.133045148895293,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033937062896676563,
      "loss": 4.6154,
      "mean_token_accuracy": 0.26248856335878373,
      "num_tokens": 98648905.0,
      "step": 43025
    },
    {
      "entropy": 5.1063636302947994,
      "epoch": 4.133525456292027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033933777763970506,
      "loss": 4.6152,
      "mean_token_accuracy": 0.25363693237304685,
      "num_tokens": 98661246.0,
      "step": 43030
    },
    {
      "entropy": 5.042243623733521,
      "epoch": 4.13400576368876,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003393049248189124,
      "loss": 4.5108,
      "mean_token_accuracy": 0.2693849042057991,
      "num_tokens": 98671837.0,
      "step": 43035
    },
    {
      "entropy": 5.09065203666687,
      "epoch": 4.134486071085495,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003392720705051504,
      "loss": 4.5596,
      "mean_token_accuracy": 0.2648086830973625,
      "num_tokens": 98683364.0,
      "step": 43040
    },
    {
      "entropy": 5.143191528320313,
      "epoch": 4.134966378482229,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033923921469918174,
      "loss": 4.6706,
      "mean_token_accuracy": 0.25287363678216934,
      "num_tokens": 98695412.0,
      "step": 43045
    },
    {
      "entropy": 5.095544004440308,
      "epoch": 4.1354466858789625,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003392063574017692,
      "loss": 4.5797,
      "mean_token_accuracy": 0.26653775870800017,
      "num_tokens": 98707114.0,
      "step": 43050
    },
    {
      "entropy": 5.1034016609191895,
      "epoch": 4.135926993275697,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033917349861367575,
      "loss": 4.5445,
      "mean_token_accuracy": 0.2654679849743843,
      "num_tokens": 98719524.0,
      "step": 43055
    },
    {
      "entropy": 5.06990613937378,
      "epoch": 4.13640730067243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003391406383356642,
      "loss": 4.5582,
      "mean_token_accuracy": 0.2560317635536194,
      "num_tokens": 98731311.0,
      "step": 43060
    },
    {
      "entropy": 4.942803812026978,
      "epoch": 4.136887608069165,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033910777656849745,
      "loss": 4.4289,
      "mean_token_accuracy": 0.2665160208940506,
      "num_tokens": 98743095.0,
      "step": 43065
    },
    {
      "entropy": 5.034766101837159,
      "epoch": 4.137367915465898,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003390749133129386,
      "loss": 4.484,
      "mean_token_accuracy": 0.26287697702646257,
      "num_tokens": 98754224.0,
      "step": 43070
    },
    {
      "entropy": 5.189194583892823,
      "epoch": 4.137848222862632,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003390420485697504,
      "loss": 4.6749,
      "mean_token_accuracy": 0.25006477236747743,
      "num_tokens": 98765825.0,
      "step": 43075
    },
    {
      "entropy": 5.090351867675781,
      "epoch": 4.138328530259366,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033900918233969606,
      "loss": 4.5559,
      "mean_token_accuracy": 0.253775979578495,
      "num_tokens": 98777091.0,
      "step": 43080
    },
    {
      "entropy": 5.045848321914673,
      "epoch": 4.1388088376561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033897631462353854,
      "loss": 4.4951,
      "mean_token_accuracy": 0.2620090126991272,
      "num_tokens": 98788354.0,
      "step": 43085
    },
    {
      "entropy": 5.084249544143677,
      "epoch": 4.1392891450528335,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000338943445422041,
      "loss": 4.562,
      "mean_token_accuracy": 0.26244322061538694,
      "num_tokens": 98799385.0,
      "step": 43090
    },
    {
      "entropy": 5.148423290252685,
      "epoch": 4.139769452449568,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033891057473596655,
      "loss": 4.653,
      "mean_token_accuracy": 0.2588789016008377,
      "num_tokens": 98810719.0,
      "step": 43095
    },
    {
      "entropy": 5.03000659942627,
      "epoch": 4.140249759846301,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003388777025660783,
      "loss": 4.469,
      "mean_token_accuracy": 0.2657980605959892,
      "num_tokens": 98822451.0,
      "step": 43100
    },
    {
      "entropy": 5.15441403388977,
      "epoch": 4.140730067243036,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003388448289131395,
      "loss": 4.5793,
      "mean_token_accuracy": 0.2584465011954308,
      "num_tokens": 98833562.0,
      "step": 43105
    },
    {
      "entropy": 5.135707378387451,
      "epoch": 4.141210374639769,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033881195377791337,
      "loss": 4.5723,
      "mean_token_accuracy": 0.25540262162685395,
      "num_tokens": 98844591.0,
      "step": 43110
    },
    {
      "entropy": 5.121141242980957,
      "epoch": 4.141690682036503,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033877907716116314,
      "loss": 4.6332,
      "mean_token_accuracy": 0.2505588471889496,
      "num_tokens": 98856505.0,
      "step": 43115
    },
    {
      "entropy": 5.0792440414428714,
      "epoch": 4.142170989433238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000338746199063652,
      "loss": 4.4798,
      "mean_token_accuracy": 0.27329216003417967,
      "num_tokens": 98867762.0,
      "step": 43120
    },
    {
      "entropy": 5.142489004135132,
      "epoch": 4.142651296829971,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003387133194861435,
      "loss": 4.6322,
      "mean_token_accuracy": 0.2597357019782066,
      "num_tokens": 98879608.0,
      "step": 43125
    },
    {
      "entropy": 5.091566896438598,
      "epoch": 4.1431316042267055,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00033868043842940097,
      "loss": 4.5708,
      "mean_token_accuracy": 0.2610999748110771,
      "num_tokens": 98892367.0,
      "step": 43130
    },
    {
      "entropy": 5.069744062423706,
      "epoch": 4.143611911623439,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033864755589418764,
      "loss": 4.5881,
      "mean_token_accuracy": 0.2605648174881935,
      "num_tokens": 98905207.0,
      "step": 43135
    },
    {
      "entropy": 5.0666711807250975,
      "epoch": 4.144092219020173,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033861467188126714,
      "loss": 4.5468,
      "mean_token_accuracy": 0.26304484009742735,
      "num_tokens": 98916126.0,
      "step": 43140
    },
    {
      "entropy": 5.067850399017334,
      "epoch": 4.144572526416907,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033858178639140286,
      "loss": 4.5253,
      "mean_token_accuracy": 0.25946626961231234,
      "num_tokens": 98927575.0,
      "step": 43145
    },
    {
      "entropy": 5.085066413879394,
      "epoch": 4.145052833813641,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003385488994253583,
      "loss": 4.5602,
      "mean_token_accuracy": 0.2624214440584183,
      "num_tokens": 98939563.0,
      "step": 43150
    },
    {
      "entropy": 5.205444622039795,
      "epoch": 4.145533141210374,
      "grad_norm": 1.03125,
      "learning_rate": 0.000338516010983897,
      "loss": 4.7054,
      "mean_token_accuracy": 0.2491161644458771,
      "num_tokens": 98950308.0,
      "step": 43155
    },
    {
      "entropy": 5.095419263839721,
      "epoch": 4.146013448607109,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033848312106778247,
      "loss": 4.5027,
      "mean_token_accuracy": 0.2636447697877884,
      "num_tokens": 98961366.0,
      "step": 43160
    },
    {
      "entropy": 4.969708013534546,
      "epoch": 4.146493756003842,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033845022967777833,
      "loss": 4.4425,
      "mean_token_accuracy": 0.2702391520142555,
      "num_tokens": 98971855.0,
      "step": 43165
    },
    {
      "entropy": 5.120429277420044,
      "epoch": 4.1469740634005765,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033841733681464837,
      "loss": 4.5786,
      "mean_token_accuracy": 0.25759520679712294,
      "num_tokens": 98984592.0,
      "step": 43170
    },
    {
      "entropy": 5.114663887023926,
      "epoch": 4.14745437079731,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033838444247915613,
      "loss": 4.5923,
      "mean_token_accuracy": 0.2595915973186493,
      "num_tokens": 98996319.0,
      "step": 43175
    },
    {
      "entropy": 5.084191465377808,
      "epoch": 4.147934678194044,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003383515466720653,
      "loss": 4.5505,
      "mean_token_accuracy": 0.26128833144903185,
      "num_tokens": 99007990.0,
      "step": 43180
    },
    {
      "entropy": 5.046680498123169,
      "epoch": 4.148414985590778,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033831864939413975,
      "loss": 4.5627,
      "mean_token_accuracy": 0.2622947797179222,
      "num_tokens": 99019921.0,
      "step": 43185
    },
    {
      "entropy": 5.156048679351807,
      "epoch": 4.148895292987512,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003382857506461432,
      "loss": 4.6206,
      "mean_token_accuracy": 0.25329324454069135,
      "num_tokens": 99031345.0,
      "step": 43190
    },
    {
      "entropy": 5.128985500335693,
      "epoch": 4.149375600384246,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00033825285042883937,
      "loss": 4.5914,
      "mean_token_accuracy": 0.25820703953504565,
      "num_tokens": 99042928.0,
      "step": 43195
    },
    {
      "entropy": 4.982667922973633,
      "epoch": 4.14985590778098,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033821994874299216,
      "loss": 4.4729,
      "mean_token_accuracy": 0.2682273268699646,
      "num_tokens": 99054808.0,
      "step": 43200
    },
    {
      "entropy": 4.9482067108154295,
      "epoch": 4.150336215177714,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003381870455893655,
      "loss": 4.4921,
      "mean_token_accuracy": 0.2733342170715332,
      "num_tokens": 99068231.0,
      "step": 43205
    },
    {
      "entropy": 5.127028560638427,
      "epoch": 4.1508165225744476,
      "grad_norm": 1.0,
      "learning_rate": 0.00033815414096872326,
      "loss": 4.6472,
      "mean_token_accuracy": 0.25479676127433776,
      "num_tokens": 99080329.0,
      "step": 43210
    },
    {
      "entropy": 5.1644186019897464,
      "epoch": 4.151296829971182,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003381212348818294,
      "loss": 4.6166,
      "mean_token_accuracy": 0.2566849261522293,
      "num_tokens": 99091481.0,
      "step": 43215
    },
    {
      "entropy": 5.168700695037842,
      "epoch": 4.151777137367915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003380883273294479,
      "loss": 4.671,
      "mean_token_accuracy": 0.2573960915207863,
      "num_tokens": 99101328.0,
      "step": 43220
    },
    {
      "entropy": 5.111501550674438,
      "epoch": 4.15225744476465,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033805541831234275,
      "loss": 4.5601,
      "mean_token_accuracy": 0.25452041029930117,
      "num_tokens": 99112574.0,
      "step": 43225
    },
    {
      "entropy": 5.140850400924682,
      "epoch": 4.152737752161383,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000338022507831278,
      "loss": 4.6185,
      "mean_token_accuracy": 0.254491513967514,
      "num_tokens": 99124325.0,
      "step": 43230
    },
    {
      "entropy": 5.027794361114502,
      "epoch": 4.153218059558117,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033798959588701786,
      "loss": 4.5019,
      "mean_token_accuracy": 0.26651804596185685,
      "num_tokens": 99136198.0,
      "step": 43235
    },
    {
      "entropy": 5.1068236351013185,
      "epoch": 4.153698366954851,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003379566824803263,
      "loss": 4.6496,
      "mean_token_accuracy": 0.2538469836115837,
      "num_tokens": 99147860.0,
      "step": 43240
    },
    {
      "entropy": 5.110669755935669,
      "epoch": 4.154178674351585,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003379237676119675,
      "loss": 4.5559,
      "mean_token_accuracy": 0.25867956727743147,
      "num_tokens": 99159791.0,
      "step": 43245
    },
    {
      "entropy": 5.135907936096191,
      "epoch": 4.154658981748319,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003378908512827057,
      "loss": 4.5967,
      "mean_token_accuracy": 0.2593014001846313,
      "num_tokens": 99170455.0,
      "step": 43250
    },
    {
      "entropy": 5.047655630111694,
      "epoch": 4.155139289145053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033785793349330495,
      "loss": 4.4678,
      "mean_token_accuracy": 0.27088149785995486,
      "num_tokens": 99181926.0,
      "step": 43255
    },
    {
      "entropy": 5.0324376106262205,
      "epoch": 4.155619596541786,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033782501424452975,
      "loss": 4.5539,
      "mean_token_accuracy": 0.26703798919916155,
      "num_tokens": 99194331.0,
      "step": 43260
    },
    {
      "entropy": 5.131289052963257,
      "epoch": 4.156099903938521,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033779209353714424,
      "loss": 4.6842,
      "mean_token_accuracy": 0.2458409145474434,
      "num_tokens": 99206231.0,
      "step": 43265
    },
    {
      "entropy": 5.114633226394654,
      "epoch": 4.156580211335255,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003377591713719128,
      "loss": 4.5824,
      "mean_token_accuracy": 0.26459487676620486,
      "num_tokens": 99216632.0,
      "step": 43270
    },
    {
      "entropy": 5.183669471740723,
      "epoch": 4.1570605187319885,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003377262477495997,
      "loss": 4.644,
      "mean_token_accuracy": 0.2567704990506172,
      "num_tokens": 99228687.0,
      "step": 43275
    },
    {
      "entropy": 5.0847211360931395,
      "epoch": 4.157540826128723,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003376933226709694,
      "loss": 4.5443,
      "mean_token_accuracy": 0.26349322497844696,
      "num_tokens": 99240157.0,
      "step": 43280
    },
    {
      "entropy": 5.087152338027954,
      "epoch": 4.158021133525456,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033766039613678634,
      "loss": 4.6137,
      "mean_token_accuracy": 0.2572331607341766,
      "num_tokens": 99252061.0,
      "step": 43285
    },
    {
      "entropy": 5.068025922775268,
      "epoch": 4.1585014409221905,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00033762746814781496,
      "loss": 4.573,
      "mean_token_accuracy": 0.25973335951566695,
      "num_tokens": 99263743.0,
      "step": 43290
    },
    {
      "entropy": 5.0830577373504635,
      "epoch": 4.158981748318924,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003375945387048197,
      "loss": 4.49,
      "mean_token_accuracy": 0.2677479237318039,
      "num_tokens": 99275176.0,
      "step": 43295
    },
    {
      "entropy": 5.078335332870483,
      "epoch": 4.159462055715658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003375616078085651,
      "loss": 4.5907,
      "mean_token_accuracy": 0.25166025310754775,
      "num_tokens": 99286654.0,
      "step": 43300
    },
    {
      "entropy": 5.00983772277832,
      "epoch": 4.159942363112392,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033752867545981576,
      "loss": 4.5216,
      "mean_token_accuracy": 0.26275093108415604,
      "num_tokens": 99298118.0,
      "step": 43305
    },
    {
      "entropy": 5.0509934425354,
      "epoch": 4.160422670509126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033749574165933626,
      "loss": 4.5491,
      "mean_token_accuracy": 0.25784042179584504,
      "num_tokens": 99309831.0,
      "step": 43310
    },
    {
      "entropy": 5.206718254089355,
      "epoch": 4.1609029779058595,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003374628064078912,
      "loss": 4.6616,
      "mean_token_accuracy": 0.255277743935585,
      "num_tokens": 99321691.0,
      "step": 43315
    },
    {
      "entropy": 5.0918989181518555,
      "epoch": 4.161383285302594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033742986970624526,
      "loss": 4.5608,
      "mean_token_accuracy": 0.2653098374605179,
      "num_tokens": 99333452.0,
      "step": 43320
    },
    {
      "entropy": 5.046872758865357,
      "epoch": 4.161863592699327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033739693155516315,
      "loss": 4.4499,
      "mean_token_accuracy": 0.2691402181982994,
      "num_tokens": 99344594.0,
      "step": 43325
    },
    {
      "entropy": 5.097774934768677,
      "epoch": 4.162343900096062,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033736399195540955,
      "loss": 4.6107,
      "mean_token_accuracy": 0.2595031812787056,
      "num_tokens": 99356029.0,
      "step": 43330
    },
    {
      "entropy": 5.048937129974365,
      "epoch": 4.162824207492795,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033733105090774917,
      "loss": 4.5856,
      "mean_token_accuracy": 0.26450394093990326,
      "num_tokens": 99368752.0,
      "step": 43335
    },
    {
      "entropy": 5.055762910842896,
      "epoch": 4.163304514889529,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000337298108412947,
      "loss": 4.5329,
      "mean_token_accuracy": 0.2640812173485756,
      "num_tokens": 99380770.0,
      "step": 43340
    },
    {
      "entropy": 5.109798336029053,
      "epoch": 4.163784822286263,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033726516447176764,
      "loss": 4.6021,
      "mean_token_accuracy": 0.25135222971439364,
      "num_tokens": 99392825.0,
      "step": 43345
    },
    {
      "entropy": 5.098224353790283,
      "epoch": 4.164265129682997,
      "grad_norm": 1.125,
      "learning_rate": 0.0003372322190849761,
      "loss": 4.6124,
      "mean_token_accuracy": 0.2535936594009399,
      "num_tokens": 99403770.0,
      "step": 43350
    },
    {
      "entropy": 5.0413871765136715,
      "epoch": 4.164745437079731,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003371992722533372,
      "loss": 4.5679,
      "mean_token_accuracy": 0.2644712969660759,
      "num_tokens": 99414628.0,
      "step": 43355
    },
    {
      "entropy": 5.099226379394532,
      "epoch": 4.165225744476465,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003371663239776159,
      "loss": 4.555,
      "mean_token_accuracy": 0.25941442847251894,
      "num_tokens": 99426009.0,
      "step": 43360
    },
    {
      "entropy": 5.007165193557739,
      "epoch": 4.165706051873199,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033713337425857713,
      "loss": 4.4272,
      "mean_token_accuracy": 0.27104564756155014,
      "num_tokens": 99436740.0,
      "step": 43365
    },
    {
      "entropy": 5.123434448242188,
      "epoch": 4.166186359269933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000337100423096986,
      "loss": 4.6264,
      "mean_token_accuracy": 0.24936665892601012,
      "num_tokens": 99449265.0,
      "step": 43370
    },
    {
      "entropy": 5.098297214508056,
      "epoch": 4.166666666666667,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003370674704936074,
      "loss": 4.5271,
      "mean_token_accuracy": 0.26776676177978515,
      "num_tokens": 99460591.0,
      "step": 43375
    },
    {
      "entropy": 5.057177734375,
      "epoch": 4.1671469740634,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033703451644920637,
      "loss": 4.541,
      "mean_token_accuracy": 0.26680720746517184,
      "num_tokens": 99470744.0,
      "step": 43380
    },
    {
      "entropy": 5.071495246887207,
      "epoch": 4.167627281460135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033700156096454816,
      "loss": 4.5431,
      "mean_token_accuracy": 0.2643969252705574,
      "num_tokens": 99482454.0,
      "step": 43385
    },
    {
      "entropy": 5.040865707397461,
      "epoch": 4.168107588856868,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033696860404039777,
      "loss": 4.5075,
      "mean_token_accuracy": 0.2690462008118629,
      "num_tokens": 99494216.0,
      "step": 43390
    },
    {
      "entropy": 5.166667413711548,
      "epoch": 4.1685878962536025,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033693564567752036,
      "loss": 4.6695,
      "mean_token_accuracy": 0.2470731183886528,
      "num_tokens": 99506207.0,
      "step": 43395
    },
    {
      "entropy": 5.1779262065887455,
      "epoch": 4.169068203650336,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003369026858766813,
      "loss": 4.6741,
      "mean_token_accuracy": 0.24630312472581864,
      "num_tokens": 99518155.0,
      "step": 43400
    },
    {
      "entropy": 5.0633673667907715,
      "epoch": 4.16954851104707,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033686972463864553,
      "loss": 4.5113,
      "mean_token_accuracy": 0.2666232600808144,
      "num_tokens": 99528655.0,
      "step": 43405
    },
    {
      "entropy": 5.122004795074463,
      "epoch": 4.170028818443804,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003368367619641785,
      "loss": 4.5645,
      "mean_token_accuracy": 0.2546557798981667,
      "num_tokens": 99540059.0,
      "step": 43410
    },
    {
      "entropy": 5.204217576980591,
      "epoch": 4.170509125840538,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003368037978540455,
      "loss": 4.6807,
      "mean_token_accuracy": 0.24819720387458802,
      "num_tokens": 99551486.0,
      "step": 43415
    },
    {
      "entropy": 5.059299468994141,
      "epoch": 4.1709894332372714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003367708323090119,
      "loss": 4.4793,
      "mean_token_accuracy": 0.26868111938238143,
      "num_tokens": 99562366.0,
      "step": 43420
    },
    {
      "entropy": 5.068058252334595,
      "epoch": 4.171469740634006,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000336737865329843,
      "loss": 4.5686,
      "mean_token_accuracy": 0.2589056044816971,
      "num_tokens": 99574702.0,
      "step": 43425
    },
    {
      "entropy": 5.088884782791138,
      "epoch": 4.17195004803074,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033670489691730405,
      "loss": 4.5579,
      "mean_token_accuracy": 0.2630153030157089,
      "num_tokens": 99587514.0,
      "step": 43430
    },
    {
      "entropy": 5.093131732940674,
      "epoch": 4.1724303554274735,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003366719270721607,
      "loss": 4.5586,
      "mean_token_accuracy": 0.25987991094589236,
      "num_tokens": 99598505.0,
      "step": 43435
    },
    {
      "entropy": 5.0107927322387695,
      "epoch": 4.172910662824208,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033663895579517835,
      "loss": 4.4497,
      "mean_token_accuracy": 0.2703227311372757,
      "num_tokens": 99609503.0,
      "step": 43440
    },
    {
      "entropy": 5.061689710617065,
      "epoch": 4.173390970220941,
      "grad_norm": 1.0,
      "learning_rate": 0.00033660598308712247,
      "loss": 4.5908,
      "mean_token_accuracy": 0.2548350840806961,
      "num_tokens": 99620353.0,
      "step": 43445
    },
    {
      "entropy": 5.036106109619141,
      "epoch": 4.173871277617676,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033657300894875857,
      "loss": 4.5073,
      "mean_token_accuracy": 0.2641583174467087,
      "num_tokens": 99631662.0,
      "step": 43450
    },
    {
      "entropy": 5.08406777381897,
      "epoch": 4.174351585014409,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033654003338085225,
      "loss": 4.5308,
      "mean_token_accuracy": 0.26477701514959334,
      "num_tokens": 99642860.0,
      "step": 43455
    },
    {
      "entropy": 5.044551086425781,
      "epoch": 4.174831892411143,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00033650705638416914,
      "loss": 4.508,
      "mean_token_accuracy": 0.2662366583943367,
      "num_tokens": 99654858.0,
      "step": 43460
    },
    {
      "entropy": 5.11461124420166,
      "epoch": 4.175312199807877,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003364740779594747,
      "loss": 4.581,
      "mean_token_accuracy": 0.26286386251449584,
      "num_tokens": 99666577.0,
      "step": 43465
    },
    {
      "entropy": 5.051496601104736,
      "epoch": 4.175792507204611,
      "grad_norm": 1.0,
      "learning_rate": 0.00033644109810753483,
      "loss": 4.5051,
      "mean_token_accuracy": 0.2607878789305687,
      "num_tokens": 99677301.0,
      "step": 43470
    },
    {
      "entropy": 5.05249252319336,
      "epoch": 4.176272814601345,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000336408116829115,
      "loss": 4.6243,
      "mean_token_accuracy": 0.2632339119911194,
      "num_tokens": 99688075.0,
      "step": 43475
    },
    {
      "entropy": 5.010852527618408,
      "epoch": 4.176753121998079,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00033637513412498117,
      "loss": 4.5434,
      "mean_token_accuracy": 0.26405146569013593,
      "num_tokens": 99700709.0,
      "step": 43480
    },
    {
      "entropy": 5.122352600097656,
      "epoch": 4.177233429394812,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003363421499958989,
      "loss": 4.582,
      "mean_token_accuracy": 0.26962532103061676,
      "num_tokens": 99712867.0,
      "step": 43485
    },
    {
      "entropy": 5.017931222915649,
      "epoch": 4.177713736791547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033630916444263407,
      "loss": 4.3949,
      "mean_token_accuracy": 0.26972639560699463,
      "num_tokens": 99723932.0,
      "step": 43490
    },
    {
      "entropy": 5.077942705154419,
      "epoch": 4.17819404418828,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003362761774659525,
      "loss": 4.585,
      "mean_token_accuracy": 0.258216692507267,
      "num_tokens": 99735377.0,
      "step": 43495
    },
    {
      "entropy": 5.055341148376465,
      "epoch": 4.178674351585014,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033624318906662,
      "loss": 4.5321,
      "mean_token_accuracy": 0.2660984888672829,
      "num_tokens": 99746803.0,
      "step": 43500
    },
    {
      "entropy": 5.157320785522461,
      "epoch": 4.179154658981748,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033621019924540255,
      "loss": 4.6992,
      "mean_token_accuracy": 0.2475883424282074,
      "num_tokens": 99760401.0,
      "step": 43505
    },
    {
      "entropy": 5.105210018157959,
      "epoch": 4.179634966378482,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000336177208003066,
      "loss": 4.5025,
      "mean_token_accuracy": 0.2657831460237503,
      "num_tokens": 99771774.0,
      "step": 43510
    },
    {
      "entropy": 4.992625427246094,
      "epoch": 4.1801152737752165,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003361442153403763,
      "loss": 4.4431,
      "mean_token_accuracy": 0.2634072318673134,
      "num_tokens": 99783719.0,
      "step": 43515
    },
    {
      "entropy": 5.071287155151367,
      "epoch": 4.18059558117195,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003361112212580995,
      "loss": 4.5123,
      "mean_token_accuracy": 0.2635758310556412,
      "num_tokens": 99794815.0,
      "step": 43520
    },
    {
      "entropy": 5.056607961654663,
      "epoch": 4.181075888568684,
      "grad_norm": 1.125,
      "learning_rate": 0.0003360782257570017,
      "loss": 4.5633,
      "mean_token_accuracy": 0.2580934017896652,
      "num_tokens": 99804660.0,
      "step": 43525
    },
    {
      "entropy": 5.130277442932129,
      "epoch": 4.181556195965418,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003360452288378488,
      "loss": 4.5866,
      "mean_token_accuracy": 0.2605630174279213,
      "num_tokens": 99815643.0,
      "step": 43530
    },
    {
      "entropy": 5.100850772857666,
      "epoch": 4.182036503362152,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033601223050140687,
      "loss": 4.5412,
      "mean_token_accuracy": 0.2576057255268097,
      "num_tokens": 99826485.0,
      "step": 43535
    },
    {
      "entropy": 5.0733717918396,
      "epoch": 4.1825168107588855,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003359792307484423,
      "loss": 4.5771,
      "mean_token_accuracy": 0.26609748750925066,
      "num_tokens": 99837401.0,
      "step": 43540
    },
    {
      "entropy": 5.0643068790435795,
      "epoch": 4.18299711815562,
      "grad_norm": 1.0,
      "learning_rate": 0.00033594622957972096,
      "loss": 4.6596,
      "mean_token_accuracy": 0.24637209475040436,
      "num_tokens": 99849080.0,
      "step": 43545
    },
    {
      "entropy": 5.14524211883545,
      "epoch": 4.183477425552353,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003359132269960091,
      "loss": 4.6238,
      "mean_token_accuracy": 0.249951733648777,
      "num_tokens": 99860603.0,
      "step": 43550
    },
    {
      "entropy": 5.1225138187408445,
      "epoch": 4.183957732949088,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000335880222998073,
      "loss": 4.58,
      "mean_token_accuracy": 0.25630970150232313,
      "num_tokens": 99872290.0,
      "step": 43555
    },
    {
      "entropy": 5.114331436157227,
      "epoch": 4.184438040345821,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003358472175866789,
      "loss": 4.4737,
      "mean_token_accuracy": 0.2755713552236557,
      "num_tokens": 99883593.0,
      "step": 43560
    },
    {
      "entropy": 5.034954643249511,
      "epoch": 4.184918347742555,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003358142107625931,
      "loss": 4.5234,
      "mean_token_accuracy": 0.2624052599072456,
      "num_tokens": 99893720.0,
      "step": 43565
    },
    {
      "entropy": 5.04785647392273,
      "epoch": 4.185398655139289,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003357812025265819,
      "loss": 4.5803,
      "mean_token_accuracy": 0.2604000300168991,
      "num_tokens": 99905572.0,
      "step": 43570
    },
    {
      "entropy": 5.034826231002808,
      "epoch": 4.185878962536023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033574819287941174,
      "loss": 4.5277,
      "mean_token_accuracy": 0.2571780174970627,
      "num_tokens": 99917323.0,
      "step": 43575
    },
    {
      "entropy": 5.082757234573364,
      "epoch": 4.1863592699327565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033571518182184886,
      "loss": 4.5266,
      "mean_token_accuracy": 0.2667724758386612,
      "num_tokens": 99928424.0,
      "step": 43580
    },
    {
      "entropy": 5.055783081054687,
      "epoch": 4.186839577329491,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033568216935465976,
      "loss": 4.5407,
      "mean_token_accuracy": 0.2651453673839569,
      "num_tokens": 99939552.0,
      "step": 43585
    },
    {
      "entropy": 5.1324786186218265,
      "epoch": 4.187319884726225,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003356491554786109,
      "loss": 4.642,
      "mean_token_accuracy": 0.25074547827243804,
      "num_tokens": 99950954.0,
      "step": 43590
    },
    {
      "entropy": 5.1839344024658205,
      "epoch": 4.187800192122959,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033561614019446867,
      "loss": 4.6901,
      "mean_token_accuracy": 0.24598041474819182,
      "num_tokens": 99961575.0,
      "step": 43595
    },
    {
      "entropy": 5.234807968139648,
      "epoch": 4.188280499519693,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00033558312350299963,
      "loss": 4.717,
      "mean_token_accuracy": 0.2446226254105568,
      "num_tokens": 99974585.0,
      "step": 43600
    },
    {
      "entropy": 5.080097627639771,
      "epoch": 4.188760806916426,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00033555010540497045,
      "loss": 4.5254,
      "mean_token_accuracy": 0.27075666189193726,
      "num_tokens": 99987377.0,
      "step": 43605
    },
    {
      "entropy": 5.1071093559265135,
      "epoch": 4.189241114313161,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003355170859011476,
      "loss": 4.4978,
      "mean_token_accuracy": 0.26543130427598954,
      "num_tokens": 99998320.0,
      "step": 43610
    },
    {
      "entropy": 5.029394292831421,
      "epoch": 4.189721421709894,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003354840649922977,
      "loss": 4.5829,
      "mean_token_accuracy": 0.26429990530014036,
      "num_tokens": 100008334.0,
      "step": 43615
    },
    {
      "entropy": 5.0137636184692385,
      "epoch": 4.1902017291066285,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003354510426791874,
      "loss": 4.5122,
      "mean_token_accuracy": 0.2651247143745422,
      "num_tokens": 100019607.0,
      "step": 43620
    },
    {
      "entropy": 5.0420496463775635,
      "epoch": 4.190682036503362,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033541801896258344,
      "loss": 4.4863,
      "mean_token_accuracy": 0.2640573397278786,
      "num_tokens": 100031695.0,
      "step": 43625
    },
    {
      "entropy": 5.080882406234741,
      "epoch": 4.191162343900096,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003353849938432524,
      "loss": 4.5561,
      "mean_token_accuracy": 0.26354463696479796,
      "num_tokens": 100043441.0,
      "step": 43630
    },
    {
      "entropy": 5.099695014953613,
      "epoch": 4.19164265129683,
      "grad_norm": 1.125,
      "learning_rate": 0.0003353519673219612,
      "loss": 4.5856,
      "mean_token_accuracy": 0.24989248514175416,
      "num_tokens": 100054306.0,
      "step": 43635
    },
    {
      "entropy": 5.026974678039551,
      "epoch": 4.192122958693564,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003353189393994765,
      "loss": 4.4736,
      "mean_token_accuracy": 0.27499869763851165,
      "num_tokens": 100065813.0,
      "step": 43640
    },
    {
      "entropy": 5.0717284202575685,
      "epoch": 4.192603266090297,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033528591007656516,
      "loss": 4.5863,
      "mean_token_accuracy": 0.25918800979852674,
      "num_tokens": 100075573.0,
      "step": 43645
    },
    {
      "entropy": 5.077979183197021,
      "epoch": 4.193083573487032,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003352528793539939,
      "loss": 4.5592,
      "mean_token_accuracy": 0.2540609300136566,
      "num_tokens": 100086744.0,
      "step": 43650
    },
    {
      "entropy": 5.040577840805054,
      "epoch": 4.193563880883765,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033521984723252973,
      "loss": 4.4814,
      "mean_token_accuracy": 0.2679770827293396,
      "num_tokens": 100098147.0,
      "step": 43655
    },
    {
      "entropy": 4.97122049331665,
      "epoch": 4.1940441882804995,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00033518681371293965,
      "loss": 4.4917,
      "mean_token_accuracy": 0.27052861750125884,
      "num_tokens": 100109968.0,
      "step": 43660
    },
    {
      "entropy": 5.04087986946106,
      "epoch": 4.194524495677234,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033515377879599035,
      "loss": 4.552,
      "mean_token_accuracy": 0.2704032674431801,
      "num_tokens": 100122712.0,
      "step": 43665
    },
    {
      "entropy": 5.074889850616455,
      "epoch": 4.195004803073967,
      "grad_norm": 1.0,
      "learning_rate": 0.00033512074248244895,
      "loss": 4.4734,
      "mean_token_accuracy": 0.26729782819747927,
      "num_tokens": 100134404.0,
      "step": 43670
    },
    {
      "entropy": 5.110729122161866,
      "epoch": 4.195485110470702,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003350877047730825,
      "loss": 4.5694,
      "mean_token_accuracy": 0.25819419473409655,
      "num_tokens": 100146449.0,
      "step": 43675
    },
    {
      "entropy": 5.062724351882935,
      "epoch": 4.195965417867435,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033505466566865783,
      "loss": 4.5443,
      "mean_token_accuracy": 0.2645687013864517,
      "num_tokens": 100157643.0,
      "step": 43680
    },
    {
      "entropy": 5.140062618255615,
      "epoch": 4.196445725264169,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003350216251699422,
      "loss": 4.6321,
      "mean_token_accuracy": 0.25599507838487623,
      "num_tokens": 100168420.0,
      "step": 43685
    },
    {
      "entropy": 5.048064231872559,
      "epoch": 4.196926032660903,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033498858327770276,
      "loss": 4.5432,
      "mean_token_accuracy": 0.26974464654922486,
      "num_tokens": 100178811.0,
      "step": 43690
    },
    {
      "entropy": 5.068032550811767,
      "epoch": 4.197406340057637,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033495553999270643,
      "loss": 4.474,
      "mean_token_accuracy": 0.2657276228070259,
      "num_tokens": 100189837.0,
      "step": 43695
    },
    {
      "entropy": 5.009793281555176,
      "epoch": 4.197886647454371,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033492249531572054,
      "loss": 4.5472,
      "mean_token_accuracy": 0.26551550775766375,
      "num_tokens": 100201319.0,
      "step": 43700
    },
    {
      "entropy": 5.117717599868774,
      "epoch": 4.198366954851105,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003348894492475123,
      "loss": 4.591,
      "mean_token_accuracy": 0.25641404688358305,
      "num_tokens": 100213540.0,
      "step": 43705
    },
    {
      "entropy": 5.036622714996338,
      "epoch": 4.198847262247838,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003348564017888488,
      "loss": 4.538,
      "mean_token_accuracy": 0.2675735056400299,
      "num_tokens": 100226360.0,
      "step": 43710
    },
    {
      "entropy": 5.023042917251587,
      "epoch": 4.199327569644573,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033482335294049735,
      "loss": 4.5611,
      "mean_token_accuracy": 0.26173949986696243,
      "num_tokens": 100237461.0,
      "step": 43715
    },
    {
      "entropy": 5.046018981933594,
      "epoch": 4.199807877041306,
      "grad_norm": 1.0,
      "learning_rate": 0.0003347903027032254,
      "loss": 4.5169,
      "mean_token_accuracy": 0.2661618575453758,
      "num_tokens": 100249710.0,
      "step": 43720
    },
    {
      "entropy": 5.0553586959838865,
      "epoch": 4.20028818443804,
      "grad_norm": 1.0,
      "learning_rate": 0.0003347572510778001,
      "loss": 4.428,
      "mean_token_accuracy": 0.2761304348707199,
      "num_tokens": 100261797.0,
      "step": 43725
    },
    {
      "entropy": 5.127534341812134,
      "epoch": 4.200768491834774,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003347241980649889,
      "loss": 4.6699,
      "mean_token_accuracy": 0.2533848717808723,
      "num_tokens": 100273361.0,
      "step": 43730
    },
    {
      "entropy": 5.076258659362793,
      "epoch": 4.201248799231508,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033469114366555913,
      "loss": 4.5592,
      "mean_token_accuracy": 0.2624511390924454,
      "num_tokens": 100284566.0,
      "step": 43735
    },
    {
      "entropy": 5.077702808380127,
      "epoch": 4.2017291066282425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033465808788027824,
      "loss": 4.5194,
      "mean_token_accuracy": 0.2633530765771866,
      "num_tokens": 100296390.0,
      "step": 43740
    },
    {
      "entropy": 5.130677223205566,
      "epoch": 4.202209414024976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033462503070991374,
      "loss": 4.6093,
      "mean_token_accuracy": 0.2589279547333717,
      "num_tokens": 100309661.0,
      "step": 43745
    },
    {
      "entropy": 5.073776149749756,
      "epoch": 4.20268972142171,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003345919721552331,
      "loss": 4.5362,
      "mean_token_accuracy": 0.26235702633857727,
      "num_tokens": 100320794.0,
      "step": 43750
    },
    {
      "entropy": 5.057130289077759,
      "epoch": 4.203170028818444,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033455891221700375,
      "loss": 4.5282,
      "mean_token_accuracy": 0.2689665541052818,
      "num_tokens": 100332543.0,
      "step": 43755
    },
    {
      "entropy": 4.987134027481079,
      "epoch": 4.203650336215178,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003345258508959933,
      "loss": 4.4763,
      "mean_token_accuracy": 0.27750977873802185,
      "num_tokens": 100343364.0,
      "step": 43760
    },
    {
      "entropy": 5.025041580200195,
      "epoch": 4.2041306436119115,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003344927881929695,
      "loss": 4.5124,
      "mean_token_accuracy": 0.2679272383451462,
      "num_tokens": 100354643.0,
      "step": 43765
    },
    {
      "entropy": 5.120384979248047,
      "epoch": 4.204610951008646,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033445972410869955,
      "loss": 4.6056,
      "mean_token_accuracy": 0.257839497923851,
      "num_tokens": 100365417.0,
      "step": 43770
    },
    {
      "entropy": 5.082676887512207,
      "epoch": 4.205091258405379,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003344266586439516,
      "loss": 4.5305,
      "mean_token_accuracy": 0.2597820207476616,
      "num_tokens": 100376050.0,
      "step": 43775
    },
    {
      "entropy": 5.093678665161133,
      "epoch": 4.2055715658021136,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033439359179949303,
      "loss": 4.5675,
      "mean_token_accuracy": 0.2637362390756607,
      "num_tokens": 100386740.0,
      "step": 43780
    },
    {
      "entropy": 5.035937547683716,
      "epoch": 4.206051873198847,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033436052357609157,
      "loss": 4.5566,
      "mean_token_accuracy": 0.26039221584796907,
      "num_tokens": 100398048.0,
      "step": 43785
    },
    {
      "entropy": 5.1419250011444095,
      "epoch": 4.206532180595581,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003343274539745151,
      "loss": 4.6576,
      "mean_token_accuracy": 0.25524075478315356,
      "num_tokens": 100408827.0,
      "step": 43790
    },
    {
      "entropy": 5.204173231124878,
      "epoch": 4.207012487992315,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003342943829955313,
      "loss": 4.6578,
      "mean_token_accuracy": 0.2531020134687424,
      "num_tokens": 100420326.0,
      "step": 43795
    },
    {
      "entropy": 5.1001105308532715,
      "epoch": 4.207492795389049,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033426131063990787,
      "loss": 4.538,
      "mean_token_accuracy": 0.2620605230331421,
      "num_tokens": 100432111.0,
      "step": 43800
    },
    {
      "entropy": 5.059865474700928,
      "epoch": 4.2079731027857825,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033422823690841286,
      "loss": 4.5842,
      "mean_token_accuracy": 0.2571199342608452,
      "num_tokens": 100443590.0,
      "step": 43805
    },
    {
      "entropy": 5.093248462677002,
      "epoch": 4.208453410182517,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003341951618018141,
      "loss": 4.6153,
      "mean_token_accuracy": 0.25908278226852416,
      "num_tokens": 100456281.0,
      "step": 43810
    },
    {
      "entropy": 5.131957626342773,
      "epoch": 4.208933717579251,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033416208532087937,
      "loss": 4.5726,
      "mean_token_accuracy": 0.25672959834337233,
      "num_tokens": 100468338.0,
      "step": 43815
    },
    {
      "entropy": 5.181240367889404,
      "epoch": 4.209414024975985,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003341290074663768,
      "loss": 4.6751,
      "mean_token_accuracy": 0.2510397583246231,
      "num_tokens": 100479666.0,
      "step": 43820
    },
    {
      "entropy": 5.012366724014282,
      "epoch": 4.209894332372719,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003340959282390742,
      "loss": 4.4865,
      "mean_token_accuracy": 0.2677836611866951,
      "num_tokens": 100491580.0,
      "step": 43825
    },
    {
      "entropy": 5.102659893035889,
      "epoch": 4.210374639769452,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033406284763973964,
      "loss": 4.599,
      "mean_token_accuracy": 0.2569173887372017,
      "num_tokens": 100503050.0,
      "step": 43830
    },
    {
      "entropy": 5.106734657287598,
      "epoch": 4.210854947166187,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003340297656691411,
      "loss": 4.6516,
      "mean_token_accuracy": 0.25140462070703506,
      "num_tokens": 100515880.0,
      "step": 43835
    },
    {
      "entropy": 5.085408163070679,
      "epoch": 4.21133525456292,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003339966823280467,
      "loss": 4.5505,
      "mean_token_accuracy": 0.2589405611157417,
      "num_tokens": 100526530.0,
      "step": 43840
    },
    {
      "entropy": 5.114202260971069,
      "epoch": 4.2118155619596545,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033396359761722453,
      "loss": 4.583,
      "mean_token_accuracy": 0.25839371979236603,
      "num_tokens": 100537544.0,
      "step": 43845
    },
    {
      "entropy": 5.129346942901611,
      "epoch": 4.212295869356388,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033393051153744274,
      "loss": 4.5919,
      "mean_token_accuracy": 0.25347372591495515,
      "num_tokens": 100549239.0,
      "step": 43850
    },
    {
      "entropy": 5.13774733543396,
      "epoch": 4.212776176753122,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00033389742408946937,
      "loss": 4.6342,
      "mean_token_accuracy": 0.25772848427295686,
      "num_tokens": 100560985.0,
      "step": 43855
    },
    {
      "entropy": 5.056217336654663,
      "epoch": 4.213256484149856,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003338643352740728,
      "loss": 4.5823,
      "mean_token_accuracy": 0.26155262589454653,
      "num_tokens": 100574793.0,
      "step": 43860
    },
    {
      "entropy": 5.193784570693969,
      "epoch": 4.21373679154659,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003338312450920211,
      "loss": 4.5922,
      "mean_token_accuracy": 0.26315720230340955,
      "num_tokens": 100586089.0,
      "step": 43865
    },
    {
      "entropy": 5.070697546005249,
      "epoch": 4.214217098943323,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003337981535440826,
      "loss": 4.5879,
      "mean_token_accuracy": 0.2616797789931297,
      "num_tokens": 100597031.0,
      "step": 43870
    },
    {
      "entropy": 5.115230226516724,
      "epoch": 4.214697406340058,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033376506063102556,
      "loss": 4.6263,
      "mean_token_accuracy": 0.25936448723077776,
      "num_tokens": 100607788.0,
      "step": 43875
    },
    {
      "entropy": 5.054033708572388,
      "epoch": 4.215177713736791,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003337319663536183,
      "loss": 4.5686,
      "mean_token_accuracy": 0.26921599060297013,
      "num_tokens": 100620828.0,
      "step": 43880
    },
    {
      "entropy": 5.009990167617798,
      "epoch": 4.2156580211335255,
      "grad_norm": 1.0,
      "learning_rate": 0.0003336988707126293,
      "loss": 4.4742,
      "mean_token_accuracy": 0.26638001054525373,
      "num_tokens": 100631903.0,
      "step": 43885
    },
    {
      "entropy": 5.1397112846374515,
      "epoch": 4.216138328530259,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003336657737088267,
      "loss": 4.6616,
      "mean_token_accuracy": 0.2552663668990135,
      "num_tokens": 100643805.0,
      "step": 43890
    },
    {
      "entropy": 5.037113189697266,
      "epoch": 4.216618635926993,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033363267534297896,
      "loss": 4.4855,
      "mean_token_accuracy": 0.26577842235565186,
      "num_tokens": 100654437.0,
      "step": 43895
    },
    {
      "entropy": 5.076376152038574,
      "epoch": 4.217098943323728,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003335995756158548,
      "loss": 4.5431,
      "mean_token_accuracy": 0.26272904574871064,
      "num_tokens": 100665182.0,
      "step": 43900
    },
    {
      "entropy": 5.070990943908692,
      "epoch": 4.217579250720461,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033356647452822236,
      "loss": 4.5743,
      "mean_token_accuracy": 0.25891211777925494,
      "num_tokens": 100675995.0,
      "step": 43905
    },
    {
      "entropy": 5.080698919296265,
      "epoch": 4.218059558117195,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033353337208085035,
      "loss": 4.5018,
      "mean_token_accuracy": 0.26677522510290147,
      "num_tokens": 100686951.0,
      "step": 43910
    },
    {
      "entropy": 5.048834753036499,
      "epoch": 4.218539865513929,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003335002682745072,
      "loss": 4.5181,
      "mean_token_accuracy": 0.27013237923383715,
      "num_tokens": 100697377.0,
      "step": 43915
    },
    {
      "entropy": 4.996704912185669,
      "epoch": 4.219020172910663,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003334671631099615,
      "loss": 4.5138,
      "mean_token_accuracy": 0.26139477342367173,
      "num_tokens": 100708259.0,
      "step": 43920
    },
    {
      "entropy": 5.087206125259399,
      "epoch": 4.2195004803073966,
      "grad_norm": 1.046875,
      "learning_rate": 0.000333434056587982,
      "loss": 4.5891,
      "mean_token_accuracy": 0.2601077720522881,
      "num_tokens": 100719042.0,
      "step": 43925
    },
    {
      "entropy": 5.107570457458496,
      "epoch": 4.219980787704131,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033340094870933714,
      "loss": 4.5707,
      "mean_token_accuracy": 0.25831174552440644,
      "num_tokens": 100730763.0,
      "step": 43930
    },
    {
      "entropy": 5.0880927562713625,
      "epoch": 4.220461095100864,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033336783947479566,
      "loss": 4.561,
      "mean_token_accuracy": 0.2570699334144592,
      "num_tokens": 100743323.0,
      "step": 43935
    },
    {
      "entropy": 5.123632001876831,
      "epoch": 4.220941402497599,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033333472888512626,
      "loss": 4.6141,
      "mean_token_accuracy": 0.25791123807430266,
      "num_tokens": 100754622.0,
      "step": 43940
    },
    {
      "entropy": 5.0102057456970215,
      "epoch": 4.221421709894332,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003333016169410978,
      "loss": 4.5108,
      "mean_token_accuracy": 0.26521526277065277,
      "num_tokens": 100766686.0,
      "step": 43945
    },
    {
      "entropy": 5.105022048950195,
      "epoch": 4.221902017291066,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033326850364347876,
      "loss": 4.617,
      "mean_token_accuracy": 0.25985223054885864,
      "num_tokens": 100778297.0,
      "step": 43950
    },
    {
      "entropy": 5.069679069519043,
      "epoch": 4.2223823246878,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003332353889930381,
      "loss": 4.5824,
      "mean_token_accuracy": 0.26190035939216616,
      "num_tokens": 100790046.0,
      "step": 43955
    },
    {
      "entropy": 5.218585062026977,
      "epoch": 4.222862632084534,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003332022729905448,
      "loss": 4.7219,
      "mean_token_accuracy": 0.24427978843450546,
      "num_tokens": 100801562.0,
      "step": 43960
    },
    {
      "entropy": 5.134727096557617,
      "epoch": 4.223342939481268,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003331691556367674,
      "loss": 4.5592,
      "mean_token_accuracy": 0.26532873064279555,
      "num_tokens": 100812184.0,
      "step": 43965
    },
    {
      "entropy": 5.051114702224732,
      "epoch": 4.223823246878002,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003331360369324749,
      "loss": 4.5846,
      "mean_token_accuracy": 0.25738628804683683,
      "num_tokens": 100823137.0,
      "step": 43970
    },
    {
      "entropy": 5.052187728881836,
      "epoch": 4.224303554274736,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003331029168784363,
      "loss": 4.5902,
      "mean_token_accuracy": 0.25528241842985155,
      "num_tokens": 100835717.0,
      "step": 43975
    },
    {
      "entropy": 5.165354824066162,
      "epoch": 4.22478386167147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033306979547542047,
      "loss": 4.6551,
      "mean_token_accuracy": 0.2567913308739662,
      "num_tokens": 100846991.0,
      "step": 43980
    },
    {
      "entropy": 5.091523885726929,
      "epoch": 4.225264169068204,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003330366727241965,
      "loss": 4.549,
      "mean_token_accuracy": 0.27386525422334673,
      "num_tokens": 100857727.0,
      "step": 43985
    },
    {
      "entropy": 5.060871267318726,
      "epoch": 4.2257444764649374,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003330035486255333,
      "loss": 4.5787,
      "mean_token_accuracy": 0.26121391355991364,
      "num_tokens": 100868903.0,
      "step": 43990
    },
    {
      "entropy": 5.078212118148803,
      "epoch": 4.226224783861672,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033297042318019995,
      "loss": 4.5585,
      "mean_token_accuracy": 0.2639164015650749,
      "num_tokens": 100880149.0,
      "step": 43995
    },
    {
      "entropy": 5.0577844142913815,
      "epoch": 4.226705091258405,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003329372963889654,
      "loss": 4.5108,
      "mean_token_accuracy": 0.26207399666309356,
      "num_tokens": 100892611.0,
      "step": 44000
    },
    {
      "entropy": 5.139122200012207,
      "epoch": 4.2271853986551395,
      "grad_norm": 0.890625,
      "learning_rate": 0.000332904168252599,
      "loss": 4.5994,
      "mean_token_accuracy": 0.25611788481473924,
      "num_tokens": 100904803.0,
      "step": 44005
    },
    {
      "entropy": 5.133137226104736,
      "epoch": 4.227665706051873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033287103877186966,
      "loss": 4.5525,
      "mean_token_accuracy": 0.2673941969871521,
      "num_tokens": 100915475.0,
      "step": 44010
    },
    {
      "entropy": 5.070243644714355,
      "epoch": 4.228146013448607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003328379079475466,
      "loss": 4.5583,
      "mean_token_accuracy": 0.2640482738614082,
      "num_tokens": 100927402.0,
      "step": 44015
    },
    {
      "entropy": 5.035873126983643,
      "epoch": 4.228626320845341,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003328047757803992,
      "loss": 4.5507,
      "mean_token_accuracy": 0.2604124411940575,
      "num_tokens": 100938461.0,
      "step": 44020
    },
    {
      "entropy": 5.0564323425292965,
      "epoch": 4.229106628242075,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003327716422711965,
      "loss": 4.5217,
      "mean_token_accuracy": 0.2658177524805069,
      "num_tokens": 100950410.0,
      "step": 44025
    },
    {
      "entropy": 5.172316074371338,
      "epoch": 4.2295869356388085,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033273850742070784,
      "loss": 4.6299,
      "mean_token_accuracy": 0.265496164560318,
      "num_tokens": 100961600.0,
      "step": 44030
    },
    {
      "entropy": 5.065817213058471,
      "epoch": 4.230067243035543,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003327053712297024,
      "loss": 4.5435,
      "mean_token_accuracy": 0.2574667617678642,
      "num_tokens": 100972654.0,
      "step": 44035
    },
    {
      "entropy": 5.053082704544067,
      "epoch": 4.230547550432276,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003326722336989497,
      "loss": 4.5642,
      "mean_token_accuracy": 0.26056842505931854,
      "num_tokens": 100983536.0,
      "step": 44040
    },
    {
      "entropy": 5.023456048965454,
      "epoch": 4.231027857829011,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003326390948292189,
      "loss": 4.4463,
      "mean_token_accuracy": 0.2782868415117264,
      "num_tokens": 100995307.0,
      "step": 44045
    },
    {
      "entropy": 4.987803506851196,
      "epoch": 4.231508165225744,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003326059546212795,
      "loss": 4.4377,
      "mean_token_accuracy": 0.2754312604665756,
      "num_tokens": 101007091.0,
      "step": 44050
    },
    {
      "entropy": 5.091264629364014,
      "epoch": 4.231988472622478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003325728130759009,
      "loss": 4.6017,
      "mean_token_accuracy": 0.2518177852034569,
      "num_tokens": 101017121.0,
      "step": 44055
    },
    {
      "entropy": 5.087401247024536,
      "epoch": 4.232468780019213,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003325396701938525,
      "loss": 4.5915,
      "mean_token_accuracy": 0.2557737618684769,
      "num_tokens": 101027835.0,
      "step": 44060
    },
    {
      "entropy": 4.973232221603394,
      "epoch": 4.232949087415946,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003325065259759039,
      "loss": 4.4538,
      "mean_token_accuracy": 0.2725698545575142,
      "num_tokens": 101039950.0,
      "step": 44065
    },
    {
      "entropy": 5.119759368896484,
      "epoch": 4.23342939481268,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003324733804228244,
      "loss": 4.6778,
      "mean_token_accuracy": 0.2519716054201126,
      "num_tokens": 101051417.0,
      "step": 44070
    },
    {
      "entropy": 5.151421594619751,
      "epoch": 4.233909702209414,
      "grad_norm": 1.125,
      "learning_rate": 0.00033244023353538367,
      "loss": 4.6266,
      "mean_token_accuracy": 0.25874412804841995,
      "num_tokens": 101063123.0,
      "step": 44075
    },
    {
      "entropy": 5.076116132736206,
      "epoch": 4.234390009606148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003324070853143514,
      "loss": 4.5482,
      "mean_token_accuracy": 0.25996452420949934,
      "num_tokens": 101074548.0,
      "step": 44080
    },
    {
      "entropy": 5.089324045181274,
      "epoch": 4.234870317002882,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033237393576049704,
      "loss": 4.5369,
      "mean_token_accuracy": 0.2601709797978401,
      "num_tokens": 101085606.0,
      "step": 44085
    },
    {
      "entropy": 5.042159986495972,
      "epoch": 4.235350624399616,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033234078487459016,
      "loss": 4.4924,
      "mean_token_accuracy": 0.2582436680793762,
      "num_tokens": 101096414.0,
      "step": 44090
    },
    {
      "entropy": 4.963140058517456,
      "epoch": 4.235830931796349,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003323076326574006,
      "loss": 4.3983,
      "mean_token_accuracy": 0.2751618310809135,
      "num_tokens": 101107708.0,
      "step": 44095
    },
    {
      "entropy": 5.073054313659668,
      "epoch": 4.236311239193084,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000332274479109698,
      "loss": 4.5467,
      "mean_token_accuracy": 0.2615604430437088,
      "num_tokens": 101119142.0,
      "step": 44100
    },
    {
      "entropy": 5.074555921554565,
      "epoch": 4.236791546589817,
      "grad_norm": 1.0,
      "learning_rate": 0.000332241324232252,
      "loss": 4.5498,
      "mean_token_accuracy": 0.263915029168129,
      "num_tokens": 101130095.0,
      "step": 44105
    },
    {
      "entropy": 5.046715021133423,
      "epoch": 4.2372718539865515,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00033220816802583247,
      "loss": 4.5444,
      "mean_token_accuracy": 0.2663159310817719,
      "num_tokens": 101141497.0,
      "step": 44110
    },
    {
      "entropy": 5.058362722396851,
      "epoch": 4.237752161383285,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003321750104912092,
      "loss": 4.5555,
      "mean_token_accuracy": 0.2633063837885857,
      "num_tokens": 101152773.0,
      "step": 44115
    },
    {
      "entropy": 5.097430419921875,
      "epoch": 4.238232468780019,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033214185162915185,
      "loss": 4.6043,
      "mean_token_accuracy": 0.25999863594770434,
      "num_tokens": 101163977.0,
      "step": 44120
    },
    {
      "entropy": 5.114026260375977,
      "epoch": 4.238712776176753,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033210869144043037,
      "loss": 4.5369,
      "mean_token_accuracy": 0.2581385880708694,
      "num_tokens": 101176206.0,
      "step": 44125
    },
    {
      "entropy": 5.07733850479126,
      "epoch": 4.239193083573487,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033207552992581474,
      "loss": 4.5429,
      "mean_token_accuracy": 0.2658164739608765,
      "num_tokens": 101188135.0,
      "step": 44130
    },
    {
      "entropy": 5.031521129608154,
      "epoch": 4.239673390970221,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00033204236708607476,
      "loss": 4.4936,
      "mean_token_accuracy": 0.2729038804769516,
      "num_tokens": 101199887.0,
      "step": 44135
    },
    {
      "entropy": 5.1248702049255375,
      "epoch": 4.240153698366955,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003320092029219804,
      "loss": 4.5882,
      "mean_token_accuracy": 0.26116538643836973,
      "num_tokens": 101211824.0,
      "step": 44140
    },
    {
      "entropy": 5.0928421974182125,
      "epoch": 4.240634005763689,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003319760374343016,
      "loss": 4.5833,
      "mean_token_accuracy": 0.25908476710319517,
      "num_tokens": 101223372.0,
      "step": 44145
    },
    {
      "entropy": 4.966018104553223,
      "epoch": 4.2411143131604225,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003319428706238085,
      "loss": 4.4578,
      "mean_token_accuracy": 0.2710027784109116,
      "num_tokens": 101233571.0,
      "step": 44150
    },
    {
      "entropy": 5.119378519058228,
      "epoch": 4.241594620557157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003319097024912709,
      "loss": 4.5669,
      "mean_token_accuracy": 0.2598879337310791,
      "num_tokens": 101245535.0,
      "step": 44155
    },
    {
      "entropy": 4.981774950027466,
      "epoch": 4.24207492795389,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000331876533037459,
      "loss": 4.5127,
      "mean_token_accuracy": 0.2664400592446327,
      "num_tokens": 101257517.0,
      "step": 44160
    },
    {
      "entropy": 5.052295160293579,
      "epoch": 4.242555235350625,
      "grad_norm": 1.078125,
      "learning_rate": 0.000331843362263143,
      "loss": 4.5107,
      "mean_token_accuracy": 0.2621279045939445,
      "num_tokens": 101268051.0,
      "step": 44165
    },
    {
      "entropy": 5.130923557281494,
      "epoch": 4.243035542747358,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033181019016909286,
      "loss": 4.6112,
      "mean_token_accuracy": 0.2597930908203125,
      "num_tokens": 101278774.0,
      "step": 44170
    },
    {
      "entropy": 5.0523998737335205,
      "epoch": 4.243515850144092,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003317770167560788,
      "loss": 4.554,
      "mean_token_accuracy": 0.2617679938673973,
      "num_tokens": 101289606.0,
      "step": 44175
    },
    {
      "entropy": 5.031502389907837,
      "epoch": 4.243996157540826,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033174384202487096,
      "loss": 4.5502,
      "mean_token_accuracy": 0.25363944172859193,
      "num_tokens": 101300274.0,
      "step": 44180
    },
    {
      "entropy": 5.049261474609375,
      "epoch": 4.24447646493756,
      "grad_norm": 1.0,
      "learning_rate": 0.0003317106659762397,
      "loss": 4.5308,
      "mean_token_accuracy": 0.25655580759048463,
      "num_tokens": 101312545.0,
      "step": 44185
    },
    {
      "entropy": 5.152039241790772,
      "epoch": 4.244956772334294,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033167748861095503,
      "loss": 4.6248,
      "mean_token_accuracy": 0.25480167120695113,
      "num_tokens": 101323862.0,
      "step": 44190
    },
    {
      "entropy": 5.023580646514892,
      "epoch": 4.245437079731028,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003316443099297874,
      "loss": 4.5533,
      "mean_token_accuracy": 0.2632831782102585,
      "num_tokens": 101335587.0,
      "step": 44195
    },
    {
      "entropy": 5.082656955718994,
      "epoch": 4.245917387127761,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033161112993350725,
      "loss": 4.5804,
      "mean_token_accuracy": 0.2609292149543762,
      "num_tokens": 101345825.0,
      "step": 44200
    },
    {
      "entropy": 5.123209047317505,
      "epoch": 4.246397694524496,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033157794862288467,
      "loss": 4.576,
      "mean_token_accuracy": 0.2594839632511139,
      "num_tokens": 101356641.0,
      "step": 44205
    },
    {
      "entropy": 5.1400515079498295,
      "epoch": 4.24687800192123,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033154476599869006,
      "loss": 4.5826,
      "mean_token_accuracy": 0.254146845638752,
      "num_tokens": 101369926.0,
      "step": 44210
    },
    {
      "entropy": 5.1509991645812985,
      "epoch": 4.247358309317963,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000331511582061694,
      "loss": 4.6501,
      "mean_token_accuracy": 0.25583566874265673,
      "num_tokens": 101380913.0,
      "step": 44215
    },
    {
      "entropy": 4.947863388061523,
      "epoch": 4.247838616714698,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003314783968126667,
      "loss": 4.4414,
      "mean_token_accuracy": 0.26619844883680344,
      "num_tokens": 101393209.0,
      "step": 44220
    },
    {
      "entropy": 5.026416015625,
      "epoch": 4.248318924111431,
      "grad_norm": 1.0,
      "learning_rate": 0.0003314452102523788,
      "loss": 4.5435,
      "mean_token_accuracy": 0.263668529689312,
      "num_tokens": 101404920.0,
      "step": 44225
    },
    {
      "entropy": 4.986096000671386,
      "epoch": 4.2487992315081655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033141202238160075,
      "loss": 4.4289,
      "mean_token_accuracy": 0.2725865185260773,
      "num_tokens": 101415822.0,
      "step": 44230
    },
    {
      "entropy": 5.116063165664673,
      "epoch": 4.249279538904899,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033137883320110296,
      "loss": 4.6154,
      "mean_token_accuracy": 0.26142917573451996,
      "num_tokens": 101427451.0,
      "step": 44235
    },
    {
      "entropy": 5.120642137527466,
      "epoch": 4.249759846301633,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003313456427116561,
      "loss": 4.6012,
      "mean_token_accuracy": 0.2603535458445549,
      "num_tokens": 101438787.0,
      "step": 44240
    },
    {
      "entropy": 5.046610164642334,
      "epoch": 4.250240153698367,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003313124509140308,
      "loss": 4.5382,
      "mean_token_accuracy": 0.2649527326226234,
      "num_tokens": 101450687.0,
      "step": 44245
    },
    {
      "entropy": 5.149113368988037,
      "epoch": 4.250720461095101,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033127925780899754,
      "loss": 4.5783,
      "mean_token_accuracy": 0.2610769048333168,
      "num_tokens": 101461827.0,
      "step": 44250
    },
    {
      "entropy": 4.993135166168213,
      "epoch": 4.2512007684918345,
      "grad_norm": 1.03125,
      "learning_rate": 0.000331246063397327,
      "loss": 4.4532,
      "mean_token_accuracy": 0.27829382121562957,
      "num_tokens": 101472288.0,
      "step": 44255
    },
    {
      "entropy": 4.983849716186524,
      "epoch": 4.251681075888569,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033121286767979,
      "loss": 4.4984,
      "mean_token_accuracy": 0.27298232913017273,
      "num_tokens": 101485140.0,
      "step": 44260
    },
    {
      "entropy": 5.059900999069214,
      "epoch": 4.252161383285302,
      "grad_norm": 0.953125,
      "learning_rate": 0.000331179670657157,
      "loss": 4.5182,
      "mean_token_accuracy": 0.2664122968912125,
      "num_tokens": 101496261.0,
      "step": 44265
    },
    {
      "entropy": 5.083528900146485,
      "epoch": 4.252641690682037,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003311464723301989,
      "loss": 4.6045,
      "mean_token_accuracy": 0.2513607829809189,
      "num_tokens": 101506822.0,
      "step": 44270
    },
    {
      "entropy": 5.069339656829834,
      "epoch": 4.25312199807877,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003311132726996866,
      "loss": 4.5425,
      "mean_token_accuracy": 0.2600252702832222,
      "num_tokens": 101519572.0,
      "step": 44275
    },
    {
      "entropy": 5.0726570129394535,
      "epoch": 4.253602305475504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003310800717663905,
      "loss": 4.5551,
      "mean_token_accuracy": 0.26894240379333495,
      "num_tokens": 101530785.0,
      "step": 44280
    },
    {
      "entropy": 5.089382314682007,
      "epoch": 4.254082612872239,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033104686953108173,
      "loss": 4.5658,
      "mean_token_accuracy": 0.26539135724306107,
      "num_tokens": 101542562.0,
      "step": 44285
    },
    {
      "entropy": 5.105643844604492,
      "epoch": 4.254562920268972,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003310136659945311,
      "loss": 4.6,
      "mean_token_accuracy": 0.25924497544765474,
      "num_tokens": 101555199.0,
      "step": 44290
    },
    {
      "entropy": 5.134792327880859,
      "epoch": 4.255043227665706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033098046115750946,
      "loss": 4.6334,
      "mean_token_accuracy": 0.25716025829315187,
      "num_tokens": 101567665.0,
      "step": 44295
    },
    {
      "entropy": 5.097273111343384,
      "epoch": 4.25552353506244,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003309472550207877,
      "loss": 4.6497,
      "mean_token_accuracy": 0.25497688800096513,
      "num_tokens": 101580152.0,
      "step": 44300
    },
    {
      "entropy": 5.025710487365723,
      "epoch": 4.256003842459174,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033091404758513686,
      "loss": 4.5051,
      "mean_token_accuracy": 0.26160633862018584,
      "num_tokens": 101592054.0,
      "step": 44305
    },
    {
      "entropy": 5.023127269744873,
      "epoch": 4.256484149855908,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033088083885132784,
      "loss": 4.4764,
      "mean_token_accuracy": 0.2679894581437111,
      "num_tokens": 101603303.0,
      "step": 44310
    },
    {
      "entropy": 5.070474529266358,
      "epoch": 4.256964457252642,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033084762882013176,
      "loss": 4.6082,
      "mean_token_accuracy": 0.2587990924715996,
      "num_tokens": 101615586.0,
      "step": 44315
    },
    {
      "entropy": 5.141975402832031,
      "epoch": 4.257444764649375,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00033081441749231944,
      "loss": 4.6502,
      "mean_token_accuracy": 0.25529417097568513,
      "num_tokens": 101627482.0,
      "step": 44320
    },
    {
      "entropy": 5.176727962493897,
      "epoch": 4.25792507204611,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003307812048686621,
      "loss": 4.6664,
      "mean_token_accuracy": 0.252188041806221,
      "num_tokens": 101637878.0,
      "step": 44325
    },
    {
      "entropy": 5.129313707351685,
      "epoch": 4.258405379442843,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003307479909499309,
      "loss": 4.6325,
      "mean_token_accuracy": 0.2660203009843826,
      "num_tokens": 101649498.0,
      "step": 44330
    },
    {
      "entropy": 5.118651294708252,
      "epoch": 4.2588856868395775,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003307147757368969,
      "loss": 4.6397,
      "mean_token_accuracy": 0.251424977183342,
      "num_tokens": 101660635.0,
      "step": 44335
    },
    {
      "entropy": 5.090250825881958,
      "epoch": 4.259365994236311,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033068155923033114,
      "loss": 4.5786,
      "mean_token_accuracy": 0.2656813427805901,
      "num_tokens": 101671997.0,
      "step": 44340
    },
    {
      "entropy": 5.032398986816406,
      "epoch": 4.259846301633045,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033064834143100495,
      "loss": 4.4869,
      "mean_token_accuracy": 0.2690926045179367,
      "num_tokens": 101682732.0,
      "step": 44345
    },
    {
      "entropy": 5.049794435501099,
      "epoch": 4.260326609029779,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003306151223396896,
      "loss": 4.5745,
      "mean_token_accuracy": 0.2643753409385681,
      "num_tokens": 101693978.0,
      "step": 44350
    },
    {
      "entropy": 5.114124870300293,
      "epoch": 4.260806916426513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003305819019571562,
      "loss": 4.578,
      "mean_token_accuracy": 0.2635829746723175,
      "num_tokens": 101705802.0,
      "step": 44355
    },
    {
      "entropy": 5.046844911575318,
      "epoch": 4.261287223823247,
      "grad_norm": 1.0,
      "learning_rate": 0.00033054868028417606,
      "loss": 4.5649,
      "mean_token_accuracy": 0.26593167930841444,
      "num_tokens": 101717362.0,
      "step": 44360
    },
    {
      "entropy": 5.028510808944702,
      "epoch": 4.261767531219981,
      "grad_norm": 0.875,
      "learning_rate": 0.00033051545732152047,
      "loss": 4.4786,
      "mean_token_accuracy": 0.26928257644176484,
      "num_tokens": 101729774.0,
      "step": 44365
    },
    {
      "entropy": 5.0090631484985355,
      "epoch": 4.262247838616715,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003304822330699609,
      "loss": 4.4943,
      "mean_token_accuracy": 0.2701886877417564,
      "num_tokens": 101740897.0,
      "step": 44370
    },
    {
      "entropy": 5.062175750732422,
      "epoch": 4.2627281460134485,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003304490075302686,
      "loss": 4.5355,
      "mean_token_accuracy": 0.2660201653838158,
      "num_tokens": 101752760.0,
      "step": 44375
    },
    {
      "entropy": 5.1107165813446045,
      "epoch": 4.263208453410183,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033041578070321494,
      "loss": 4.5388,
      "mean_token_accuracy": 0.262556491792202,
      "num_tokens": 101764389.0,
      "step": 44380
    },
    {
      "entropy": 5.08427882194519,
      "epoch": 4.263688760806916,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003303825525895714,
      "loss": 4.5747,
      "mean_token_accuracy": 0.2585351556539536,
      "num_tokens": 101776161.0,
      "step": 44385
    },
    {
      "entropy": 5.076357889175415,
      "epoch": 4.264169068203651,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033034932319010946,
      "loss": 4.558,
      "mean_token_accuracy": 0.25655672550201414,
      "num_tokens": 101787337.0,
      "step": 44390
    },
    {
      "entropy": 5.039587116241455,
      "epoch": 4.264649375600384,
      "grad_norm": 1.0,
      "learning_rate": 0.00033031609250560066,
      "loss": 4.5204,
      "mean_token_accuracy": 0.2636445090174675,
      "num_tokens": 101798616.0,
      "step": 44395
    },
    {
      "entropy": 5.073539686203003,
      "epoch": 4.265129682997118,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003302828605368164,
      "loss": 4.5675,
      "mean_token_accuracy": 0.2640291079878807,
      "num_tokens": 101810556.0,
      "step": 44400
    },
    {
      "entropy": 5.152903318405151,
      "epoch": 4.265609990393852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033024962728452826,
      "loss": 4.5912,
      "mean_token_accuracy": 0.2546470731496811,
      "num_tokens": 101821665.0,
      "step": 44405
    },
    {
      "entropy": 5.033795309066773,
      "epoch": 4.266090297790586,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003302163927495078,
      "loss": 4.6271,
      "mean_token_accuracy": 0.2599362090229988,
      "num_tokens": 101834629.0,
      "step": 44410
    },
    {
      "entropy": 5.053190422058106,
      "epoch": 4.26657060518732,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003301831569325267,
      "loss": 4.5182,
      "mean_token_accuracy": 0.2653361737728119,
      "num_tokens": 101844921.0,
      "step": 44415
    },
    {
      "entropy": 5.091714906692505,
      "epoch": 4.267050912584054,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033014991983435654,
      "loss": 4.5782,
      "mean_token_accuracy": 0.25716689974069595,
      "num_tokens": 101855942.0,
      "step": 44420
    },
    {
      "entropy": 4.970720911026001,
      "epoch": 4.267531219980787,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003301166814557691,
      "loss": 4.4291,
      "mean_token_accuracy": 0.2711897104978561,
      "num_tokens": 101866760.0,
      "step": 44425
    },
    {
      "entropy": 5.052925491333008,
      "epoch": 4.268011527377522,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033008344179753584,
      "loss": 4.552,
      "mean_token_accuracy": 0.2634182274341583,
      "num_tokens": 101877948.0,
      "step": 44430
    },
    {
      "entropy": 5.144617128372192,
      "epoch": 4.268491834774256,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033005020086042874,
      "loss": 4.7222,
      "mean_token_accuracy": 0.24762854725122452,
      "num_tokens": 101890038.0,
      "step": 44435
    },
    {
      "entropy": 5.078931617736816,
      "epoch": 4.268972142170989,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033001695864521934,
      "loss": 4.5488,
      "mean_token_accuracy": 0.26210779696702957,
      "num_tokens": 101901169.0,
      "step": 44440
    },
    {
      "entropy": 5.100177574157715,
      "epoch": 4.269452449567724,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032998371515267954,
      "loss": 4.6223,
      "mean_token_accuracy": 0.26291104555130007,
      "num_tokens": 101913254.0,
      "step": 44445
    },
    {
      "entropy": 5.127922105789184,
      "epoch": 4.269932756964457,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003299504703835812,
      "loss": 4.5757,
      "mean_token_accuracy": 0.2651260390877724,
      "num_tokens": 101923785.0,
      "step": 44450
    },
    {
      "entropy": 5.171281719207764,
      "epoch": 4.2704130643611915,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003299172243386961,
      "loss": 4.5964,
      "mean_token_accuracy": 0.2524265170097351,
      "num_tokens": 101935752.0,
      "step": 44455
    },
    {
      "entropy": 5.045860004425049,
      "epoch": 4.270893371757925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003298839770187961,
      "loss": 4.4539,
      "mean_token_accuracy": 0.27060349732637407,
      "num_tokens": 101946016.0,
      "step": 44460
    },
    {
      "entropy": 5.012447214126587,
      "epoch": 4.271373679154659,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032985072842465314,
      "loss": 4.5286,
      "mean_token_accuracy": 0.26129133254289627,
      "num_tokens": 101957646.0,
      "step": 44465
    },
    {
      "entropy": 5.045280790328979,
      "epoch": 4.271853986551393,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032981747855703915,
      "loss": 4.5332,
      "mean_token_accuracy": 0.2628894865512848,
      "num_tokens": 101967972.0,
      "step": 44470
    },
    {
      "entropy": 5.02220139503479,
      "epoch": 4.272334293948127,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032978422741672603,
      "loss": 4.5466,
      "mean_token_accuracy": 0.2585185572504997,
      "num_tokens": 101979839.0,
      "step": 44475
    },
    {
      "entropy": 5.153263330459595,
      "epoch": 4.2728146013448605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003297509750044858,
      "loss": 4.606,
      "mean_token_accuracy": 0.2624754771590233,
      "num_tokens": 101992148.0,
      "step": 44480
    },
    {
      "entropy": 5.018824672698974,
      "epoch": 4.273294908741595,
      "grad_norm": 1.0,
      "learning_rate": 0.0003297177213210907,
      "loss": 4.4825,
      "mean_token_accuracy": 0.2672931969165802,
      "num_tokens": 102003990.0,
      "step": 44485
    },
    {
      "entropy": 5.028857421875,
      "epoch": 4.273775216138328,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003296844663673124,
      "loss": 4.4997,
      "mean_token_accuracy": 0.26715411990880966,
      "num_tokens": 102015824.0,
      "step": 44490
    },
    {
      "entropy": 5.0509411811828615,
      "epoch": 4.2742555235350626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003296512101439232,
      "loss": 4.5237,
      "mean_token_accuracy": 0.2606134802103043,
      "num_tokens": 102026180.0,
      "step": 44495
    },
    {
      "entropy": 5.170035743713379,
      "epoch": 4.274735830931796,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032961795265169525,
      "loss": 4.6728,
      "mean_token_accuracy": 0.25170220881700517,
      "num_tokens": 102038569.0,
      "step": 44500
    },
    {
      "entropy": 5.189231967926025,
      "epoch": 4.27521613832853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003295846938914005,
      "loss": 4.6761,
      "mean_token_accuracy": 0.2517354145646095,
      "num_tokens": 102050822.0,
      "step": 44505
    },
    {
      "entropy": 5.090476989746094,
      "epoch": 4.275696445725265,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003295514338638114,
      "loss": 4.5659,
      "mean_token_accuracy": 0.2701396256685257,
      "num_tokens": 102061589.0,
      "step": 44510
    },
    {
      "entropy": 5.025851631164551,
      "epoch": 4.276176753121998,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003295181725696999,
      "loss": 4.5217,
      "mean_token_accuracy": 0.26774953305721283,
      "num_tokens": 102073206.0,
      "step": 44515
    },
    {
      "entropy": 5.104561614990234,
      "epoch": 4.276657060518732,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003294849100098383,
      "loss": 4.6408,
      "mean_token_accuracy": 0.2604869112372398,
      "num_tokens": 102084381.0,
      "step": 44520
    },
    {
      "entropy": 5.080414295196533,
      "epoch": 4.277137367915466,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003294516461849988,
      "loss": 4.5577,
      "mean_token_accuracy": 0.2577751114964485,
      "num_tokens": 102096597.0,
      "step": 44525
    },
    {
      "entropy": 5.154457664489746,
      "epoch": 4.2776176753122,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000329418381095954,
      "loss": 4.5843,
      "mean_token_accuracy": 0.25310677140951154,
      "num_tokens": 102107754.0,
      "step": 44530
    },
    {
      "entropy": 5.046805381774902,
      "epoch": 4.278097982708934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003293851147434758,
      "loss": 4.5762,
      "mean_token_accuracy": 0.25929042547941206,
      "num_tokens": 102119994.0,
      "step": 44535
    },
    {
      "entropy": 5.069373655319214,
      "epoch": 4.278578290105668,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032935184712833674,
      "loss": 4.5316,
      "mean_token_accuracy": 0.26754012554883955,
      "num_tokens": 102131637.0,
      "step": 44540
    },
    {
      "entropy": 5.127988195419311,
      "epoch": 4.279058597502401,
      "grad_norm": 1.0,
      "learning_rate": 0.0003293185782513093,
      "loss": 4.6004,
      "mean_token_accuracy": 0.2566997691988945,
      "num_tokens": 102143363.0,
      "step": 44545
    },
    {
      "entropy": 5.098930168151855,
      "epoch": 4.279538904899136,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032928530811316564,
      "loss": 4.5873,
      "mean_token_accuracy": 0.25568495988845824,
      "num_tokens": 102154613.0,
      "step": 44550
    },
    {
      "entropy": 5.124676370620728,
      "epoch": 4.280019212295869,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003292520367146784,
      "loss": 4.6556,
      "mean_token_accuracy": 0.2533414363861084,
      "num_tokens": 102166751.0,
      "step": 44555
    },
    {
      "entropy": 5.140055418014526,
      "epoch": 4.2804995196926034,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003292187640566199,
      "loss": 4.6156,
      "mean_token_accuracy": 0.2571514382958412,
      "num_tokens": 102178683.0,
      "step": 44560
    },
    {
      "entropy": 5.058174657821655,
      "epoch": 4.280979827089337,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032918549013976267,
      "loss": 4.55,
      "mean_token_accuracy": 0.2593823343515396,
      "num_tokens": 102191293.0,
      "step": 44565
    },
    {
      "entropy": 5.11408486366272,
      "epoch": 4.281460134486071,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003291522149648794,
      "loss": 4.5162,
      "mean_token_accuracy": 0.26007364839315417,
      "num_tokens": 102202886.0,
      "step": 44570
    },
    {
      "entropy": 5.127225112915039,
      "epoch": 4.281940441882805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032911893853274234,
      "loss": 4.668,
      "mean_token_accuracy": 0.25215446650981904,
      "num_tokens": 102214993.0,
      "step": 44575
    },
    {
      "entropy": 5.134648418426513,
      "epoch": 4.282420749279539,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003290856608441243,
      "loss": 4.551,
      "mean_token_accuracy": 0.26329675167798994,
      "num_tokens": 102225161.0,
      "step": 44580
    },
    {
      "entropy": 5.1288755416870115,
      "epoch": 4.282901056676272,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003290523818997978,
      "loss": 4.6148,
      "mean_token_accuracy": 0.2567339390516281,
      "num_tokens": 102236160.0,
      "step": 44585
    },
    {
      "entropy": 5.098296356201172,
      "epoch": 4.283381364073007,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032901910170053553,
      "loss": 4.5924,
      "mean_token_accuracy": 0.25824546366930007,
      "num_tokens": 102248055.0,
      "step": 44590
    },
    {
      "entropy": 5.06054654121399,
      "epoch": 4.28386167146974,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003289858202471101,
      "loss": 4.5628,
      "mean_token_accuracy": 0.2668904781341553,
      "num_tokens": 102259517.0,
      "step": 44595
    },
    {
      "entropy": 5.136925411224365,
      "epoch": 4.2843419788664745,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032895253754029415,
      "loss": 4.6707,
      "mean_token_accuracy": 0.25033715516328814,
      "num_tokens": 102271016.0,
      "step": 44600
    },
    {
      "entropy": 5.0967542171478275,
      "epoch": 4.284822286263209,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032891925358086063,
      "loss": 4.5221,
      "mean_token_accuracy": 0.26390642374753953,
      "num_tokens": 102282699.0,
      "step": 44605
    },
    {
      "entropy": 5.104277849197388,
      "epoch": 4.285302593659942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032888596836958206,
      "loss": 4.5791,
      "mean_token_accuracy": 0.26075673997402193,
      "num_tokens": 102293420.0,
      "step": 44610
    },
    {
      "entropy": 5.0682884693145756,
      "epoch": 4.285782901056677,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032885268190723134,
      "loss": 4.5505,
      "mean_token_accuracy": 0.26358385384082794,
      "num_tokens": 102305537.0,
      "step": 44615
    },
    {
      "entropy": 5.155219316482544,
      "epoch": 4.28626320845341,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003288193941945812,
      "loss": 4.6432,
      "mean_token_accuracy": 0.2568045198917389,
      "num_tokens": 102316163.0,
      "step": 44620
    },
    {
      "entropy": 5.086761522293091,
      "epoch": 4.286743515850144,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003287861052324047,
      "loss": 4.5424,
      "mean_token_accuracy": 0.2691433161497116,
      "num_tokens": 102327399.0,
      "step": 44625
    },
    {
      "entropy": 5.165593194961548,
      "epoch": 4.287223823246878,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003287528150214744,
      "loss": 4.7201,
      "mean_token_accuracy": 0.2511637479066849,
      "num_tokens": 102339755.0,
      "step": 44630
    },
    {
      "entropy": 5.083091974258423,
      "epoch": 4.287704130643612,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032871952356256336,
      "loss": 4.5296,
      "mean_token_accuracy": 0.2728526577353477,
      "num_tokens": 102350852.0,
      "step": 44635
    },
    {
      "entropy": 5.149697399139404,
      "epoch": 4.2881844380403455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003286862308564445,
      "loss": 4.6591,
      "mean_token_accuracy": 0.26196784526109695,
      "num_tokens": 102360942.0,
      "step": 44640
    },
    {
      "entropy": 5.057823133468628,
      "epoch": 4.28866474543708,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032865293690389076,
      "loss": 4.5739,
      "mean_token_accuracy": 0.26034584641456604,
      "num_tokens": 102373060.0,
      "step": 44645
    },
    {
      "entropy": 5.084482336044312,
      "epoch": 4.289145052833813,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003286196417056752,
      "loss": 4.5569,
      "mean_token_accuracy": 0.2599236011505127,
      "num_tokens": 102383320.0,
      "step": 44650
    },
    {
      "entropy": 5.03424038887024,
      "epoch": 4.289625360230548,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032858634526257077,
      "loss": 4.5476,
      "mean_token_accuracy": 0.26538113355636594,
      "num_tokens": 102394074.0,
      "step": 44655
    },
    {
      "entropy": 4.954782438278198,
      "epoch": 4.290105667627281,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003285530475753505,
      "loss": 4.3941,
      "mean_token_accuracy": 0.27461579740047454,
      "num_tokens": 102405745.0,
      "step": 44660
    },
    {
      "entropy": 5.012332487106323,
      "epoch": 4.290585975024015,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00032851974864478743,
      "loss": 4.5705,
      "mean_token_accuracy": 0.26174415498971937,
      "num_tokens": 102417273.0,
      "step": 44665
    },
    {
      "entropy": 5.149373531341553,
      "epoch": 4.291066282420749,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00032848644847165484,
      "loss": 4.5952,
      "mean_token_accuracy": 0.25891787111759185,
      "num_tokens": 102429096.0,
      "step": 44670
    },
    {
      "entropy": 5.077334022521972,
      "epoch": 4.291546589817483,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003284531470567256,
      "loss": 4.5345,
      "mean_token_accuracy": 0.2648103147745132,
      "num_tokens": 102440962.0,
      "step": 44675
    },
    {
      "entropy": 5.011559581756591,
      "epoch": 4.2920268972142175,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003284198444007731,
      "loss": 4.527,
      "mean_token_accuracy": 0.25950742810964583,
      "num_tokens": 102451560.0,
      "step": 44680
    },
    {
      "entropy": 5.097563552856445,
      "epoch": 4.292507204610951,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032838654050457047,
      "loss": 4.5933,
      "mean_token_accuracy": 0.2534120589494705,
      "num_tokens": 102462210.0,
      "step": 44685
    },
    {
      "entropy": 5.083821153640747,
      "epoch": 4.292987512007685,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003283532353688908,
      "loss": 4.6043,
      "mean_token_accuracy": 0.25666172057390213,
      "num_tokens": 102474016.0,
      "step": 44690
    },
    {
      "entropy": 5.0788942813873295,
      "epoch": 4.293467819404419,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032831992899450745,
      "loss": 4.5608,
      "mean_token_accuracy": 0.26434915959835054,
      "num_tokens": 102485787.0,
      "step": 44695
    },
    {
      "entropy": 5.113919830322265,
      "epoch": 4.293948126801153,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032828662138219375,
      "loss": 4.548,
      "mean_token_accuracy": 0.2618255838751793,
      "num_tokens": 102495961.0,
      "step": 44700
    },
    {
      "entropy": 5.0558568954467775,
      "epoch": 4.2944284341978864,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003282533125327228,
      "loss": 4.5473,
      "mean_token_accuracy": 0.25207003504037856,
      "num_tokens": 102507987.0,
      "step": 44705
    },
    {
      "entropy": 5.063663101196289,
      "epoch": 4.294908741594621,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003282200024468682,
      "loss": 4.5882,
      "mean_token_accuracy": 0.25601168125867846,
      "num_tokens": 102519689.0,
      "step": 44710
    },
    {
      "entropy": 5.127630138397217,
      "epoch": 4.295389048991354,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003281866911254031,
      "loss": 4.6147,
      "mean_token_accuracy": 0.2575711369514465,
      "num_tokens": 102530498.0,
      "step": 44715
    },
    {
      "entropy": 5.046017742156982,
      "epoch": 4.2958693563880885,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000328153378569101,
      "loss": 4.4748,
      "mean_token_accuracy": 0.26670549958944323,
      "num_tokens": 102541091.0,
      "step": 44720
    },
    {
      "entropy": 5.02529125213623,
      "epoch": 4.296349663784822,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003281200647787352,
      "loss": 4.4766,
      "mean_token_accuracy": 0.26291784197092055,
      "num_tokens": 102552905.0,
      "step": 44725
    },
    {
      "entropy": 4.969589519500732,
      "epoch": 4.296829971181556,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032808674975507937,
      "loss": 4.4902,
      "mean_token_accuracy": 0.26154722571372985,
      "num_tokens": 102565062.0,
      "step": 44730
    },
    {
      "entropy": 5.1193482875823975,
      "epoch": 4.29731027857829,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003280534334989067,
      "loss": 4.5528,
      "mean_token_accuracy": 0.26038527190685273,
      "num_tokens": 102575829.0,
      "step": 44735
    },
    {
      "entropy": 4.918604612350464,
      "epoch": 4.297790585975024,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032802011601099095,
      "loss": 4.397,
      "mean_token_accuracy": 0.2798259362578392,
      "num_tokens": 102587962.0,
      "step": 44740
    },
    {
      "entropy": 5.007194709777832,
      "epoch": 4.2982708933717575,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032798679729210555,
      "loss": 4.5122,
      "mean_token_accuracy": 0.26786320507526395,
      "num_tokens": 102599655.0,
      "step": 44745
    },
    {
      "entropy": 5.008139181137085,
      "epoch": 4.298751200768492,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000327953477343024,
      "loss": 4.4434,
      "mean_token_accuracy": 0.2613228037953377,
      "num_tokens": 102611004.0,
      "step": 44750
    },
    {
      "entropy": 5.122451257705689,
      "epoch": 4.299231508165226,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032792015616452,
      "loss": 4.587,
      "mean_token_accuracy": 0.2677863910794258,
      "num_tokens": 102622939.0,
      "step": 44755
    },
    {
      "entropy": 4.997371673583984,
      "epoch": 4.29971181556196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000327886833757367,
      "loss": 4.4714,
      "mean_token_accuracy": 0.26714710146188736,
      "num_tokens": 102633500.0,
      "step": 44760
    },
    {
      "entropy": 5.0536435604095455,
      "epoch": 4.300192122958694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003278535101223388,
      "loss": 4.5489,
      "mean_token_accuracy": 0.2673764944076538,
      "num_tokens": 102644741.0,
      "step": 44765
    },
    {
      "entropy": 5.110957288742066,
      "epoch": 4.300672430355427,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003278201852602091,
      "loss": 4.6727,
      "mean_token_accuracy": 0.2583541676402092,
      "num_tokens": 102657718.0,
      "step": 44770
    },
    {
      "entropy": 5.165959692001342,
      "epoch": 4.301152737752162,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032778685917175153,
      "loss": 4.5992,
      "mean_token_accuracy": 0.256752672791481,
      "num_tokens": 102669303.0,
      "step": 44775
    },
    {
      "entropy": 5.085873317718506,
      "epoch": 4.301633045148895,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032775353185773983,
      "loss": 4.4923,
      "mean_token_accuracy": 0.26354487985372543,
      "num_tokens": 102679655.0,
      "step": 44780
    },
    {
      "entropy": 4.971346139907837,
      "epoch": 4.302113352545629,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032772020331894776,
      "loss": 4.5458,
      "mean_token_accuracy": 0.26051697582006456,
      "num_tokens": 102690727.0,
      "step": 44785
    },
    {
      "entropy": 5.165554189682007,
      "epoch": 4.302593659942363,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003276868735561491,
      "loss": 4.6508,
      "mean_token_accuracy": 0.2572144716978073,
      "num_tokens": 102702557.0,
      "step": 44790
    },
    {
      "entropy": 5.221826505661011,
      "epoch": 4.303073967339097,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00032765354257011765,
      "loss": 4.6287,
      "mean_token_accuracy": 0.2514147162437439,
      "num_tokens": 102716724.0,
      "step": 44795
    },
    {
      "entropy": 5.13641390800476,
      "epoch": 4.303554274735831,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032762021036162735,
      "loss": 4.5736,
      "mean_token_accuracy": 0.2567949637770653,
      "num_tokens": 102727760.0,
      "step": 44800
    },
    {
      "entropy": 5.098563623428345,
      "epoch": 4.304034582132565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003275868769314521,
      "loss": 4.6535,
      "mean_token_accuracy": 0.24942730069160463,
      "num_tokens": 102739189.0,
      "step": 44805
    },
    {
      "entropy": 5.0372960567474365,
      "epoch": 4.304514889529298,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00032755354228036554,
      "loss": 4.5497,
      "mean_token_accuracy": 0.2707751512527466,
      "num_tokens": 102751515.0,
      "step": 44810
    },
    {
      "entropy": 5.106947231292724,
      "epoch": 4.304995196926033,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003275202064091419,
      "loss": 4.612,
      "mean_token_accuracy": 0.2627533584833145,
      "num_tokens": 102762593.0,
      "step": 44815
    },
    {
      "entropy": 5.05414719581604,
      "epoch": 4.305475504322766,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032748686931855495,
      "loss": 4.5247,
      "mean_token_accuracy": 0.2611304372549057,
      "num_tokens": 102775347.0,
      "step": 44820
    },
    {
      "entropy": 5.1641600131988525,
      "epoch": 4.3059558117195005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003274535310093788,
      "loss": 4.6273,
      "mean_token_accuracy": 0.2578777983784676,
      "num_tokens": 102785877.0,
      "step": 44825
    },
    {
      "entropy": 5.03865795135498,
      "epoch": 4.306436119116235,
      "grad_norm": 1.0,
      "learning_rate": 0.0003274201914823874,
      "loss": 4.4682,
      "mean_token_accuracy": 0.268404433131218,
      "num_tokens": 102797620.0,
      "step": 44830
    },
    {
      "entropy": 5.063390445709229,
      "epoch": 4.306916426512968,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032738685073835475,
      "loss": 4.4986,
      "mean_token_accuracy": 0.26714832335710526,
      "num_tokens": 102808210.0,
      "step": 44835
    },
    {
      "entropy": 5.006105518341064,
      "epoch": 4.307396733909703,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000327353508778055,
      "loss": 4.4566,
      "mean_token_accuracy": 0.2684914365410805,
      "num_tokens": 102820336.0,
      "step": 44840
    },
    {
      "entropy": 5.00662145614624,
      "epoch": 4.307877041306436,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00032732016560226235,
      "loss": 4.4723,
      "mean_token_accuracy": 0.2665442511439323,
      "num_tokens": 102832632.0,
      "step": 44845
    },
    {
      "entropy": 5.049189043045044,
      "epoch": 4.30835734870317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032728682121175077,
      "loss": 4.5429,
      "mean_token_accuracy": 0.2673038303852081,
      "num_tokens": 102843744.0,
      "step": 44850
    },
    {
      "entropy": 5.010554695129395,
      "epoch": 4.308837656099904,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032725347560729433,
      "loss": 4.5229,
      "mean_token_accuracy": 0.2676745310425758,
      "num_tokens": 102855529.0,
      "step": 44855
    },
    {
      "entropy": 5.042962265014649,
      "epoch": 4.309317963496638,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032722012878966744,
      "loss": 4.4878,
      "mean_token_accuracy": 0.2688456490635872,
      "num_tokens": 102867831.0,
      "step": 44860
    },
    {
      "entropy": 5.10370283126831,
      "epoch": 4.3097982708933715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032718678075964425,
      "loss": 4.5516,
      "mean_token_accuracy": 0.26277157068252566,
      "num_tokens": 102879207.0,
      "step": 44865
    },
    {
      "entropy": 5.046253633499146,
      "epoch": 4.310278578290106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003271534315179989,
      "loss": 4.5361,
      "mean_token_accuracy": 0.264535753428936,
      "num_tokens": 102891188.0,
      "step": 44870
    },
    {
      "entropy": 5.124748945236206,
      "epoch": 4.310758885686839,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032712008106550584,
      "loss": 4.5906,
      "mean_token_accuracy": 0.26350053697824477,
      "num_tokens": 102901250.0,
      "step": 44875
    },
    {
      "entropy": 5.114408302307129,
      "epoch": 4.311239193083574,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032708672940293923,
      "loss": 4.5996,
      "mean_token_accuracy": 0.2605550065636635,
      "num_tokens": 102913992.0,
      "step": 44880
    },
    {
      "entropy": 4.983069658279419,
      "epoch": 4.311719500480307,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032705337653107336,
      "loss": 4.4277,
      "mean_token_accuracy": 0.2737168475985527,
      "num_tokens": 102925165.0,
      "step": 44885
    },
    {
      "entropy": 5.13154296875,
      "epoch": 4.312199807877041,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00032702002245068267,
      "loss": 4.622,
      "mean_token_accuracy": 0.25852309018373487,
      "num_tokens": 102938112.0,
      "step": 44890
    },
    {
      "entropy": 5.074195432662964,
      "epoch": 4.312680115273775,
      "grad_norm": 1.0,
      "learning_rate": 0.00032698666716254155,
      "loss": 4.5756,
      "mean_token_accuracy": 0.25872686207294465,
      "num_tokens": 102950523.0,
      "step": 44895
    },
    {
      "entropy": 5.074637603759766,
      "epoch": 4.313160422670509,
      "grad_norm": 1.0,
      "learning_rate": 0.0003269533106674244,
      "loss": 4.5126,
      "mean_token_accuracy": 0.26998933851718904,
      "num_tokens": 102960984.0,
      "step": 44900
    },
    {
      "entropy": 5.058228302001953,
      "epoch": 4.3136407300672435,
      "grad_norm": 1.0,
      "learning_rate": 0.00032691995296610566,
      "loss": 4.5261,
      "mean_token_accuracy": 0.26462444961071013,
      "num_tokens": 102974348.0,
      "step": 44905
    },
    {
      "entropy": 4.991805601119995,
      "epoch": 4.314121037463977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032688659405935973,
      "loss": 4.4534,
      "mean_token_accuracy": 0.2723309427499771,
      "num_tokens": 102985927.0,
      "step": 44910
    },
    {
      "entropy": 5.0119647026062015,
      "epoch": 4.314601344860711,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003268532339479612,
      "loss": 4.5426,
      "mean_token_accuracy": 0.26474773734807966,
      "num_tokens": 102996961.0,
      "step": 44915
    },
    {
      "entropy": 5.017851066589356,
      "epoch": 4.315081652257445,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003268198726326846,
      "loss": 4.5295,
      "mean_token_accuracy": 0.2670641392469406,
      "num_tokens": 103008675.0,
      "step": 44920
    },
    {
      "entropy": 5.046277856826782,
      "epoch": 4.315561959654179,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003267865101143043,
      "loss": 4.5197,
      "mean_token_accuracy": 0.2694770351052284,
      "num_tokens": 103019249.0,
      "step": 44925
    },
    {
      "entropy": 5.019059753417968,
      "epoch": 4.316042267050912,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032675314639359505,
      "loss": 4.5608,
      "mean_token_accuracy": 0.2628733113408089,
      "num_tokens": 103029690.0,
      "step": 44930
    },
    {
      "entropy": 5.102677583694458,
      "epoch": 4.316522574447647,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003267197814713315,
      "loss": 4.5559,
      "mean_token_accuracy": 0.2615921899676323,
      "num_tokens": 103041273.0,
      "step": 44935
    },
    {
      "entropy": 5.066519641876221,
      "epoch": 4.31700288184438,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032668641534828807,
      "loss": 4.514,
      "mean_token_accuracy": 0.2718195587396622,
      "num_tokens": 103052351.0,
      "step": 44940
    },
    {
      "entropy": 4.9942389011383055,
      "epoch": 4.3174831892411145,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032665304802523965,
      "loss": 4.5035,
      "mean_token_accuracy": 0.2627409532666206,
      "num_tokens": 103064013.0,
      "step": 44945
    },
    {
      "entropy": 5.0468353748321535,
      "epoch": 4.317963496637848,
      "grad_norm": 1.0,
      "learning_rate": 0.00032661967950296084,
      "loss": 4.5769,
      "mean_token_accuracy": 0.26167799681425097,
      "num_tokens": 103074556.0,
      "step": 44950
    },
    {
      "entropy": 5.051586055755616,
      "epoch": 4.318443804034582,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003265863097822263,
      "loss": 4.5352,
      "mean_token_accuracy": 0.268604701757431,
      "num_tokens": 103084842.0,
      "step": 44955
    },
    {
      "entropy": 5.076031970977783,
      "epoch": 4.318924111431316,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00032655293886381083,
      "loss": 4.6038,
      "mean_token_accuracy": 0.2588009387254715,
      "num_tokens": 103097271.0,
      "step": 44960
    },
    {
      "entropy": 5.142802619934082,
      "epoch": 4.31940441882805,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003265195667484892,
      "loss": 4.6554,
      "mean_token_accuracy": 0.25755585730075836,
      "num_tokens": 103109370.0,
      "step": 44965
    },
    {
      "entropy": 5.119687080383301,
      "epoch": 4.3198847262247835,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032648619343703626,
      "loss": 4.5863,
      "mean_token_accuracy": 0.2625979855656624,
      "num_tokens": 103120156.0,
      "step": 44970
    },
    {
      "entropy": 4.985166311264038,
      "epoch": 4.320365033621518,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003264528189302267,
      "loss": 4.5064,
      "mean_token_accuracy": 0.2732252299785614,
      "num_tokens": 103131462.0,
      "step": 44975
    },
    {
      "entropy": 5.042249441146851,
      "epoch": 4.320845341018252,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00032641944322883557,
      "loss": 4.5183,
      "mean_token_accuracy": 0.2683578163385391,
      "num_tokens": 103143001.0,
      "step": 44980
    },
    {
      "entropy": 5.066150856018067,
      "epoch": 4.321325648414986,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003263860663336376,
      "loss": 4.5411,
      "mean_token_accuracy": 0.2563976854085922,
      "num_tokens": 103155397.0,
      "step": 44985
    },
    {
      "entropy": 5.097312879562378,
      "epoch": 4.32180595581172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032635268824540774,
      "loss": 4.5751,
      "mean_token_accuracy": 0.26273681223392487,
      "num_tokens": 103166137.0,
      "step": 44990
    },
    {
      "entropy": 5.05346941947937,
      "epoch": 4.322286263208453,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032631930896492097,
      "loss": 4.5482,
      "mean_token_accuracy": 0.26483093947172165,
      "num_tokens": 103177572.0,
      "step": 44995
    },
    {
      "entropy": 5.004176235198974,
      "epoch": 4.322766570605188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003262859284929522,
      "loss": 4.4679,
      "mean_token_accuracy": 0.26414899677038195,
      "num_tokens": 103189847.0,
      "step": 45000
    },
    {
      "epoch": 4.322766570605188,
      "eval_entropy": 4.850394265701688,
      "eval_loss": 4.713542938232422,
      "eval_mean_token_accuracy": 0.26084024209378404,
      "eval_num_tokens": 103189847.0,
      "eval_runtime": 26.5254,
      "eval_samples_per_second": 1237.118,
      "eval_steps_per_second": 154.644,
      "step": 45000
    }
  ],
  "logging_steps": 5,
  "max_steps": 104090,
  "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": 1.57890279884544e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}